Archives: NNT Reviews

Source

Long B, Gottlieb M. Video laryngoscopy versus fiberoptic bronchoscopy for awake tracheal intubation. Academic Emergency Medicine. Published online November 13, 2024:acem.15051.

Study Population: 873 participants from 11 studies of adults undergoing awake intubation

Efficacy Endpoints

Time to intubation, lower risk of oxygen desaturation <90%, successful intubation on first attempt, patient-reported satisfaction

Harm Endpoints

Failed intubation and adverse events arising from intubation procedure (sore throat/hoarseness)

Narrative

Tracheal intubation is a high-risk procedure particularly as it may require sedation and paralysis.1^, 2^, 3 Awake intubation is a potential option for patients with anticipated or known difficult anatomic or physiologic airways, as it allows for spontaneous breathing and maintenance of airway tone by utilizing topical anesthetics with or without sedative agents in the absence of paralytic administration during the intubation attempt.2^, 3^, 4^, 5 Indications for awake intubation include patients with significant risk of a difficult anatomic or physiologic airway (e.g., angioedema, Ludwig's angina, oral mass) who do not require immediate airway protection, are able to tolerate the procedure, and are at low risk of vomiting. One method of awake intubation is using fiberoptic bronchoscopy, which requires familiarity with the necessary equipment, extensive knowledge of airway anatomy, and an understanding of local anesthesia and sedation strategies.6^, 7^, 8^, 9 On the other hand, video laryngoscopes provide improved glottic visualization and first-pass success, and they are typically easier to use and master compared to fiberoptic bronchoscopes.1^, 3^, 10 Thus, video laryngoscopes have been used as an alternative device for use in awake intubation. Two prior meta-analyses comparing the two techniques found shorter intubation time with video laryngoscopy, but there was no difference in other outcomes. Both meta-analyses reported significant heterogeneity among the trials.11^, 12

The updated systematic review summarized here included 11 randomized controlled trials (RCTs; n = 873 participants) and compared video laryngoscopy and fiberoptic bronchoscopy for patients undergoing awake intubation.13 The primary outcome was time to intubation. Secondary outcomes included rate of successful intubation on first attempt, rates of failed intubation, patient-reported satisfaction, oxygen desaturation <90%, and any other complications or adverse events (sore throat/hoarseness) from the intubation procedure. They utilized trial sequential analysis to evaluate the conclusiveness of the evidence concerning intubation time.

All 11 RCTs were published between 2012 and 2023 and evaluated awake intubation performed by anesthesiologists for elective bariatric, oral and maxillofacial, otolaryngology, cervical spine, gynecologic, abdominal, or urologic surgeries.13 Two trials utilized the nasal route for intubation, while the remaining nine RCTs used the oral route. Video laryngoscope devices included AceScope, Airtraq, Bullard, C-MAC D-BLADE, GlideScope, McGrath, and Pentax AWS.

Compared to fiberoptic bronchoscopy, video laryngoscopy was associated with reduced time to intubation (standardized mean difference [SMD] −1.97 min, 95% CI −2.78 to −1.15 min, 10 studies). Three studies evaluated GlideScope and found reduced time to intubation (SMD −2.50 min, 95% CI −4.87 to −0.13 min). Other video laryngoscope devices were also associated with reduced time to intubation (SMD −1.77 min, 95% CI −2.66 to −0.87 min). Video laryngoscopy was associated with reduced risk of oxygen desaturation <90% (risk ratio [RR] −0.70, 95% CI −1.40 to −0.01, seven studies). Regarding oxygen desaturation <90%, video laryngoscopy was associated with an absolute risk reduction of 5.2%, number needed to treat 20, when compared to fiberoptic laryngoscopy. There was no difference in first-attempt successful intubation (RR 0.01, 95% CI −0.06 to 0.09, nine studies), failed intubation (RR 0.46, 95% CI −0.52 to 1.44, nine studies), or sore throat/hoarseness (RR 0.07, 95% CI −0.48 to 0.62, 3 studies). Patient satisfaction did not differ in seven studies. Trial sequential analysis for time to intubation suggested the results were conclusive.13

Caveats

There are several factors that influence the interpretation of these results.13 First, there was significant heterogeneity among the included trials involving several factors including the type of intubation device, indication for awake intubation, operator experience, types of medications used for sedation and topical anesthesia, and trial inclusion criteria. This heterogeneity limited subgroup analysis, particularly for the video laryngoscope devices used. Second, the analysis of intubation time as the primary outcome varied among the included studies, with five RCTs using mean and standard deviation and the other five trials using median and interquartile range. Third, all RCTs were conducted in the anesthesiology setting in elective surgeries. Indeed, five studies did not require inclusion of patients with known or anticipated difficult airways, which is the primary indication for awake intubation. This limits the application of the findings of this systematic review to the emergency department (ED) setting. Fourth, the sedation protocols and sedation targets varied among the included trials. Ketamine, a commonly utilized medication in the ED setting, was not used in any of the included RCTs. Rather, the included studies utilized a variety of medications such as remifentanil, midazolam, or propofol. Ten trials utilized topical anesthesia with lidocaine, while one used transtracheal injection and two used superior laryngeal nerve blockade and tracheal block. Fifth, blinding of operators and outcomes assessors was not possible, which may lead to bias. Finally, the definition of the airway operator as “expert” varied with both devices across all the included studies, with different levels of experience in the included trials.

Based on current data, video laryngoscopy may be associated with reduced time to intubation and risk of oxygen desaturation <90% compared to fiberoptic bronchoscopy in patients undergoing awake intubation for elective surgery. However, these results may not be generalizable to the ED setting. In addition, significant heterogeneity is a serious validity threat that limits our ability to draw conclusions from the existing evidence. Therefore, we have assigned a color recommendation of yellow (more data needed) for the evaluation of video laryngoscopy compared to fiberoptic bronchoscopy in awake intubation. Importantly, there is no consensus that fiberoptic laryngoscopy is more successful than video laryngoscopy. For predicted difficult airways where awake intubation is being considered without paralysis during the attempt in the ED setting, video laryngoscopy appears to be as effective with lower complications rates.14 Thus, emergency physicians should consider video laryngoscopy as an option for awake intubation, rather than delaying a necessary procedure for fiberoptic equipment. Further data are needed in the ED and critical care settings using clearly defined standardized sedation and anesthesia protocols and outcomes.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

November 26, 2024

Source

Johari F, Johari P. Cranberry products for preventing urinary tract infections. afp. 2024;110(1):23A-23B.

Study Population: 50 trials including 8,857 patients considered susceptible to recurrent urinary tract infections (UTIs)

Efficacy Endpoints

Number of participants with symptomatic, culture-verified UTIs

Harm Endpoints

Adverse events (e.g., gastrointestinal problems)

Narrative

UTIs affect more than 150 million individuals globally each year and impose a substantial financial burden on health care systems.1 The annual incidence of acute uncomplicated UTIs reaches 7% among women of all ages, and 30% of these women may experience a recurrence within 6 to 12 months.2^, 3

For decades, cranberry products have been used in the treatment and prevention of UTIs. Cranberries' mechanism of action uses proanthocyanidins, which inhibit the adherence of p-fimbriated bacteria (e.g., Escherichia coli) to urothelial cells.4 The systematic review and meta-analysis discussed here explores the effectiveness of cranberry products in preventing UTI.

The Cochrane review included 50 randomized controlled trials (N = 8,857) and compared cranberry products with placebo, no treatment, or other interventions (e.g., antibiotics, probiotics) for prevention of UTI.5 The review focused on six populations that are susceptible to UTI: women with recurrent UTI (usually two or more in the past 12 months), older men and women in institutions, pregnant women, children, adults with neuromuscular dysfunction of the bladder and incomplete bladder emptying, and people susceptible to UTIs following an intervention (e.g., bladder radiotherapy, urogenital surgery, kidney transplantation).5

The Cochrane review showed moderate-certainty evidence that compared with placebo or control (water or nonspecific treatment), cranberry products reduced the risk of UTI in women with a history of recurrent UTI (risk ratio [RR] = 0.74; 95% CI, 0.55 to 0.99; absolute risk reduction [ARR] = 6.3%; number needed to treat [NNT] = 16) and in children (RR = 0.53; 95% CI, 0.36 to 0.78; ARR = 13.6%; NNT = 8). The review showed low-certainty evidence of benefit for those susceptible to UTI due to an intervention (RR = 0.47; 95% CI, 0.37 to 0.61; ARR = 12.2%; NNT = 9). However, in older men and women who are institutionalized, pregnant women, and adults with neuromuscular bladder dysfunction with incomplete bladder emptying, no apparent treatment benefit was observed.5

The most commonly reported harms associated with cranberry products were gastrointestinal adverse effects; moderate-certainty evidence demonstrated no difference between the groups.

Caveats

The trials used different cranberry forms (i.e., juice, tablet, capsule) and amounts; therefore, there is no consensus dose of proanthocyanidins for UTI risk reduction. The available information is derived from small-scale studies, which introduces uncertainty. Statistical heterogeneity among the trials was mostly low or moderate. Approximately one-half of the trials reported intervention compliance rates, and no apparent relationship was found between compliance and the efficacy end points.5

About one-half of the trials in the review demonstrated low risk of bias, with selection bias noted as the major issue. Most of the included trials compared cranberry products with no treatment, water, or placebo. The Cochrane review found insufficient data to compare cranberry products with antibiotics or probiotics for UTI risk reduction.4

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Fatima Johari, MD; Paniz Johari, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

November 21, 2024

Source

Long B, Gottlieb M. Ketamine versus etomidate for induction of intubation in critically ill patients. Academic Emergency Medicine. 2024;31(9):937-938.

Study Population: 2978 participants from eight studies of critically ill adults sedated for endotracheal intubation

Efficacy Endpoints

All-cause mortality at longest follow-up available

Harm Endpoints

Not reported

Narrative

Endotracheal intubation (ETI) is a common procedure performed in the prehospital, emergency department, and intensive care unit settings that carries a significant risk of morbidity and mortality.1^, 2 Rapid sequence intubation involves the use of induction agents (sedatives), of which ketamine and etomidate are two of the most commonly used. It has been suggested that the delayed effects of etomidate may be problematic due to increased rates of adrenal insufficiency.3^, 4^, 5 However, several studies have compared the two with conflicting results.4^, 5

The systematic review summarized here included seven randomized controlled trials and one propensity-matched study (n = 2978 participants) between 2009 and 2023 of critically ill adult patients undergoing emergency ETI who received ketamine or etomidate.6 Here we summarize only the data from randomized trials, comparing ketamine with other sedatives for sedation in patients with critical illness. The primary outcome was mortality at the longest available follow-up. Secondary outcomes included Sequential Organ Failure Assessment (SOFA) score, vasopressor-free days, ventilator-free days, postsedation mean arterial pressure, and successful intubation on first attempt. The review utilized a Bayesian random-effects meta-analysis, reporting posterior probability distributions for relative treatment effects using risk ratios. The authors also performed a sensitivity analysis for the primary outcome using a frequentist approach with a Mantel–Haenszel random-effects model.

Of the included studies, four trials were conducted in the United States, one trial was in France, one was in the Netherlands, and one was in Thailand. All studies except one were single center.7 Ketamine dosing ranged from 1 to 2 mg/kg in all studies but one, which combined 0.5 mg/kg ketamine with 0.5 mg/kg propofol.8 The time point of mortality assessment was hospital discharge in three studies, 28 days in three studies, and 30 days in one study.

Compared to etomidate, ketamine did not reduce mortality (RR 0.96, 95% CI 0.8–1.1). No difference was found in any secondary outcome including SOFA score, vasopressor-free days, ventilator-free days, blood pressure, and first-attempt success. On Bayesian analysis of trial results, the authors report a probability of 68.6% that ketamine reduced mortality by up to 1% and a probability of 41.6% that ketamine reduced mortality by ≥1% when compared to etomidate.

Caveats

There are several factors that complicate interpreting these results.6 First, while the review utilized Bayesian methods, this has limitations including the need for estimating prior probability (the likelihood of one agent being worse or better based on information available before the analysis), which can be difficult and subjective but directly informs numerical results. The review found ketamine has a 68.6% probability of lowering mortality by up to 1% and a probability of 41.6% that ketamine reduced mortality by ≥1% when compared to etomidate, but a traditional frequentist interpretation would characterize this as “no difference” (see comments below for more on Bayesian vs. frequentist). Second, most larger trials were open label, introducing bias. Third, there was variation in the timepoints for mortality assessment in the included studies, with assessment including 28 days, 30 days, and at hospital discharge. Fourth, there was no standardization of peri-intubation interventions (e.g., paralytics, opioids, vasopressors), introducing further potential confounders. Fifth, one study combined ketamine with propofol, confounding the drug's effects.8

Of note, a 2022 meta-analysis of nine studies found less postintubation hypotension with etomidate, but included six retrospective studies,4 while a 2023 meta-analysis of 11 trials found increased mortality with etomidate but evaluated multiple agents (propofol, midazolam, thiopental, and ketamine).5 Based on weak methodology in the first review, and the desire to focus on comparing the two agents head to head, we did not summarize these meta-analyses.

On the reliability and utility of “Bayesian” versus traditional “frequentist” methods of analyzing, the issue is hotly debated. For the purposes of this paper, one distinction is crucial. Traditional frequentist methods demand that any difference in numerical results cross a certain threshold (usually 95% mathematical likelihood) to be considered a “true” difference, providing a dichotomous (yes/no) answer. Conversely, Bayesian methods assign a sliding scale of probability to the existence of a difference, providing a percentage likelihood. In both cases; however, the calculations are embedded with assumptions such as a total absence of bias and methodologic perfection in the underlying data. Of course, this might not be true, but mathematical representations of study results are two dimensional—they are just numbers. This is why studies require careful, expert interpretation, and numerical results must be placed in context.

Based on current data it is unclear whether ketamine lowers mortality compared to etomidate. Therefore, we have selected a color recommendation of yellow (equal efficacy) for the use of ketamine in critically ill adult patients requiring induction for ETI. Further data are needed using clearly defined patient populations, outcomes, and standardized peri-induction management and dosing of etomidate and ketamine.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

November 12, 2024

Source

MacDonald BJ, Turgeon RD, McCormack J. Metformin for Type 2 diabetes mellitus. Academic Emergency Medicine. 2024;31(8):832-834.

Study Population: 17,470 patients with type 2 diabetes mellitus from 14 studies

Efficacy Endpoints

Death, cardiovascular death, myocardial infarction, stroke, peripheral vascular events, amputation

Harm Endpoints

Severe hypoglycemia (requiring help or intervention), gastrointestinal events (nausea, vomiting, diarrhea, or abdominal discomfort)

Narrative

Metformin has historically been the first-line drug option for Type 2 diabetes mellitus, likely based on the UK Prospective Diabetes Study (UKPDS-34 trial), which showed benefits in mortality and myocardial infarction with metformin compared to diet control in overweight patients.1 The finding, however, was muddied by a UKPDS substudy showing metformin increased mortality when added to a sulfonylurea. Neither finding was explained or replicated.1 To better assess metformin for Type 2 diabetes, we examined the evidence from systematic reviews and randomized controlled trials (RCTs) to determine the benefits and harms of metformin in management of Type 2 diabetes compared to standard care or placebo.

The primary source of this evidence-based summary is a 2012 meta-analysis of 13 studies by Boussageon et al.2 including 13,110 patients with Type 2 diabetes mellitus. Study participants had a mean age of 57 years, 47% were female, mean body mass index was 30, and the average hemoglobin A1c was 8.2%. Comparisons included metformin versus diet alone or placebo, sulfonylurea plus metformin versus sulfonylurea alone, and metformin withdrawal versus continuation.

The authors found no significant reduction in any efficacy outcome (death, cardiovascular death, myocardial infarction, stroke, peripheral vascular events, and amputation), and no increase in severe hypoglycemic episodes. Adverse events were not reported. Separately, a single trial comparing metformin to rosiglitazone or glyburide showed a 16% absolute increase in gastrointestinal adverse events (number needed to harm [NNH] 6).3

Caveats

A network meta-analysis published in 2023 by Shi et al.4 also found metformin did not reduce death or cardiovascular outcomes but did not report which trials contributed to this finding. The authors would not provide the information when contacted; therefore, we could not confirm or replicate their results. A Cochrane review from 2020 did not find any qualifying studies comparing metformin to placebo or standard care and reporting major clinical outcomes.5 Three other reviews of metformin were excluded because they either included observational studies6^, , 7 (without separate reporting of RCTs) or included trials that did not answer the clinical question of interest8 (i.e., included trials of patients without diabetes [e.g., with prediabetes] and pooled together placebo- and active-controlled trials).

These findings contrast with those of UKPDS-34, which demonstrated significant reductions in mortality and myocardial infarction compared to diet alone.1 While UKPDS-34 had a low risk of bias, there was uncertainty in the results, with wide CIs for mortality (relative risk [RR] 0.6, 95% CI 0.5–0.9, absolute difference 7% over 11 years, number needed to treat [NNT] 14) and myocardial infarction (RR 0.6, 95% CI 0.4–0.9, absolute difference 6% over 11 years, NNT 16). In a separate UKPDS trial, metformin plus sulfonylurea increased mortality compared to a sulfonylurea alone (RR 2.0, 95% CI 1.0–2.5, absolute risk difference 6% over 7 years, NNH 17), with no difference in myocardial infarctions.1 Regarding safety, UKPDS-34 found no increase in severe hypoglycemia or gastrointestinal events compared to diet alone.

The primary source of uncertainty in the metformin evidence seems to be imprecision: the CIs are too wide to rule out clinically important benefits or harms. For instance, in the review by Boussageon et al. the 95% CI for the RR of mortality with metformin was 0.8–1.3.2 Regarding quality of the evidence, the authors rated all trials as 3 or 4 out of 5 on the Jadad score (indicating high risk of bias in at least one domain).9 The three systematic reviews also had notable shortcomings. Boussageon et al. made the questionable decision to pool trials comparing metformin to diet with trials adding metformin to other drugs. The review by Shi et al. lacked transparency. The Cochrane review was unable to find data they considered reliable enough on clinical outcomes.

This uncertainty is reflected in some guidelines. The American Diabetes Association recommends either sodium-glucose cotransporter-2 (SGLT2) or glucagon-like peptide (GLP)-1 drugs for improving heart and kidney outcomes in high-risk patients10 while metformin is described as providing “potential benefit.” Similar recommendations are made by the American Association of Clinical Endocrinology,11 which recommends SGLT2 or GLP-1 drugs ahead of or without metformin.12^, 13 However, metformin continues to be the most prescribed medication for Type 2 diabetes in practice.14^, 15

Finally, it is important to note that while metformin may have unclear benefits, few alternatives are better. Within the review by Shi et al., nine of 11 other drugs did not reduce cardiovascular outcomes.4 Furthermore, sulfonylureas and insulin cause weight gain16 and severe hypoglycemia,4 while thiazolidinediones4 and some dipeptidyl peptidase-4 inhibitors17^, 18^, 19 increase heart failure. SGLT2 and GLP-1 drugs are the only ones shown in meta-analyses to reduce death and cardiovascular events, but also cause harms (SGLT2s4 increase genital infections and ketoacidosis and GLP-1s4^, 13 cause gastrointestinal events).

In summary, while the effects of metformin were not statistically significant, the broad CIs cannot rule out potentially meaningful clinical benefits or harms. We have therefore chosen a yellow color recommendation to reflect the uncertainty of metformin's effects. High-quality RCTs are needed to determine if common practice is benefiting people with Type 2 diabetes prescribed metformin.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Blair J. MacDonald BA, PharmD; Ricky D. Turgeon BSc (Pharm), ACPR, PharmD; James McCormack BSc, BSc (Pharm), PharmD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

October 24, 2024

Source

Long B, Gottlieb M. Fluid volumes in adults with sepsis. Academic Emergency Medicine. Published online April 3, 2024:acem.14912.

Study Population: 4,006 participants in 13 trials of adult patients with sepsis

Efficacy Endpoints

All-cause mortality

Harm Endpoints

Serious adverse events

Narrative

Sepsis, defined by organ dysfunction and a dysregulated response to infection, accounts for up to one in six hospital admissions in the United States and nearly 270,000 deaths annually.1^, 2 Treatment includes source control, antibiotics, and ensuring oxygenated blood flow supported by intravenous (IV) fluid administration. However, limited evidence on “more versus less” IV fluids has led to conflicting, uncertain recommendations.3^, 4^, 5^, 6 The 2021 international guidelines for management of sepsis and septic shock “Surviving Sepsis Campaign” suggests administering 30 mL/kg of fluids in the first 3 h.3 However, other studies have shown either no benefit or potentially increased mortality with such a strategy.4^, 5^, 6 A systematic review published in 2020 found no difference in mortality between lower and higher fluid volumes,7 but several large, high-quality randomized trials have been published since then.8^, 9^, 10^, 11

The review summarized here included 13 trials of 4006 adult participants with sepsis.12 The review included trials comparing fluid strategies intended to “obtain a separation in IV fluid volumes.” The authors focused on clinical outcomes (e.g., mortality) and excluded trials comparing different fluids and those focused on burn patients or severe blood loss. The primary outcomes were assessed at roughly 90 days and included (1) mortality, (2) adverse events, and (3) health-related quality of life. Secondary outcomes included duration and need for mechanical ventilation, need for renal replacement therapy (e.g., dialysis), and acute kidney injury. Additional outcomes included blood products, intensive care unit (ICU) length of stay, and hospital length of stay.

All 13 included trials were published between 2015 and 2023. Seven were multicenter, 10 were conducted in the ICU, and three in the emergency department (ED). All enrolled septic patients with hypoperfusion, severe sepsis, or septic shock.

Meta-analysis of all 13 trials found that less versus more fluid had no effect on mortality (relative risk [RR] 0.98, 95% confidence interval [CI] 0.9–1.1, moderate-quality evidence) or serious adverse events (RR 0.95, 95% CI 0.8–1.1, low-quality evidence).12 No data were reported on health-related quality of life, and no difference was found in any other outcome. The average fluid volume for resuscitation was 1679 mL in the less fluid group and 2775 mL in the more fluid group. The mean total volume of fluid was 6950 mL in the less fluid group and 7828 mL in the more fluid group.

Caveats

There are several important considerations when interpreting these results. First, five of 13 trials were unable to separate fluid volumes between the groups. However, analysis of only trials with separation also found no differences. Moreover, analysis of only high-quality trials found no difference in mortality. Second, definitions of sepsis and septic shock varied across studies. Third, the patient populations and interventions were significantly different. Seven trials included ICU patients only, three ED patients only, and three combined both. Importantly, several trials used complex protocols with a bundle of interventions, and in most, patients received IV fluids before inclusion. Crystalloids were the most common fluid; however, one trial used hydroxyethyl starch, which may be harmful.13^, 14 This heterogeneity in populations and interventions could dilute any effects. Fourth, serious adverse events are often underreported. Accordingly, only six trials evaluated serious adverse events using varying definitions.

It is reassuring, however, that the two largest and highest quality trials, representing more than 85% of all subjects in high-quality trials, tested similar methods and populations and achieved significant separation in fluid volumes.8^, 10 In both trials, mortality, the primary outcome, was unchanged by fluid strategy.

Based on current data, it seems highly likely there is no meaningful difference in clinical outcomes between less versus more IV fluid strategies for people experiencing severe sepsis. We have therefore selected a color recommendation of red (no benefits) for the choice of fluid volumes in adult patients with sepsis. Further randomized controlled trials using well-defined fluid protocols in well-defined populations, as well as those comparing timing of vasopressors, will hopefully shed further light on this topic.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 26, 2024

Source

Johari F, Verma R. Paxlovid for nonhospitalized patients with COVID ‐19. Academic Emergency Medicine. Published online March 22, 2024:acem.14896.

Study Population: Two randomized controlled trials with 3286 nonhospitalized symptomatic adults with acute mild to moderate COVID-19

Efficacy Endpoints

All-cause mortality and hospitalization

Harm Endpoints

Adverse events (e.g., rebound, dysgeusia, diarrhea)

Narrative

Paxlovid is an oral medication that combines two drugs, nirmatrelvir and ritonavir. Nirmatrelvir is a protease inhibitor, and ritonavir is used to increase the levels of nirmatrelvir. In 2021 Paxlovid was granted Emergency Use Authorization by regulatory authorities in various countries including the United States for the treatment of mild-to-moderate COVID-19 in high-risk individuals. In 2023 the U.S. Food and Drug Administration (FDA) granted Paxlovid full approval for this purpose.1^, 2 Here we summarize data from randomized trials relevant to the FDA-approved use of the drug.

As of February 2024, we are aware of two relevant randomized controlled trials (RCTs), both sponsored and conducted by Pfizer.3^, 4 A third trial is under way by a nonindustry group in the United Kingdom.5 The first trial, called EPIC-HR,3 enrolled 2246 outpatients with <5 days of symptomatic COVID infection and at least one high-risk criterion for worsening. The most common criteria were obesity (80%), smoking, and hypertension. Any patients vaccinated against COVID or previously exposed to COVID were excluded from the trial. In this group, the drug reduced a composite endpoint of hospitalization for COVID or death: 6.3% with placebo versus 0.8% with Paxlovid (absolute risk difference [ARD] 5.5%, p < 0.001, number needed to treat [NNT] 18). This includes 12 deaths during the study period, all in the placebo group. Of note, a Cochrane systematic review in 2023 included the EPIC-HR trial and did not note or mention the results of the unpublished EPIC-SR trial.6

The risk of treatment-specific adverse events increases with Paxlovid (relative risk [RR] 2.1, 95% confidence interval [CI] 1.4–3.0; ARD 4%; number needed to harm [NNH] 25).6 The most common adverse events are dysgeusia and gastrointestinal symptoms such as diarrhea.

The second trial, EPIC-SR, remains unpublished but data have been uploaded to a trial registry site.4 The results report on 1288 symptomatic outpatients with acute COVID-19 given Paxlovid or placebo. No high-risk criteria were necessary for enrollment and COVID-vaccinated and previously exposed people were eligible. The trial found no difference between groups in hospitalization for COVID or death (0.8% vs. 1.6%, p = 0.2) and was stopped early for futility.

According to the only high-quality, prospective study reporting on virologic rebound that we are aware of, 127 nonhospitalized subjects with acute COVID were followed and tested. Virologic rebound (a substantial increase in viral load occurring after initial recovery) occurred in 21% of those who took Paxlovid versus 2% in the control group (ARD 19%, p = 0.001, NNH 5).7 Patients experienced a second clinical illness in 26% of those taking Paxlovid versus 15% in the control group.

Caveats

The data on Paxlovid for acute COVID-19 suffer from several limitations warranting caution. First, EPIC-HR, performed early in the pandemic, excluded people with a history of exposure to COVID or COVID vaccination.3 This makes the trial inapplicable to the world's current population. The second trial,4 with a more generalizable population, did not have this exclusion but the findings remain piecemeal and unpublished. Most participants in both trials, however, were younger than 65, of White ethnicity, and predominantly from high-income and upper-middle-income countries, an additional limitation for generalizing.

On the subject of efficacy Pfizer's first trial EPIC-HR, which found a benefit, it was ostensibly performed in high-risk subjects (hence “HR”), while the second trial EPIC-SR (standard risk) found no benefit. However, the actual risk levels in the two trials, as indicated by the rates of hospitalization or death in both control groups, were similar at 6% and 2%, respectively. The 6% in EPIC-HR would, moreover, be lower today since unvaccinated, unexposed people have significantly decreased and vaccination has proven to lower hospitalization or death.8 While factors like obesity and chronic illness theoretically differentiated high versus standard risk in the two trials, vaccination status may have been the main driver of the small difference. Therefore, EPIC-SR seems much more likely to be relevant to current practice. In addition, EPIC-SR's finding of no benefit is more consistent with Paxlovid's failure in RCTs for patients hospitalized with COVID,9 for long COVID,10 and as a preventive after exposure to COVID.11 Similar to EPIC-SR, Pfizer has not published the results of the latter two studies. Other than EPIC-HR, performed in unvaccinated, unexposed patients, we are not aware of any trial of Paxlovid finding a benefit in COVID. Perhaps the ongoing UK trial will shed further light.

Two additional concerns with Paxlovid include, firstly, the strong cytochrome P450 effects of ritonavir, the drug's boosting agent. Concomitant use with some medications might significantly interfere with drug levels and metabolism,12 which raises the specter of harm and has led to a daunting list of common medications that render patients ineligible.

Secondly, virologic rebound following Paxlovid is a concern. There have been multiple reports and several studies of the subject, and an MMWR review suggests most studies do not show differences between persons receiving Paxlovid and those not.13 However, only one study we are aware of included careful prospective data collection and sequential viral load measurement. Published after the MMWR report, the study found a much higher rate of virologic rebound with Paxlovid and a higher rate of clinical worsening, though there were limitations. The Paxlovid group included more elderly and immunologically compromised patients, and no matching or randomization was attempted. However, rebound was more common with Paxlovid regardless of immunocompromise and other characteristics.6 The clinical significance of these rebound changes remain fuzzy, but harms of a new drug are often missed or underestimated initially; therefore, we are approaching the data cautiously. Hopefully further prospective data will be forthcoming.

Despite these limitations and concerns, and despite the very low risk of hospitalization and death, the Infectious Diseases Society of America (IDSA) recommends treatment with Paxlovid in ambulatory patients with mild to moderate COVID-19 at high risk for progression to severe disease. Their guideline recommends starting Paxlovid as soon as possible, preferably within 5 days of onset of symptoms.14 This may partly be due to the IDSA being unaware of the three unpublished Pfizer trials including EPIC-SR, in which the drug failed in cohorts more directly relevant to current practice.

Based on the failure of Paxlovid in EPIC-SR, likely harms of the drug, and a consistent trend of not publishing trial results gives rise to concerns about potential deliberate lack of transparency, we have assigned an NNT color recommendation of yellow (unclear if benefits/more data needed) to Paxlovid for COVID. While the drug might have a role in treating high-risk and particularly unvaccinated individuals infected with COVID-19, until further data from independent trials become available, the current evidence does not support its widespread use.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Fatima Johari, MD; Rajesh Verma, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 23, 2024

Source

Davila E, McCormack J. sglt ‐2 inhibitors and glp ‐1 receptor agonists for type 2 diabetes. Academic Emergency Medicine. 2024;31(4):408-411.

Study Population: 421,346 patients with Type 2 diabetes, already on standard treatments, followed for 24 weeks or longer

Efficacy Endpoints

Death, nonfatal heart attack, nonfatal stroke, end-stage kidney disease, body weight change

Harm Endpoints

Severe hypoglycemia, severe gastrointestinal events, genital infection, amputation, ketoacidosis

Narrative

Type 2 diabetes is a condition that can affect many organs and can lead to serious complications. Recently, new classes of medications have been introduced for the treatment of Type 2 diabetes, including glucagon-like peptide 1 (GLP-1) receptor agonists and sodium-glucose transport protein 2 (SGLT-2) inhibitors. SGLT-2 medications increase the elimination of glucose and sodium in the urine by blocking the reuptake of filtered glucose in the kidney. GLP-1 receptor medications mimic the intestinal hormone incretin, increasing glucose-dependent endogenous insulin excretion. Both medications slow gastric emptying, decrease appetite, and regulate insulin and glucagon.1 Several trials have shown benefits prompting some guidelines to recommend these class of medications for patients with Type 2 diabetes (T2DM).1^, 2^, 3

The systematic review and network meta-analysis summarized here4 included 764 randomized trials testing SGLT-2 inhibitors or GLP-1 receptor agonists typically added to other antidiabetes medications. Trial groups received SGLT-2 or GLP-1 medications while control groups received placebos. However, both trial arms were on—and stayed on—standard background treatments that could include a variety of other medications (metformin, sulfonylureas, dipeptidyl peptidase-4 inhibitors, thiazolidinediones, alpha-glucosidase inhibitors, glitinides, or insulin).4 A total of 421,346 patients were involved in the 764 studies.

Outcomes of interest included death, nonfatal stroke, end-stage kidney disease, nonfatal heart attack, body weight change, severe hypoglycemia, severe gastrointestinal events, genital infection, ketoacidosis, amputation, and hyperkalemia. End-stage kidney disease is defined in studies as a glomerular filtration rate <15 mL/min (per 1.73 m²) or initiation of dialysis. Tables 1 and 2 show the results found in the review. Of note, the review reports medication effects according to a patient's baseline cardiovascular risk. This is because the magnitude of the effect of diabetes medications varies according to a person's chance of developing the problem the medication is aiming to prevent. For instance, among those already at little to no risk of having a stroke, a medication's ability to demonstrate a reduction in strokes is obviously small. For those who are at higher risk of future strokes, there is a greater possibility for improvement, and effective medications can have a greater impact. Effective medications therefore tend to have different impacts in people with different risks—the higher the risk, likely the greater the impact. The authors therefore analyzed and reported medication effects separately for people in each of the following risk categories: very low (fewer than three cardiovascular risk factors), low (three or more risk factors), moderate (patients who already had known cardiovascular disease), high (those with chronic kidney disease), and very high (known cardiovascular and kidney disease).4

Major harms included severe hypoglycemic episodes, severe gastrointestinal adverse events, genital infections, amputation, and ketoacidosis. SGLT-2 inhibitors increased genital infections (odds ratio [OR] 3.5, 95% confidence interval [CI] 3.0–4.0, absolute risk difference [ARD] 14%; NNH 7; high certainty). GLP-1 receptor agonists increased the risk of severe gastrointestinal symptoms (OR 2.5, 95% CI 1.2–5.0; ARD 6%; NNH 17; low certainty). There were no other differences in risks of harm between SGLT-2 inhibitors and GLP-1 receptor agonists.

Another recent review by Shi et al.5 conducted a similar analysis; however, it included trials comparing SGLT-2 and GLP-1 medications to standard treatments without a placebo control. After reviewing a number of original trials, we made the judgment that comparing these medications to placebo paints a more accurate picture of their effects, and we therefore used the Palmer meta-analysis as our primary source (though most relative risks in the two reviews are quite similar).4

The meta-analysis by Shi et al. also, however, included trials of the new medication tirzepatide which is a glucose-dependent insulinotropic polypeptide (GIP) plus a GLP-1 medication. For completeness we are including here the weight loss impact of tirzepatide (average 8.6 kg, 95% CI 7.8–9.4 kg) and its attendant increase in severe gastrointestinal symptoms (OR 4.6, 95% CI 1.9–11.1; ARD 12.5%; NNH 8; moderate certainty), estimates from the review by Shi et al.5 There was no statistical difference between tirzepatide and standard treatment for all other outcomes of interest.

Importantly, when any of these medication classes were compared directly to metformin, no additional benefit was seen other than greater weight loss of varying degrees (0.2–7.7 kg).4^, 5^, 6 However, patients on metformin had fewer genital infections and gastrointestinal adverse events.4^, 5^, 6 While head-to-head trials are comparatively few, a large population-based study of the new medications from 2022 appears to confirm their lack of benefit over metformin, showing almost 9000 on SGLT-2 medications, when matched to over 17,000 on metformin, had identical rates of heart attack, stroke, and death but higher rates of genital infection.7

Caveats

Notable limitations exist in these data. Background treatments varied from trial to trial, making precise comparisons impossible. A total of 687 of the included 764 trials were rated as high risk of bias in at least one of six domains.4 Many endpoints were surrogate outcomes or laboratory values; however, the huge number of studies meant there were enough reporting patient-oriented outcomes to overcome this. There are a paucity of studies comparing SGLT-2 to either dual GIP/GLP-1 or GLP-1 medications directly, prohibiting conclusions about one class over another. There were also significant differences in trial follow-up periods, which could result in over- or underestimating effects.

In summary, while SGLT-2 and GLP-1 medications appear to reduce mortality, heart attack, and end-stage kidney disease compared to many other classes, available evidence does not show them to be better than metformin, though some (particularly tirzapetide) may lead to greater weight loss. The newer medications also cause genital infections, particularly SGLT-2s, at a NNH of 7. The cost and harms of SGLT-2 and GLP-1 medications therefore should be balanced against weight loss and other effects compared to metformin.

We have assigned an NNT color recommendation of green (benefits > risk) for SGLT-2 inhibitors and GLP-1 receptor agonists in patients with Type 2 diabetes for improving most patient-centered outcomes over most medications (as well as in addition to a patient's current diabetes medication regimen). However, there are important increases in harms, particularly genital infections with SGLT-2 inhibitors and severe gastrointestinal adverse events with GLP-1 medications. Therefore, if exclusively comparing the medications to metformin we would assign a color of yellow, as more head-to-head trials are needed to determine if the new medications improve outcomes greater than other agents (for instance metformin).

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Esteban Davila, MD; James McCormack, BSc (Pharm), PharmD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Long B, Gottlieb M. Therapeutic hypothermia following cardiac arrest. Academic Emergency Medicine. Published online July 31, 2023:acem.14785.

Study Population: 2,900 participants in 6 trials, most enrolling comatose survivors of out-of-hospital cardiac arrest

Efficacy Endpoints

Survival and favorable neurologic outcome

Harm Endpoints

Not assessed

Narrative

Cardiac arrest and its aftermath represent common conditions managed in the emergency department. Timely cardiopulmonary resuscitation, rapid defibrillation of shockable rhythms, and post-arrest care are integral to improving outcomes.1 However survival remains poor, ranging between 6% and 20%.1^, 2^, 3 Several randomized trials in the past two decades have evaluated therapeutic hypothermia, also known as targeted temperature management (TTM), during the post-arrest phase to improve survival and neurologic outcomes.4^, 5^, 6^, 7^, 8 TTM has been refined over time with regard to patient selection, time of initiation, temperature goals, means and duration of cooling, and research design.9^, 10^, 11^, 12^, 13 Perhaps unsurprisingly therefore, trials during this period have yielded varying results. Herein, we summarize a systematic review of randomized trial data addressing TTM for survivors of cardiac arrest.11

The systematic review discussed here included 6 randomized trials (n = 2,900 participants) of comatose post-cardiac arrest patients. Trials randomized patients to receive either hypothermic TTM (32-34℃) versus normal-temperature TTM (36.5–38℃) in the hours after their cardiac arrest, or to receive hypothermic TTM versus no TTM.11 Included studies provided information concerning the timing of TTM initiation, target temperature, duration, method, and rewarming rate for cooled patients. The review did not include studies that used TTM for indications other than post-cardiac arrest, or with patients <18 years of age. The systematic review reported outcomes recommended by the International Liaison Committee on Resuscitation Advanced Life Support Task Force. These outcomes included short-term survival, mid-term favorable neurologic outcome, long-term survival, and long-term favorable neurologic outcome. The systematic review defined a favorable neurologic outcome as a modified Rankin Scale (mRS) score of 0-3 or a Cerebral Performance Category score of 1 to 2.11 Other outcomes included health-related quality of life, cognitive function, and anxiety and depression. Authors also reported subgroup analyses on shockable (ventricular fibrillation and pulseless ventricular tachycardia) or non-shockable (asystole and pulseless electrical activity) initial cardiac arrest rhythms. Finally, they conducted sensitivity analyses with exclusion of studies at high risk of bias and trials more than 10 years old. They also conducted sensitivity analysis comparing 36℃ versus 32-34℃.

Of note, the two earliest trials (Bernard 2002 and HACA 2002) enrolled almost exclusively patients with shockable rhythms4^, 5 while a multicenter trial (HYPERION 2019)6 enrolled only patients with asystole or pulseless electrical activity. Other trials enrolled mixed populations with mostly shockable rhythms (72-100%).7^, 9^, 10 Only two trials included patients with in-hospital cardiac arrest;4^, 6 the remaining trials studied patients after cardiac arrest outside of the hospital.5^, 7^, 9^, 10

The findings of the meta-analysis indicates that TTM with a hypothermic target of 32-34℃ does not improve survival (5 trials; 2,776 participants) or favorable neurologic outcome at 90 to 180 days (5 trials; 2,753 participants). The results remained the same after authors performed sensitivity analyses. Evidence certainty was low for all outcomes.

Caveats

There are several important considerations when interpreting these results. First, the certainty of evidence was low. This was driven by the risk of bias from lack of blinding, inconsistency of results and heterogeneity of the trials, and imprecision of effects leading to wide confidence intervals. Second, not all trials reported all outcomes sought in the review. Third, studies were performed decades apart as TTM changed, which may particularly have impacted the control groups. The HYPERION trial used a 36.5-37.5℃ target in the control group while Dankiewicz targeted 37.5℃, intervening only if the patient’s temp was greater than 37.8℃.6^, 7 Other trials used no TTM in the control group or 37℃. Fourth, trials reported variable follow up durations. A fifth limitation was variations in time to achieving target temperature and durations of therapy. Sixth, authors were unable to conduct analyses separately for the participants whose arrests occurred in-hospital. Finally, this meta-analysis did not report adverse events which include arrhythmia, infection, and electrolyte abnormalities.12^, 13

It is important to note the first two high profile trials of hypothermic TTM from 2002 found a benefit, but these trials were unblinded, single-center, proof-of-concept studies that enrolled less than 10% of screened patients, did not standardize approaches to withdrawal of care or control arm treatment, and were stopped early for logistical reasons.4^, 5 Together they enrolled 213 subjects. By contrast, one TTM trial published in 2021 enrolled 1,821 participants and found no benefit with a target of 33℃ compared to normal temperature.7 The study was more rigorous, better controlled, and multi-center across 61 sites.

Also notable is a 2023 Cochrane systematic review of 12 RCTs and quasi-RCTs.12 The review found hypothermic TTM improved neurologic outcome (RR: 1.6, 95% CI 1.2 to 2.2) but also caused higher rates of adverse events.12 These results, however, incorporate several smaller, less controlled studies with significant bias and heterogeneity, and it is these studies that appear to drive the improvement. A second meta-analysis published in 2023 included 5 trials (n=3,614 patients) and found no improvement with TTM in survival or rate of favorable neurologic outcome, with a higher risk of arrhythmia.13

History suggests large replication studies often produce more reliable results, particularly when small early studies report benefits of a novel therapy.14 Based on current data it appears that induction of therapeutic hypothermia does not improve neurologic outcome or survival while multiple reviews suggest an increased risk of adverse events.12^, 13 Therefore, we have selected a color recommendation of Red (benefits and harms equal or equivocal) for the use of therapeutic hypothermia in survivors of cardiac arrest. This recommendation is different from the previous NNT summary which had a green recommendation.15 The older summary was based on a 2009 Cochrane systematic review.16 Since then, several large RCTs have been published. The new data have more clearly challenged any claims of benefit due to hypothermia after cardiac arrest.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

August 17, 2023

Source

Davila E, Chirayil J, Silverberg M. Early versus delayed coronary angiography after out‐of‐hospital cardiac arrest without ST‐segment elevation. Academic Emergency Medicine. Published online July 24, 2023:acem.14774.

Study Population: 1590 patients with OCHA randomized to early coronary angiography or delayed angiography enrolled in six randomized controlled trials

Efficacy Endpoints

Mortality, good neurological function

Harm Endpoints

Adverse events such as ventricular arrhythmias, major bleeding, and acute kidney injury

Narrative

Less than 40% of patients with out-of-hospital cardiac arrest (OHCA) with return of spontaneous circulation (ROSC) survive to discharge.1 A subset of these patients will have ST-elevation on their post-ROSC electrocardiogram. Multiple guidelines recommend early angiography after OHCA for these patients after ROSC citing associations with improved survival.2^, 3^, 4 The benefits of early angiography after OHCA in patients without ST-elevation are less clear. Therefore, it is important to examine the available evidence for benefits and harms of early angiography after OHCA for patients without ST-elevation.

The systematic review and meta-analysis discussed here examined the safety and efficacy of early versus delayed angiography in patients without ST-elevation after ROSC.5 The systematic review included six multicenter randomized clinical trials (RCTs) with a total of 1590 patients. All trials enrolled patients with OHCA of presumed cardiac origin and excluded patients with obvious or suspected noncardiac etiology. Of the patients randomized to early angiography, 95.6% received their angiography immediately after randomization,6^, 7 within 1 h of presentation,8 or within 2 h of presentation.9^, 10^, 11 In the delayed angiography group, 59.7% received an angiogram, ranging from within 6 h to 4 days after randomization, contingent on either evidence of neurological recovery or at the discretion of the treating physician. The outcomes of interest included mortality at longest follow-up; disability (measured by the Cerebral Performance Category or another validated scale); duration of mechanical ventilation and intensive care unit (ICU) and hospital length of stay (LOS); and adverse events including ventricular arrhythmias, major bleeding, acute kidney injury, and need for renal replacement therapy.

The pooled data for patients with OHCA without ST-segment elevation demonstrated that early angiography had no statistically significant impact on mortality compared to delayed angiography (six trials, 1590 patients, moderate quality of evidence). The difference in the rates of good neurologic outcome, ICU and hospital LOS, duration of mechanical ventilation, major bleeding, acute kidney injury, and need for dialysis were also not statistically significant between early versus delayed angiography.

Caveats

This systematic review and meta-analysis summarized here has notable limitations. The included trials were unblinded, possibly introducing bias in decisions for further treatment after randomization. Despite inclusion of patients with presumed cardiac causes of cardiac arrest, there was low incidence of acute coronary occlusion in patients without ST-elevation after OHCA, ranging from 15% to 40%. In addition, the majority of nonsurvivors died of neurologic complications after the cardiac arrest, potentially biasing against angiography. Other clinically relevant factors for possible cardiac-related causes such as suggestive patient history, dynamic electrocardiographic changes, echocardiography demonstrating regional wall motion abnormalities, or serial troponin values were not considered in the trials. It is likely that certain subgroups of patients could have benefitted from the intervention more than others. However, due to insufficient data, the authors of the systematic review were not able to perform many of the preplanned subgroup analyses. Furthermore, only comatose patients were included in the RCTs, and therefore these recommendations do not apply to noncomatose patients.

Five of the RCTs reported low risk of bias. However, one of the RCTs reported a high risk of bias from deviations from intended protocol. Though the systematic review rated the outcome of mortality as moderate certainty, all other outcome measures were rated as low to very low certainty. The clinical heterogeneity among the trials, particularly in regard to timing of early and delayed angiography as well as the timing of mortality outcome measures, threatens the validity of the results of the meta-analysis. Additionally, the majority of the included studies were underpowered for their primary outcomes. Lastly, the confidence intervals for all safety outcomes were wide and could not confidently rule out the potential for harm.

In light of available evidence, early angiography following OHCA without ST-elevation after ROSC does not appear to have an effect on mortality, neurological disability, or adverse events. Larger trials with less clinical heterogeneity are necessary to improve the certainty of the conclusions and to identify possible subgroups that may benefit from early angiography. Though of moderate to very low certainty, the findings are consistent with prior meta-analysis of RCTs exploring this research question.12 Therefore, we have assigned an NNT color recommendation of red (no benefits) for early angiography for OHCA without ST-elevation post-ROSC.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Esteban Davila, MD; Joseph Chirayil, MD; Mark Silverberg, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

August 15, 2023

Source

Long B, Gottlieb M. Direct oral anticoagulants versus conventional anticoagulants for pulmonary embolism. Academic Emergency Medicine. Published online July 5, 2023:acem.14771.

Study Population: 10 trials comprising 13,073 participants

Efficacy Endpoints

Rate of recurrent VTE, DVT, PE, all-cause mortality, and major bleeding

Harm Endpoints

Adverse treatment reactions

Narrative

Pulmonary embolism (PE) has an incidence ranging from four to 12 per 10,000 annually and is a leading cause of cardiovascular mortality in the United States, contributing to nearly 300,000 deaths per year.1^, 2^, 3^, 4 Conventional treatment includes heparins, fondaparinux, or vitamin K antagonists (VKAs), but recent data suggest treatment with oral direct thrombin inhibitors (DTIs) or oral factor Xa inhibitors, collectively known as direct oral anticoagulants (DOACs), is effective and safe.5^, 6^, 7^, 8^, 9 DOACs can be administered orally and have a more predictable effect without the need for monitoring. Here we summarize a recent update to a prior Cochrane review,10 adding 1484 new participants to meta-analyzed data on the subject.11

The Cochrane review included 10 randomized trials comprising 13,073 participants with confirmed PE allocated to treatment with a DOAC or conventional anticoagulation.11 The authors excluded trials not administering oral DOACs, trials not using conventional treatment with heparin and/or VKAs as the comparison, and trials with treatment less than 3 months. Confirmation of PE included computed tomography pulmonary angiography, ventilation/perfusion scan, or pulmonary angiography. DTIs included dabigatran and ximelagatran; factor Xa inhibitors included rivaroxaban, apixaban, betrixaban, and edoxaban; and control agents included low-molecular-weight heparin (LMWH), unfractionated heparin (UFH), or VKAs. Two trials evaluated dabigatran (2553 patients) and eight trials evaluated factor Xa inhibitors (three with apixaban [3307 patients], three with rivaroxaban [2548 patients], and two with edoxaban [2172 patients]).

The primary outcome of the review included recurrent venous thromboembolism (VTE), defined variably as clinically overt new deep vein thrombosis (DVT), PE, or either, confirmed by standard imaging during the period after PE diagnosis. Secondary outcomes included all-cause mortality, major bleeding (fatal bleeding, symptomatic bleeding in a critical area, bleeding causing a fall of hemoglobin ≥2 g/dL or requiring transfusion of ≥2 units, or any combination of these), and health-related quality of life.

Compared to conventional anticoagulation, DTIs led to a similar rate of recurrent PE (two trials, 1602 participants; moderate-certainty evidence), recurrent VTE (two trials, 1602 participants; moderate-certainty evidence), recurrent DVT (two trials, 1602 participants; moderate-certainty evidence), and major bleeding (two trials, 1527 participants; moderate-certainty evidence). Factor Xa inhibitors were also similar to conventional anticoagulation for recurrent PE (three trials, 8186 participants; moderate-certainty evidence), recurrent VTE (eight trials, 11,416 participants; moderate-certainty evidence), recurrent DVT (two trials, 8151 participants; moderate certainty evidence), all-cause mortality (one trial, 4817 participants; moderate-certainty evidence), and major bleeding (eight trials, 11,447 participants; low-certainty evidence). Authors did not evaluate quality of life.

Caveats

There are several important limitations to this systematic review. First, overall outcome event rates were low resulting in wide confidence intervals. While the included trials were mostly of good methodological quality, the certainty of evidence was considered moderate or low. The authors downgraded certainty for both imprecision and heterogeneity in the data. Second, the duration of treatment varied among the studies, ranging from 3 months to 1 year. Third, health-related quality of life was not reported. Fourth, while overall risk of bias was low in eight trials, two trials were at high risk of reporting bias, and six were open label.

Current guidelines incorporate treatment of VTE with DOACs due to efficacy, ease of use, and no need for monitoring. The American College of Chest Physicians guidelines recommend DOACs over VKAs for treatment of VTE, and these guidelines also recommend oral factor Xa inhibitors over LMWH for VTE associated with cancer.12^, 13 The 2019 European Society of Cardiology also recommends DOACs over VKAs in those requiring treatment for VTE,2 while the NICE 2020 guidelines recommend apixaban or rivaroxaban as the initial anticoagulant of choice.4

Based on these data, there seems to be little or no outcome difference between DOACs and conventional anticoagulation for PE; however, DOACs provide several practical advantages including ease of use, fixed dosing, ability to initiate in the ED without the need for LMWH injections or UFH infusions, and no need for routine laboratory monitoring. Thus, we have provided a color recommendation of green (benefits > harms) for the use of DOACs in treatment of PE. It is difficult to directly compare prices due to variations in dosages, treatment regimens, and local cost data for DOACs versus warfarin14^, 15^, 16; however, in general, DOACs are significantly more expensive than conventional anticoagulation. For example, a 1-month supply of dabigatran costs more than $400, and a 1-month supply of rivaroxaban costs more than $500, though this depends on each individual patient's insurance.17^, 18 In contrast, a 1-month supply of 5-mg tablets of warfarin (a traditional VKA) is about $11, though warfarin does require laboratory monitoring, which increases the price of use.19 Also of note, DOAC reversal may be more expensive (idarucizumab for dabigatran and prothrombin complex concentrate [PCC] or andexanet alpha for factor Xa inhibitors) compared to warfarin (vitamin K with fresh-frozen plasma or PCC).20 More studies are necessary to compare DOACs to each other, as well as to evaluate adherence, mortality, safety, quality of life, cost-effectiveness, and tolerability.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

August 10, 2023

Source

Long B, Gottlieb M. Direct oral anticoagulants versus conventional anticoagulants for deep vein thrombosis. Academic Emergency Medicine. Published online June 23, 2023:acem.14763.

Study Population: 21 trials comprising 30,895 participants with confirmed DVT

Efficacy Endpoints

Recurrent VTE, all-cause mortality, and major bleeding

Harm Endpoints

N/A

Narrative

Deep vein thrombosis (DVT) has an incidence of approximately five per 10,000 annually.1^, 2^, 3 If DVT is not appropriately managed, the thrombus may progress or dislodge leading to pulmonary embolism (PE). DVT was traditionally treated with heparin and vitamin K antagonists (VKAs).2^, 3^, 4 However, oral direct thrombin inhibitors (DTIs) and oral factor Xa inhibitors, broadly known as direct oral anticoagulants (DOACs), demonstrate several characteristics that may be favorable to heparin and VKAs. DOACs can be administered orally, have a predictable effect without the need for frequent monitoring or dose adjustment, and have few medication interactions.5^, 6 Current guidelines have incorporated DOACs for management of DVT based on recent literature, suggesting that DOACs are safe and efficacious.7^, 8^, 9^, 10^, 11^, 12^, 13^, 14 A prior Cochrane systematic review evaluated 11 randomized controlled trials (RCTs) comprising 27,945 participants15; however, the publication of several new RCTs required an updated evaluation of DOACs in the treatment of DVT.10^, 11^, 12^, 13^, 14

This updated Cochrane review published in 2023 included RCTs of participants with confirmed DVT allocated to treatment with a DOAC or conventional anticoagulation.2 They did not include studies evaluating these treatments by other routes, trials not using conventional treatment with heparin and/or VKAs as the comparison, and trials with treatment less than 3 months. Confirmation of DVT was based on standard imaging (venography, impedance plethysmography, whole-leg compression ultrasound, proximal compression ultrasound). DTIs included dabigatran and ximelagatran; factor Xa inhibitors included rivaroxaban, apixaban, and edoxaban; and comparisons included low-molecular-weight heparin (LMWH), unfractionated heparin (UFH), or VKAs (warfarin, acenocoumarol, phenprocoumon, and fluindione).

The Cochrane review utilized three primary outcomes. The first included recurrent venous thromboembolism (VTE), which was defined as a clinically overt DVT confirmed by standard imaging such as proximal leg vein ultrasound scan and/or D-dimer test or a clinically overt PE confirmed by imaging including computed tomography pulmonary angiography (CTPA) and/or ventilation/perfusion scan. The second was a recurrent DVT confirmed by standard imaging including proximal leg ultrasound or D-dimer. The third primary outcome included PE (fatal and nonfatal) confirmed by imaging. Secondary outcomes included all-cause mortality, major bleeding (fatal bleeding, symptomatic bleeding in a critical area/organ, bleeding causing a fall of hemoglobin ≥2 g/dL or requiring transfusion of ≥2 units, or any combination), post-thrombotic syndrome (PTS), and health-related quality of life (as reported by the individual studies).

The systematic review included 21 trials (n = 30,895 participants).2 Three studies investigated DTIs (two dabigatran and one ximelagatran), 17 studies investigated factor Xa inhibitors (eight rivaroxaban, five apixaban, and four edoxaban), and one trial with three arms investigated both a DTI (dabigatran) and a factor Xa inhibitor (rivaroxaban).

DTIs compared to conventional anticoagulation did not reduce the rate of recurrent VTE (odds ratio [OR] 1.17, 95% confidence interval [CI] 0.83–1.65; three studies, 5994 participants; moderate-certainty evidence), recurrent DVT (OR 1.11, 95% CI 0.74–1.66; three trials, 5994 participants; moderate-certainty evidence), fatal PE (OR 1.32, 95% CI 0.29–6.02; three trials, 5994 participants; moderate-certainty evidence), nonfatal PE (OR 1.29, 95% CI 0.64–2.59; three trials, 5994 participants; moderate-certainty evidence), or all-cause mortality (OR 0.66, 95% CI 0.41–1.08; one trial, 2489 participants; moderate-certainty evidence).16^, 17^, 18 However, DTIs reduced the risk of major bleeding (OR 0.58, 95% CI 0.38–0.89; absolute risk difference [ARD] 0.80%; number needed to treat [NNT] 125; three studies, 5994 participants; high-certainty evidence) compared to conventional anticoagulation.

Factor Xa inhibitors compared to conventional anticoagulation did not reduce rate of recurrent VTE (OR 0.85, 95% CI 0.71–1.01; 13 trials, 17,505 participants; moderate-certainty evidence), recurrent DVT (OR 0.70, 95% CI 0.49–1.01; nine trials, 16,439 participants; moderate-certainty evidence), fatal PE (OR 1.18, 95% CI 0.69–2.02; 6 trials, 15,082 participants; moderate-certainty evidence), nonfatal PE (OR 0.93, 95% CI 0.68–1.27; seven trials, 15,166 participants; moderate-certainty evidence), or all-cause mortality (OR 0.87, 95% CI 0.67 to 1.14; nine trials, 10,770 participants; moderate-certainty evidence). However, factor Xa inhibitors reduced the risk of major bleeding (OR 0.63, 95% CI 0.45–0.89; ARD 0.86%; NNT 116; 17 studies, 18,066 participants; high-certainty evidence). Only one trial evaluated PTS, and one study evaluated health-related quality of life, preventing the systematic review from drawing clear conclusions regarding these outcomes.

Caveats

There are several limitations to the findings of this systematic review.2 The overall outcome event rates were low, resulting in wide CIs and limited precision, though heterogeneity was low for all reported outcomes. The Cochrane review only included patients with DVT, so it is difficult to draw conclusions regarding treatment failures or mortality in patients with PE. Importantly, the Cochrane review included D-dimer for diagnosis of DVT or PE as part of the primary outcome, which may have led to incorrect diagnoses in some cases, as D-dimer test should only be used to exclude the diagnosis of VTE, and imaging is necessary for VTE diagnosis. The duration of treatment varied among the studies, ranging from 3 months to 1 year. Almost all participants had DVT of the lower extremities, and thus authors did not address DVT of the upper extremities. Of note, ximelagatran was withdrawn from the market in 2006 due to safety issues, though only one study evaluated this medication.18 While it is difficult to compare prices due to various dosages and treatment regimens, in general DOACs are significantly more expensive than conventional anticoagulation. For example, a 1-month supply of dabigatran costs more than $400, and a 1-month supply of rivaroxaban costs more than $500.19^, 20 In contrast, a 1-month supply of 5-mg tablet of warfarin (a traditional VKA) is approximately $11.21 However, warfarin requires regular monitoring, which is inconvenient for patients and associated with increased cost, and the medication has numerous dietary restrictions and medication interactions.

Based on the existing data, DOACs have similar efficacy but a better safety profile (lower risk of major bleeding) than conventional anticoagulation for treatment of DVT. Additionally, DOACs provide a practical benefit with ease of use, fewer medication interactions, and no requirement for frequent laboratory monitoring. Thus, we have assigned a color recommendation of green (benefits > harms) for the use of DOACs in treatment of DVT. Of note, DOACs are renally excreted and may require dose adjustment in those with chronic kidney disease (CKD). Further studies should include subjects with CKD and those with DVT in unusual sites (e.g., upper extremities) as well as those with specific health conditions such as subjects who have obesity and malignancy.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

July 18, 2023

Source

Long B, Gottlieb M. Aggressive intravenous fluid resuscitation for acute pancreatitis. Academic Emergency Medicine. Published online April 20, 2023:acem.14741.

Study Population: 953 patients with acute pancreatitis enrolled in 9 randomized controlled trials

Efficacy Endpoints

Survival rate

Harm Endpoints

Primary: mortality
Secondary: clinical improvement, fluid-related-complications, sepsis, acute respiratory failure, and total hospital days

Narrative

Acute pancreatitis (AP) has an incidence of 34 cases per 100,000 person-years and mortality of 1.6 deaths per 100,000 person-years.1^, 2^, 3^, 4 International guidelines recommend early fluid resuscitation with isotonic crystalloids to treat hypovolemia and prevent organ hypoperfusion.5^, 6^, 7^, 8^, 9 Several meta-analyses, however, have demonstrated inconsistent findings with aggressive intravenous (IV) fluid therapy in AP.10^, 11^, 12 Moreover, the recent WATERFALL trial reported a three-fold increased risk of fluid overload in those who received aggressive IV fluid hydration.13 Therefore it is important to assess the evidence for aggressive IV fluid resuscitation in patients with AP.

The systematic review summarized here included randomized controlled trials (RCTs) of adult patients diagnosed with AP based on the revised Atlanta classification.14 This classification requires at least two of the following to be present for diagnosis: abdominal pain consistent with AP, serum lipase or amylase at least three times greater than the upper limit of normal, or classic imaging findings of AP.14^, 15^, 16 The systematic review classified the severity of AP based on the Atlanta international symposium and revised Atlanta classification.15^, 16 Patients with mild(absence of organ failure and local or systemic complications) and moderately severe (transient organ failure or local or systemic complications) AP were classified into the non-severe AP group. Aggressive fluid resuscitation was defined as fluid resuscitation greater than 10 mL/kg/hour for initial management, a fluid bolus of 10 mL/kg/hour for 2 hours followed by 2-3 mL/kg/hour in the first 24 hours, or isotonic fluids >500 mL/hour for the first 12-24 hours. The comparison group received non-aggressive resuscitation, defined as fluid administration less than the thresholds defined above. The primary outcome was all-cause mortality. Secondary outcomes included rate of clinical improvement, fluid-related-complications (abdominal compartment syndrome, pulmonary or peripheral edema, and any sign of volume overload), sepsis, acute respiratory failure, and total hospital days.

The systematic review included 9 RCTs (n=953 patients).14 Two trials evaluated patients with severe AP, while 6 trials evaluated those with non-severe AP. Five studies were conducted in China, while the remainder included the United States (one study), Mexico (one study), and Thailand (one study). One study involved multiple countries (India, Italy, Mexico, and Spain). The majority of studies compared a 20 ml/kg bolus then 3 mL/kg/hour infusion (aggressive) with 10 ml/kg bolus then 1.5 mL/kg/hour infusion (non-aggressive).

Aggressive IV hydration was associated with increased risk of mortality compared to non-aggressive fluid hydration when data for patients with severe and non-severe AP were pooled together (9 RCTs, risk ratio [RR]: 2.4, 95% confidence interval [CI]: 1.4 to 4.2, absolute risk difference [ARD]: 4.7%, number needed to harm [NNH]: 21). Aggressive fluid hydration was associated with increased risk of mortality in those with severe AP (2 RCTs, pooled RR: 2.5, 95% CI: 1.4 to 4.4, ARD: 19.5%, NNH: 5) but not in patients with non-severe AP (3 RCTs). Regarding secondary outcomes, the authors found increased risk of fluid-related complications (5 RCTs, pooled RR: 2.5, 95% CI: 1.7 to 3.8, ARD: 12.5%, NNH: 7) when patients with severe and non-severe AP were pooled.

Caveats

This systematic review has several limitations. While it included only RCTs there were small numbers of patients in each trial, reducing statistical power to detect potential differences. This was particularly notable for severe AP, which comprised only 211 total patients. Second, while the review did include trials from multiple countries, 5 trials were conducted in China, limiting the applicability to other countries or settings. Third, the included RCTs did not report the total hydration volumes in hospitalized patients, so the true volumes patients in the aggressive hydration and non-aggressive hydration groups received are unclear. Fourth, there were a variety of etiologies for AP in the included patients, adding clinical heterogeneity. Fifth, the certainty of evidence for all study outcomes was determined to be low to very low, primarily due to methodological issues and small sample sizes of the included RCTs.

Based on the available evidence, aggressive fluid hydration seems to increase the risk of mortality and complications in patients with AP. Further data, more carefully documented, and from larger studies and across multiple countries are necessary to improve the certainty and reliability of these findings. Despite the limitations and low certainty, however, this evidence base appears far stronger than the largely observational evidence underlying longstanding recommendations for aggressive hydration in AP. Thus, we have assigned an NNT color recommendation of Black (Harms > Benefits) for aggressive IV hydration in those with AP and look forward to future trials that we hope will bring further clarity.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

May 9, 2023

Source

Tchouapi P, Anderson KE, Hein PN. “Intravenous magnesium as an adjunct to standard of care for treatment of atrial fibrillation with rapid ventricular response.” Academic Emergency Medicine. Published online April 6, 2023:acem.14734.

Study Population: 745 patients with atrial fibrillation and rapid ventricular response enrolled in 6 randomized controlled trials

Efficacy Endpoints

Rate control, conversion to normal sinus rhythm

Harm Endpoints

Adverse events (bradycardia, hypotension, flushing)

Narrative

Atrial fibrillation with rapid ventricular response (RVR) is the most encountered arrhythmia in the emergency department. Its natural course posits significant morbidity and mortality risk from complications such as stroke, thromboembolism, heart failure, and decreased quality of life. The current management of atrial fibrillation involves rate control or rhythm control. Intravenous (IV) magnesium use in the emergency department in conjunction with standard of care is proposed as a means for achieving rate control and potentiation of rhythm conversion in atrial fibrillation with RVR.

The systematic review and meta-analysis by Ramesh et al. discussed here, examined the effectiveness of IV magnesium as an adjunct to standard of care for rate control and rhythm control on atrial fibrillation in non-post operative patients.1 The systematic review includes 6 randomized controlled trials (n = 745 patients) from 1993 to 2018, with 448 in the IV magnesium group and 297 in the control group. Five trials used 4.5 to 10 grams of IV magnesium and one study used 3 grams. The “standard of care” that magnesium sulfate was adjunct to, was most commonly digoxin, although beta blockers and calcium channel blockers were used as well. The primary outcomes included rate control (achieving a heart rate <100 or <90 or a drop in heart rate more than 20%). And conversion to normal sinus rhythm (being in sinus rhythm at the endpoints of the study, ranging from 2.5 to 24 hours from initiation).1

The meta-analysis demonstrated that the use of IV magnesium for the treatment of atrial fibrillation with RVR in addition to routine care was superior to placebo in achieving rate control (Odds Ratio [OR]: 2.49, 95% CI, 1.80 to 3.45; Absolute Risk Difference [ARD]: 23%; Number-needed-to-treat [NNT]: 4) and conversion to sinus rhythm (OR: 1.75, 95% CI, 1.08 to 2.84: ARD: 7%, NNT= 14).1 The risk of flushing was significantly higher in the IV magnesium group compared to the placebo group (9% vs. 0.4%; OR: 19.79, 95% CI, 4.30 to 91.21; ARD: 8.6%; NNH: 11).1 The risk of bradycardia and hypotension was not significantly differenct between the groups. The heterogeneity among trials was relatively low for all primary outcomes analyses.1

Two other systematic reviews by Ho et. al (2006) and Onalan et al. (2007) assessed the effectiveness of IV magnesium in the as primary therapy and as an adjunct.2^, 3 Ho et al. included 10 randomized controlled trials (n = 515 patients) in their systematic review and found a reduction in ventricular response (defined as HR <100) in patients receiving IV magnesium as an adjunct to digoxin.2 When compared directly to calcium channel blockers or amiodarone, IV magnesium was less effective in reducing ventricular response. Of note, most of the included trials enrolled patients with rapid atrial fibrillation following cardiac surgery.

The systematic review by Onlan et. al included 8 trials (n = 476 patients) with only 4 of the included trials (n = 303) reported IV magnesium superiority in achieving rate control when directly compared to placebo, calcium channel blockers or beta blockers.3 In all eight trials, IV magnesium group had a higher chance to regain normal sinus rhythm when compared to controls (placebo, calcium channel blockers, and ajmaline). This systemic review reported significant heterogeneity in comorbidities, baseline pharmacological management, and baseline magnesium levels of patients included in the review.

Though both systematic reviews ultimately concluded that IV magnesium is safe and effective in rate control in atrial fibrillation with RVR, both as a single agent or as an adjunct, it is important to note that these meta-analyses included older studies, mostly published more than 15 years ago, less likely to reflect the current care. The systematic review by Ramesh et al. discussed here, included two more recent trials that were not included in the other two systematic reviews.

We did not discuss the most recent (2022) systematic review by Hoffer et al, because it additionally included two studies using only IV magnesium versus placebo. We consider this a separate question to using magnesium as an adjunctive treatment, and wished to focus on the more narrow question of magnesium as adjunct to first line agents.4 We used the systematic review and meta-analysis by Ramesh et al. published in 2021 as the primary source for this evidence-based summary given its recent publication, more stringent requirement for definition of heart rate control, and its superior methodology.

Caveats

The criteria for rate and rhythm control in the included studies varied both for heart rate goal and time to outcome measure. Treatment groups varied in magnesium dosage (ranging from 3-10 g) and infusion rates. Notably, while these studies discussed care in the RVR time period, they largely did not discuss care after achieving RVR for maintenance therapy (such as a proposed magnesium maintenance rate). Cardiac comorbidities, antiarrhythmic therapies, and additional medications were not consistently reported in the included studies. More than 90% of patients included in the meta-analysis were enrolled in two of the included trials and their findings significantly skewed the results of the analysis.

The routine care for treatment of atrial fibrillation was not standardized among trials. As many of the trials were older, digoxin was used as a common treatment for atrial fibrillation in many of the trials. Therefore, the standard of care used in the original trials may not reflect modern practices of first-line agents, particularly beta blockers or calcium channel blockers.

Experts in the field have raised concern about patient selection in studies evaluating magnesium as a rate and rhythm control agent.5 The clinicians are reminded that atrial fibrillation has numerous different underlying causes, both from different types of primary cardiac disease and from underlying systemic illness triggering a secondary atrial fibrillation as a result. As such, studies should target specific patients’ modifiers to better stratify magnesium’s utility based on underlying etiology. It is likely that certain patients such as those with underlying electrolyte abnormalities (e.g. hypomagnesemia or hypokalemia) may have more favorable responses to IV magnesium. Some patients such as those with severe heart failure or chronic kidney disease may be subject to additional harms from IV magnesium.

Lastly, it must be emphasized that in this review IV magnesium use is assessed as an adjunct to other treatments for atrial fibrillation. The evidence for efficacy of intravenous magnesium as a sole agent is lacking. Therefore, intravenous magnesium should not be used as the primary agent in treatment of atrial fibrillation with RVR, pending further studies.

The available evidence indicates a modest benefit form IV magnesium as an adjunct in management of atrial fibrillation with RVR with limited harms. However, significant heterogeneity particularly in regards to magnesium dose, infusion rate, and underlying pathologies or comorbidities threatens the validity of the results. Additionally, many of the patients in the included trials received digoxin as the first line treatment, a treatment that is rarely used in current practice. Therefore, we have assigned an NNT color recommendation of Yellow (Unclear if benefits/More data needed) to this intervention.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Pierre-Carole Tchouapi, MD; Kevin E. Anderson; Paul Narayan Hein, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 20, 2023

Source

Silva TW, Valerio C, Gawrys B. Magnesium sulfate infusion to prevent hospitalization for acute copd exacerbations. afp. 2022;106(5):498A-498B.

Study Population: Adults 35 years and older in seven countries presenting to an emergency department (ED) with an acute exacerbation of chronic obstructive pulmonary disease (COPD)

Efficacy Endpoints

Hospital admissions from the ED and the need for noninvasive ventilation, assisted ventilation, or intensive care unit admission

Harm Endpoints

Adverse events or serious adverse events

Narrative

COPD is a chronic, progressive disease often complicated by exacerbations that commonly lead to hospital admissions, decreased quality of life, and increased morbidity and mortality. Novel strategies have been suggested to decrease the rate of hospitalizations for acute COPD exacerbations.

Some evidence suggests that hypomagnesemia increases airway hyperreactivity, impairs pulmonary function, and increases the risk of COPD exacerbations.1^, 2 A 2008 study showed that hypomagnesemia is an independent predictor of readmission to the hospital for acute COPD exacerbations.3 Magnesium sulfate infusions are used as adjuvant therapy for asthma exacerbations because of their bronchodilatory effect.4 Similarly, magnesium sulfate may have potential as adjuvant therapy for COPD exacerbations.

A 2022 Cochrane review evaluated magnesium sulfate in the management of acute COPD exacerbations in the ED. This review included 11 randomized controlled trials (10 double-blind and one single-blind), with 762 participants in seven countries (United States, Iran, Turkey, Nepal, New Zealand, United Kingdom, Tunisia).5 The primary outcomes were hospital admissions and the need for noninvasive ventilation, assisted ventilation, or intensive care unit admission.

Seven of the studies examined magnesium sulfate infusions, three studies assessed nebulized magnesium sulfate inhalation, and one study examined both. Comparisons included placebo alone (five studies), placebo plus standard care (three studies), placebo plus alternative nebulized solution (two studies), and nebulized ipratropium plus intravenous saline (one study).

Three of the studies in the Cochrane review evaluated hospital admissions from the ED. These studies included 170 patients from the United States, Iran, and New Zealand. Low-certainty evidence suggested a reduction in hospitalizations with magnesium sulfate infusion compared with placebo (odds ratio = 0.45; 95% CI, 0.23 to 0.88; number needed to treat = 7).6^, 7^, 8 A number needed to harm could not be calculated because no adverse events were noted in these studies. The analysis found no significant difference in hospital admissions with nebulized magnesium sulfate compared with placebo, standard care, or nebulized ipratropium. The analysis did not demonstrate a statistical difference for the other primary outcomes.

Caveats

The meta-analysis had a small sample size of 762 people. Of the 11 studies included, five were considered to be at low risk of bias, and six were considered to be of unclear bias. Each study had a small number of participants. Of the studies that examined magnesium sulfate infusion for the treatment of acute COPD exacerbation in the ED, only one reported adverse events.8

None of the studies quantified COPD severity, making it difficult to know if comparison groups were similar. Despite previous data showing a correlation between low magnesium levels and hospital readmission, none of the three studies evaluating hospital admissions from the ED reported magnesium levels on initial presentation to the ED; therefore, it is difficult to determine if patients who benefited from magnesium sulfate infusion were truly hypomagnesemic at the time of initial presentation.

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Taran W. Silva, DO; Christina Valerio, MD, MPH; Breanna Gawrys, DO
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Brown B, Price ST. Mu-opioid antagonists for the treatment of opioid-induced bowel dysfunction. afp. 2023;107(2):131-132.

Study Population: 1,343 adults in 10 randomized controlled trials who have cancer (any stage) or are receiving palliative care (any terminal disease)

Efficacy Endpoints

Laxation response

Harm Endpoints

Effect on analgesia and adverse events

Narrative

Opioids are often used to manage moderate to severe pain in patients with cancer and patients with or without cancer receiving palliative care. The benefit of opioid analgesics occurs through binding to mu-, kappa-, and delta-opioid receptors. Activation of mu-opioid receptors in the wall of the gut leads to disruption of gastrointestinal propulsive motility, causing decreased peristalsis, delayed gastric emptying, and ultimately opioid-induced bowel dysfunction, a significant adverse effect of opioid use.1

Opioid-induced bowel dysfunction is a change in baseline bowel habits characterized by decreased frequency of bowel movements (less than three per week), increased straining to pass bowel movements, or the inability to complete rectal evacuation.2 Patients may also experience nausea, emesis, bloating, abdominal cramping, gastric reflux, and abdominal distention. These symptoms can cause psychological distress, agitation, and increased morbidity for patients with cancer and those receiving palliative care.

A comprehensive regimen to decrease constipation in patients with opioid-induced bowel dysfunction commonly includes a laxative stimulant, stool softeners, and continuing general laxation measures. However, despite these interventions, more than 80% of patients with opioid-induced bowel dysfunction continue to experience constipation, 33% change their pain medication regimen, and 92% of those who changed medications subsequently experience increased pain.3 Mu-opioid antagonists bind to mu-opioid receptors in the wall of the gut and neutralize the constipating effect of opioids. Mu-opioid antagonists are commonly used as laxation augmentation or as an alternative when laxatives fail.4

A systematic review of 10 randomized controlled trials evaluated the benefits of mu-opioid antagonists on laxation response, effect on analgesia, and adverse events in 1,343 adults with cancer or those receiving palliative care with or without cancer.5 Mu-opioid antagonists included oral naldemedine (Symproic) or subcutaneous methylnaltrexone (Relistor). The randomized controlled trials compared mu-opioid antagonists with placebo, at different doses, or in combination with laxative medications.

Patients taking oral naldemedine had a greater than threefold increase in spontaneous laxation at two weeks compared with placebo (risk ratio [RR] = 2.0; 95% CI, 1.59 to 2.52; absolute risk difference [ARD] = 35%; number needed to treat [NNT] = 3; moderate-certainty evidence). Naldemedine had little to no effect on analgesia and did not increase the risk of serious adverse events. The risk of nonserious adverse events (most commonly diarrhea) was higher for naldemedine compared with placebo (RR = 1.49; 95% CI, 1.19 to 1.87; ARD = 17%; number needed to harm = 6; moderate-certainty evidence). Higher doses of naldemedine may be associated with an increased rate of spontaneous laxation in the medium term (0.1 mg vs. 0.4 mg; RR = 0.69; 95% CI, 0.53 to 0.89; ARD = 25%; NNT = 4; low-certainty evidence).

The rate of spontaneous laxation was four times greater at 24 hours with subcutaneous methylnaltrexone than placebo (RR = 2.97; 95% CI, 2.13 to 4.13; ARD = 45%; NNT = 3) and 10 times greater at two weeks (RR = 8.15; 95% CI, 4.76 to 13.95; ARD = 65%; NNT = 2; moderate-certainty evidence). Although subcutaneous methylnaltrexone had a higher overall risk of adverse events compared with placebo (RR = 0.59; 95% CI, 0.38 to 0.93; ARD = 10%; number needed to harm = 10; low-certainty evidence), it reduced the risk of opioid withdrawal symptoms and did not increase the risk of serious adverse events.

Caveats

The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) rating of the evidence for the use of oral naldemedine and subcutaneous methylnaltrexone for opioid-induced bowel dysfunction was evaluated as moderate certainty; however, the evidence for adverse events was considered low certainty. Although naldemedine and methylnaltrexone demonstrated laxation benefits, studies had significant heterogeneity in comparison and delivery of mu-opioid antagonists, reporting of adverse events, and timing of follow-up. Eight of the 10 trials were industry sponsored.

The American Gastroenterological Association Institute guidelines for patients with opioid-induced constipation recommend traditiona l laxatives as first-line therapy. If laxatives do not provide an adequate laxation response, escalation of therapy to naldemedine is recommended, with methylnaltrexone following as a conditional recommendation.6 Additional controlled studies with a focused evaluation of treatment benefits and potential adverse events over a longer follow-up period are needed.

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Brandon Brown, MD, FHM; S. Tucker Price, MD, PhD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 13, 2023

Source

Haley SP, Stem LA. Corticosteroids for low back pain. afp. 2023;107(3):230A-230B.

Study Population: Adults 18 years or older with acute or chronic low back pain in one of three diagnostic groups: radicular low back pain, nonradicular low back pain, or spinal stenosis

Efficacy Endpoints

Improvement in pain and function, need for surgery, and quality of life

Harm Endpoints

Adverse events, serious adverse events, transient hyperglycemia, and withdrawal from the trial

Narrative

Acute low back pain is a common reason adults present to primary care. Low back pain is usually self-limited; however, 31% of people with low back pain do not fully recover within six months.1 Physicians often use corticosteroids to treat acute low back pain with and without radicular symptoms, but the benefits and harms are unclear. Corticosteroids, when used orally, intramuscularly, or intravenously, or when injected directly into spinal structures, theoretically work to mitigate symptoms of low back pain through their anti-inflammatory properties. The 2017 American College of Physicians guideline did not include a recommendation for the use of systemic corticosteroids for low back pain.2

This Cochrane review examined the utility of systemic corticosteroid administration in patients with low back pain.3 The review identified 13 studies (n = 1,047) that evaluated the effectiveness of systemic corticosteroids in the treatment of low back pain with radicular symptoms (nine studies), low back pain without radicular symptoms (two studies), and spinal stenosis (two studies). Nine of the studies were randomized controlled trials conducted in the United States (six trials), Australia (one trial), Brazil (one trial), and Iran (one trial). The studies included patients from primary care settings, specialty clinics, or emergency departments and compared corticosteroids vs. placebo. Corticosteroid options (prednisone, prednisolone, methylprednisolone, dexamethasone, betamethasone, hydrocortisone) and dosing regimens varied. Studies of back pain caused by cancer, cauda equina syndrome, fracture, or pregnancy were excluded.

The primary outcomes of these studies were pain (continuous pain measured by pain scales or a dichotomous outcome measured by categories) and function (defined by methods in original trials). The studies also looked for severe adverse reactions and transient hyperglycemia.

For low back pain with radiculopathy, corticosteroids were found to be more effective than placebo for short-term (two weeks to less than three months) pain control when measured as a continuous outcome on a scale of 0 to 10 (mean difference = −0.56; 95% CI, −1.08 to −0.04; n = 430), with higher numbers signifying more severe pain. No significant difference in pain control was found in nonradicular low back pain or spinal stenosis when treated with systemic corticosteroids in the immediate term (less than two weeks) or long term (12 months or greater).

No improvement in function was reported in the immediate term regardless of the type of back pain. In the short term, patients with radicular low back pain reported improved function after receiving corticosteroids compared with placebo (risk ratio [RR] = 1.52; 95% CI, 1.22 to 1.91; absolute risk difference [ARD] = 17.8%; number needed to treat [NNT] = 6; moderate-quality evidence; n = 403), but this effect was not statistically significant for patients with nonradicular pain. The one study reporting on long-term function showed a benefit with corticosteroids over placebo in radicular low back pain (RR = 1.29; 95% CI, 1.06 to 1.56; ARD = 17; NNT = 6; n = 267).

Three trials reported on transient hyperglycemia, defined as a short-term blood glucose increase of 50 mg per dL (2.77 mmol per L) or greater. There was no significant difference between corticosteroids and placebo.

Caveats

The Cochrane review included subgroup analyses, of which some demonstrated a benefit of corticosteroids in low back pain and others did not, making it difficult to determine an overarching conclusion. The differences in study type, outcomes measured, and treatments delivered among studies limit generalizability. This meta-analysis did not establish an optimal corticosteroid dose and duration for radicular low back pain.

All significant outcomes were found within the subgroup analysis of radicular back pain. Short-term pain improvement was clinically modest, at an average of a 0.5-point improvement on a scale of 0 to 10. Among the five studies that measured pain as continuous, only two had a low risk of bias in all domains. Among studies evaluating function dichotomously, the definition varied from “return to normal activities,” “no disability,” or “improvement on a disability scale.” Short-term functional improvement was based on three studies (n = 403), with only one study identified as having a low risk of bias in all domains. Improvement in long-term function was based on one study (n = 267) with a low risk of bias in all domains and was defined as a 50% or greater improvement on the Oswestry Disability Index. The definitions of pain categories, pain scales, and functions varied across studies, further limiting conclusions.

The meta-analysis reported low certainty about the risks associated with treating low back pain with corticosteroids. Many studies were missing data on relevant harms from corticosteroids, such as hyperglycemia, which limited the accurate reporting of potential harms. Of the three studies including these data, only two had information on steroid dosing. These studies focused only on disease-oriented outcomes (degree of transient hyperglycemia) without data on hospitalization, titration of medication, or long-term outcomes such as bony fractures. The inability to accurately weigh the benefits of treatment against associated risk is an important factor as physicians consider the modest benefit in pain reduction and improved function in radicular back pain, especially in patients at higher risk of complications.

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Sean P. Haley, MD, MPH; Leah A. Stem, MD, MMEd
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Beck RA, Spiva S. Corticosteroids for the treatment of kawasaki disease in children. afp. 2023;107(1):20-21.

Study Population: Children younger than 19 years diagnosed with Kawasaki disease (KD)

Efficacy Endpoints

The primary intervention outcome was any coronary artery abnormality found on cardiac angiography or echocardiography within three months of a KD diagnosis; abnormalities were defined by de Zorzi criteria (coronary lumen dimension 2.5 standard deviations or more above the mean for body surface area) or specified Japanese Ministry of Health criteria based on the patient's age and lumen diameter; secondary treatment outcomes included fever duration, length of hospitalization, and mortality

Harm Endpoints

Any serious adverse event attributable to the administration of corticosteroids

Narrative

KD (i.e., mucocutaneous syndrome) is a multisystem vasculitis that may be related to an abnormal host response to an infection.1 Although KD is the leading cause of childhood-acquired heart disease in high-income countries, the incidence in children younger than five years varies, with 20.8 per 100,000 in the United States and 239.6 per 100,000 in Japan. Because there is no diagnostic test for KD, the diagnosis is based on clinical features or symptoms defined by the American Heart Association or the Japan KD Research Committee guidelines. Determining a diagnosis is challenging because clinical symptoms of KD are prevalent in common childhood viral illnesses (e.g., fever, rash, conjunctivitis, cervical lymphadenopathy).

The most significant complication of KD is a predisposition for coronary artery abnormalities, which can lead to aneurysm formation in up to 25% of untreated patients.2 Prompt treatment decreases coronary complications. The mainstay of initial treatment is intravenous immune globulin and aspirin. Intravenous corticosteroids have been recommended for patients at high risk (e.g., history of cardiac abnormalities, diagnosis before 12 months of age, clinical features of shock) or if the initial treatment regimen fails (up to 20% of patients), although data are limited.

The Cochrane review analyzed eight randomized trials with 1,877 children (younger than 19 years) worldwide diagnosed with KD.1 All forms of corticosteroids in conjunction with any combination of placebo, no treatment, intravenous immune globulin, aspirin, or infliximab were studied. Comparators included monotherapy or a combination of the previously mentioned interventions. Corticosteroid administration was broadly grouped into a single pulsed dose of intravenous methylprednisolone or a longer tapering course of oral prednisolone.

Moderate-quality evidence showed that when compared with no corticosteroid use, corticosteroids administered in the acute phase of KD reduced coronary artery abnormalities over a two- to six-week follow-up period (odds ratio = 0.32; 95% CI, 0.14 to 0.75); absolute risk difference = 10.7%; number needed to treat = 10). There were no reported serious adverse events in the included studies.

Moderate-certainty evidence demonstrated a reduction in the length of hospital stay (mean difference = –1.01 days), and low-certainty evidence demonstrated a decrease in fever duration (mean difference = –1.34 days) in patients treated with corticosteroids compared with no corticosteroids. There were no deaths reported in the included studies.

Caveats

The Cochrane review considered trials involving children diagnosed with KD worldwide. Four of the eight trials had an unclear risk of selection bias, seven had an unclear risk of performance bias, four had an unclear risk of detection bias, and one had an unclear risk of attrition bias.

A subgroup analysis revealed that corticosteroids used as first-line therapy led to a reduction in coronary artery abnormalities, but this effect was not found in a subgroup analysis of corticosteroids used as second-line therapy (e.g., after the failure of intravenous immune globulin and aspirin).

Six of the eight included trials were performed in Japan or South China. The two North American studies involved children at lower risk and used single-dose corticosteroid regimens, limiting the overall applicability in the United States.

The subgroup analyses performed in the systematic review provided limited information because of the small sample sizes. Studies did not include long-term data involving outcomes more than one year after KD diagnosis. Larger studies are needed to target patient-centered outcomes such as mortality. The studies need to provide longer follow-up periods to properly assess the benefits and harms of this intervention.

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Robert A. Beck, MD; Sara Spiva, DO
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

January 26, 2023

Source

Long B, Gottlieb M. Anterior‐posterior versus anterior‐lateral pad positioning for electrical cardioversion of atrial fibrillation. Academic Emergency Medicine. Published online January 13, 2023:acem.14662.

Study Population: 10 trials with 1,677 patients with atrial fibrillation undergoing electrical cardioversion

Efficacy Endpoints

Rate of successful cardioversion, success at low energy

Harm Endpoints

Not reported

Narrative

Atrial fibrillation (AF) is a common arrhythmia evaluated and managed in the emergency department (ED), and treatment may include electrical cardioversion.1^, 2^, 3 Success of electrical cardioversion depends on several factors including AF duration and, possibly, pad positioning.4^, 5 The most common pad positions are the anterior-posterior (AP) and anterior-lateral (AL), and several randomized controlled trials (RCTs) have compared the two.4^, 5^, 6^, 7^, 8

The systematic review summarized here included RCTs comparing AP and AL pad positioning during elective electrical AF cardioversion.9 To be included, RCTs had to report cardioversion success rate, which was the primary outcome. Other outcomes included cardioversion success at low energy (≤200 J monophasic or ≤120 J biphasic), total number of shocks, mean transthoracic impedance, and mean shock energy.

The meta-analysis identified 10 RCTs with 1,677 subjects (831 AP, 846 AL).9 Two trials were multicenter.4^, 8 AP positioning involved placement at the right sternal border and left infrascapular region, except for one trial that used the left parasternal and left lower scapular region.4 AL positioning involved placement at the cardiac apex and right infraclavicular region. Five RCTs used biphasic cardioversion. The mean patient age was 64 years, and 32% of patients were female. Most RCTs included patients with persistent AF, except for one which had 20% with paroxysmal AF.4 Anti-arrhythmic choice varied.

There was no difference in cardioversion success between the AP and AL pad position (86.6% vs. 87.9%, risk ratio: 1.00; 95% confidence interval [CI]: 0.9 to 1.1). Meta-regression demonstrated no difference by body mass index, left atrial diameter, valvular heart disease, or AF duration. Subgroup analysis found no difference in monophasic or biphasic cardioversion. Among secondary outcomes there was no difference in success at low energy, in number of delivered shocks, or in mean energy delivered. AP pad position was associated with lower transthoracic impedance (standardized mean difference: -0.3 ohms; 95% CI: -0.5 to -0.1).

Caveats

While this systematic review found no difference in successful cardioversion for AF using AP or AL pad positioning, there are several limitations. There was significant heterogeneity in study design, setting, and outcome. The sequence and escalation of shocks varied widely or was not reported, and the included studies provide little detail of the study setting (e.g., emergency department, inpatient setting). While all studies used restoration of sinus rhythm, the time threshold varied or was not specified. In addition, six RC Ts included only persistent AF, which is less responsive to electrical cardioversion. Only patients undergoing elective cardioversion were included. Finally, numbers were small and limited for useful subgroup analysis.

Importantly, monophasic waveform defibrillators are not routinely utilized in North America as of 2023. Of the included studies evaluating biphasic shocks, two studies found no difference with AP and AL positioning, while two others found AL positioning to be superior. The largest study used biphasic shocks with a step-up energy strategy and found AL positioning to be superior.410 However, no difference was seen at energy levels >200 J.10 Of note, a prospective RCT including 244 ED patients with paroxysmal AF published in 2020 found no difference in AL versus AP pad positioning using biphasic shocks at ≥200 J.11

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

January 19, 2023

Source

Long B, Gottlieb M. intra‐articular lidocaine versus intravenous sedation for anterior shoulder dislocation reduction. Academic Emergency Medicine. Published online January 2023:acem.14653.

Study Population: 12 trials of 630 ED patients ≥15 years undergoing closed reduction for acute anterior shoulder dislocation

Efficacy Endpoints

Successful reduction, adverse events, emergency department length of stay, pain score, procedure time and patient satisfaction

Harm Endpoints

None

Narrative

Anterior shoulder dislocation is the most common large joint dislocation managed in the emergency department (ED).1^, 2^, 3^, 4 Several analgesic techniques are available to assist with reduction, including intra-articular lidocaine, nerve blocks, and procedural sedation with intravenous (IV) medications.2^, 3^, 4 Each technique has distinct advantages (e.g., reduced length of stay with intra-articular lidocaine) and disadvantages (e.g., adverse event such as respiratory depression with procedural sedation). Indeed, several studies have compared these techniques, with a 2014 meta-analysis finding similar efficacy in reduction success rate using either procedural sedation or intra-articular lidocaine with fewer adverse events in the intra-articular lidocaine group.2 Since 2014, additional randomized controlled trials (RCT) have been published, providing further data comparing intra-articular lidocaine and IV sedation.5^, 6^, 7

The systematic review and meta-analysis summarized here included RCTs comparing intra-articular lidocaine and IV sedation in ED patients with acute anterior shoulder dislocation.8 Participants were ≥15 years of age undergoing closed reduction for treatment of the dislocation. Studies with pediatric populations, posterior dislocations, fracture-dislocations, and settings other than the ED were excluded. The main outcomes included reduction success rate, adverse events, procedure time, and patient satisfaction. Successful reduction was defined as confirmation of relocation in post-reduction imaging. Adverse events were defined by the individual RCT. The meta-analysis identified 12 RCTs published between 1994 to 2020 with 630 patients (327 patients receiving intra-articular lidocaine and 303 patients receiving IV sedation) conducted in the ED setting.8 All patients in the intra-articular lidocaine group received 1% lidocaine, with 10 trials administering 20 mL of 1% lidocaine and 2 trials administering 4 mg/kg (maximum 200 mg). The type of IV sedation medications varied widely across all studies, with a benzodiazepine (diazepam, midazolam) in combination with an opioid (meperidine, morphine, or fentanyl) being the most common sedation regimen used.

There was no difference in reduction success rate nor pain scores between groups. There were, however, reductions in adverse events (RR 0.16; 95% CI, 0.07 to 0.33; number needed to treat: 6; absolute risk reduction: 19.5%), shortened ED length of stay (mean difference: -1.48 hours; 95% CI -2.48 to -0.47), and a shortened procedure time (mean difference: 8 min; 95% CI: 4.42 to 11.57) in the intra-articular lidocaine group. The reported adverse events included agitation and drowsiness associated with intra-articular lidocaine and respiratory depression (including apnea/hypoxia), hypotension, nausea/vomiting, headache, allergic reaction, and thrombophlebitis in patients assigned to the IV sedation group. The systematic review also reported patient satisfaction for each technique. The patient satisfaction rate was lower with intra-articular lidocaine (70.5% vs. 90.4%; RR 0.80: 95% CI, 0.67 to 0.95, moderate certainty). However, this outcome was subject to significant heterogeneity due to variability in the tools used for measuring it.

Of note, a recent network of meta-analysis was published in 2022 that evaluated intra-articular anesthetic injection, intravenous sedation, and peripheral nerve block for shoulder dislocation reduction.9 This meta-analysis found no differences in reduction success rate, adverse respiratory events, or patient satisfaction between the techniques. However, it did identify a longer time for the reduction procedure with intra-articular analgesia compared to IV sedation. As in the previous meta-analysis, intra-articular analgesia resulted in a shorter length of ED stay.8^, 9

Caveats

Intra-articular lidocaine and IV sedation can be effective in facilitating the reduction of anterior shoulder dislocations, with distinct advantages and disadvantages for each. This meta-analysis found no significant difference in reduction success rate or pain scores between either option.8 Intra-articular lidocaine was associated with fewer adverse events, shorter ED length of stay, and reduced procedure time, while procedural sedation was associated with higher patient satisfaction.

The presented data have several limitations. There was moderate to high heterogeneity for almost all of the evaluated outcomes except adverse events. The levels of certainty for the measured outcomes ranged from very low to moderate. Both successful reduction and pain score outcomes had low to very low levels of certainty. The significant heterogeneity and lower certainty of evidence limit the applicability of the results. Another important limitation is the variability in the specific agents used for IV sedation. Six RCTs used meperidine/pethidine with a benzodiazepine or propofol, while the other 6 RCTs used propofol, etomidate, ketamine, morphine, fentanyl, midazolam, and diazepam. The specific agent could increase the risk of adverse events, and many of the included studies relied on older regimens, as opposed to current era sedation medications (e.g., propofol, ketamine, and etomidate). Adverse events were also based on individual study definition and may have led to underreporting of some events. Moreover, reduction technique was not standardized, and success rates vary among techniques.3^, 10 Neither of these studies compare intra-articular analgesia nor IV sedation with manipulation alone. Modern reduction techniques such as the Cunningham method have demonstrated high rates of success in the hands of trained providers, without concomitant sedation.3^, 11 Finally, the efficacy of intra-articular lidocaine is dependent on clinician skill, and the included trials did not provide information concerning clinicians’ training level or whether ultrasound was used.12

Based on the available evidence, we have assigned a color recommendation of Yellow (Equal Efficacy) for the use of intra-articular lidocaine in reduction of anterior shoulder dislocation. It is likely that intra-articular lidocaine injection for anterior shoulder dislocation does not affect the procedure success rate, but it reduces the risk of adverse events and shortens ED length of stay. However, the heterogeneity among trials and suboptimal level of evidence certainty call for larger, more rigorous trials.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

December 6, 2022

Source

Long B, Gottlieb M. Surgical decompression for space‐occupying hemispheric infarction. Zehtabchi S, ed. Academic Emergency Medicine. 2021;28(12):1475-1477.

Study Population: 488 patients with space-occupying hemispheric infarction after ischemic stroke

Efficacy Endpoints

Improved functional outcome as defined by mRS ≤2, ≤3, and ≤4 at 1 year, mortality at 1 year

Harm Endpoints

No harm endpoints reported

Narrative

Ischemic stroke is associated with several significant complications. One of the most deadly is severe cerebral edema, which may occur in up to 8% of patients in the first 4 days after onset of stroke.1^, 2^, 3 Earlier meta-analyses suggested surgical decompression with hemicraniectomy and duraplasty, designed to provide space for brain swelling, may improve outcomes.4^, 5 However, study sample sizes were small with unclear results for subgroups based on aphasia, late presentation, and vascular territory.6^, 7 The review summarized here included randomized controlled trials (RCTs) of patients with supratentorial “space-occupying hemispheric infarction” (i.e., large-territory ischemic stroke expected to develop severe cerebral edema).8 Participants were randomized to surgical craniotomy versus medical treatment alone, with assessment of functional outcome at 6 to 12 months using the modified Rankin scale (mRS). The primary outcome of the review was mRS score ≤ 3 at 1 year. Secondary outcomes included death at 6 to 12 months, functional independence (mRS ≤ 2), and “reasonable” neurologic outcome (i.e., survival without severe neurologic disability; defined as mRS ≤ 4). The authors contacted investigators and requested individual patient data. In addition to overall pooling, they performed subgroup analyses based on age, sex, presence of aphasia, time to randomization, vascular territory, and National Institute of Health Stroke Scale (NIHSS). Authors adjusted analysis based on prespecified covariates, including age, sex, baseline stroke severity, aphasia, and time from stroke onset to randomization.

The review identified eight published RCTs and one complete but unpublished trial for inclusion. Of 543 patients randomized in the original RCTs, the meta-analysis included 488.8 Full data were available for seven trials, including the unpublished RCT. All trials evaluated mRS at 1 year, with all except one also evaluating mRS at 6 months. Surgical decompression increased the likelihood of a favorable outcome (mRS ≤ 3) at 1 year compared with medical treatment alone (37.2% vs. 14.6%; adjusted odds ratio [aOR] = 3.0, 95% confidence interval [CI] = 1.6 to 5.6, absolute risk reduction [ARR] = 23%, number needed to treat [NNT] = 5). Surgical decompression was also associated with reduced mortality at 1 year (aOR = 0.16, 95% CI = 0.1 to 0.2; ARR = 42%, NNT = 3), increased chance of mRS ≤ 2 at 1 year (aOR = 2.77, 95% CI = 0.97 to 7.88, ARR = 12%, NNT = 9), and an increased chance of mRS ≤ 4 at 1 year (aOR = 5.34, 95% CI = 3.26 to 8.74, ARR = 38%, NNT = 3).

Caveats

There are several limitations associated with these data. Overall outcomes were poor with less than 25% of patients surviving with an mRS ≤ 3. Of note, the mean patient age was in the 40s in the majority of included RCTs, with few patients over age 60. In patients over 60 years, estimates of treatment outcomes were less precise due to low numbers of favorable outcomes. Only 8% of patients over 60 years demonstrated a favorable outcome in the DESTINY II trial, compared with the unpublished DEMITUR trial, which demonstrated a favorable outcome in 66% of patients.8^, 9^, 10 This difference may be due to patient characteristics or adjudication of mRS outcomes. It is unclear if younger patients are being offered this intervention because they are younger or have better premorbid functional status or some other reason. Distribution of harms in this younger population is unclear as well.

Of note, the mRS is a score that measures the degree of disability or dependence in daily activities of patients with a neurologic condition from 0 points (no symptoms) to 6 points (death).11 Several studies evaluating surgical decompression for ischemic stroke have utilized mRS of ≤3 (moderate disability, unable to walk and attend to bodily needs without assistance) as a favorable outcome,12^, 13 including this current meta-analysis, while other studies evaluating thrombolysis utilize a score of ≤2 (slight disability; unable to carry out all previous activities, but able to look after own affairs without assistance).14 While there appears to be significant difference in functional ability between a score of 2 and 3, literature suggests patients with scores of 2 and 3 have similar quality of life.15 Literature also suggests a wide inter-rater reliability in assessment of mRS.16 Additionally, individual studies were small, ranging from 32 to 151 participants. Data reporting was also not complete for all trials. Due to the nature of the treatment blinding was not possible, which may have affected treatment approaches, goal-of-care decisions, and cointerventions. Outcome assessment was also not blinded, and with no way of blinding clinicians and the small number of included patients, biases are likely present. Limitations in study data availability prevented full analysis of all prespecified subgroups. In addition, quality of life was not included as an outcome due to limitations in use of these instruments in the included RCTs and the impact of survival bias. Finally, craniotomy is invasive, burdensome, and resource-intensive and may cause substantial postsurgical complications.17 Unfortunately, no harm data are reported in this review of trials testing a major surgical procedure, which is a significant limitation. Thus, any negative impact of this procedure on quality of life, particularly for populations who may not benefit, is unclear.

Based on the available evidence, the review summarized here found surgical decompression was associated with improved functional outcome in those with space-occupying hemispheric infarction. Thus, we have assigned a color recommendation of green (benefit > harm). Further study is needed evaluating surgical decompression and quality of life and subgroups such as those receiving decompression 48 hours after stroke onset. However, in patients with large territory ischemic stroke who may develop cerebral edema, emergency clinicians should consult a neurosurgical specialist and assess the risks and benefits of further intervention.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

November 17, 2022

Source

Long B, Gottlieb M. Liberal versus conservative intravenous fluid administration in pediatric diabetic ketoacidosis. Academic Emergency Medicine. Published online November 3, 2022:acem.14596.

Study Population: Three trials of 1323 children aged 0–18 years presenting to emergency departments with diabetic ketoacidosis

Efficacy Endpoints

Reduction in GSC (primary), cerebral edema, hospital length of stay

Harm Endpoints

Adverse events such as pulmonary edema, thrombosis, alteration in mental status, kidney injury, electrolyte imbalances

Narrative

Diabetic ketoacidosis (DKA), a dangerous condition of elevated blood glucose, dehydration, and acidosis, is associated with morbidity and mortality in children.1 Much of this is due to neurologic damage and deaths from DKA-related brain injury, a condition occurring in 0.3%–0.9% of children with DKA.2^, 3^, 4 Unfortunately, the cause of cerebral edema in DKA is unknown, and there is controversy about whether rate, volume, or type of intravenous (IV) fluid therapy plays a role.5^, 6^, 7

The systematic review summarized here included randomized controlled trials (RCTs) comparing liberal (fast) versus conservative (slow) IV fluid infusion during management of children with DKA.8 Participants were under 18 years of age and required IV fluid and insulin infusion. The primary outcome in the systematic review was reduction in Glasgow Coma Scale (GCS). Secondary outcomes included development of cerebral edema and hospital length of stay (LOS). Adverse events included pulmonary edema, development of kidney injury, and electrolyte imbalances.

The meta-analysis identified three RCTs conducted in the emergency department, one of which contributed 1389 out of 1457 (95%) of the total DKA episodes analyzed.9 Concentrations of saline infusion were mostly either 0.45% or 0.9%. All three trials compared 20 mL/kg (liberal) with 10 mL/kg administered as IV bolus or over 1 h; however, the definition of “liberal” and “conservative” differed among the three trials. The largest trial used a two-by-two factorial design also comparing 0.45% to 0.9% saline solution.9 No study reported DKA severity, mortality, PICU admission, or development of kidney injury.

There was no difference in reduction in GCS (n = 1361), development of cerebral edema (n = 1439), or hospital LOS (n = 1439). There were no cases of pulmonary edema or thrombosis reported. Three percent of the total study population experienced a serious adverse event, most commonly an alteration in mental status. Hyperchloremic metabolic acidosis and hypocalcemia were more common in the liberal infusion group, but rates of hypoglycemia and hypokalemia were similar.9 Subgroup analysis evaluating <0.9% saline compared with 0.9% saline found no difference in outcomes.

Caveats

No significant differences in outcomes between fluid rate or fluid choice were identified in this meta-analysis. However, the presented data have several limitations. The most significant limitation is that one trial accounts for 95% of patients and episodes in the analysis, and this trial's results have not been reproduced or externally validated.9 This trial incorporated a two-by-two factorial design based on fluid infusion amount and saline concentration.9 There was also important heterogeneity in study designs and primary outcomes among the included trials. While all studies used saline, they differed in maintenance infusions and concentrations. Moreover, “evidence certainty” was rated as low to very low because all trials were at high risk of bias due to a lack of blinding of participants, providers, or outcome assessors. The wide confidence intervals for the effect sizes are likely due to small sample sizes and heterogeneity, suggesting the need for trials with larger sample sizes with more homogenous populations. Finally, no trial used “balanced” electrolyte solutions (e.g., Ringer's lactate).

Based on the available evidence, we have assigned a color recommendation of yellow (unclear if benefit) for liberal versus conservative IV fluid regimens. The high risk of bias and lack of any trials reproducing the findings of the lone large study suggest larger, more rigorous trials will be critical. Future studies evaluating IV fluid administration and IV fluid type with clear, patient-centered outcomes are necessary. It is reassuring that the overall conclusion of the systematic review is in line with the rigorous randomized trial published in 2018, which was included in the systematic review and demonstrates no evidence that fluid rate or fluid choice has an important impact on cerebral edema or adverse outcomes in children with DKA.9

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

October 25, 2022

Source

Long B, Gottlieb M. Oral fidaxomicin versus vancomycin for Clostridioides difficile infection. Academic Emergency Medicine. Published online October 11, 2022:acem.14600.

Study Population: Six trials of 1390 patients aged ≥16 years with Clostridioides difficile infection

Efficacy Endpoints

Global cure, clinical cure, and recurrent infection

Harm Endpoints

Adverse events

Narrative

Clostridioides difficile infection (CDI) is a serious gastrointestinal disease commonly caused by antibiotics.1^, 2 In 2019, the Centers for Disease Control and Prevention reported CDI as an urgent threat to public health.3 CDI can range from mild diarrhea to severe disease with toxic megacolon.4 Unfortunately, the disease often recurs after therapy at rates between 30 and 65%.5^, 6^, 7 Vancomycin and metronidazole are common treatments for CDI, but recent Infectious Diseases Society of America (IDSA) guidelines have added oral fidaxomicin, a narrow-spectrum macrocyclic antibiotic, as a first-line option.8 The systematic review summarized here compared the efficacy of oral fidaxomicin and vancomycin for treatment of CDI.9

The review included randomized controlled trials (RCTs) comparing oral fidaxomicin and vancomycin for treatment of CDI.9 Participants were ≥16  years of age. The primary outcome was global cure, defined as clinical cure without recurrence. Secondary outcomes included clinical cure rate, recurrence rate, and adverse events.

The meta-analysis identified six RCTs (n = 1390 patients), three of which were noninferiority studies. Five studies were performed in the United States, Canada, or Europe, and one study was performed in Japan. One study included patients with initial CDI,10 one study included only patients with recurrent CDI,11 and three studies included patients with both initial and recurrent CDI. Roughly one-third of subjects in the three RCTs that reported severity had severe CDI. Follow-up ranged from 38 to 66 days in the included trials.

Oral fidaxomicin was associated with higher global cure rates (relative risk [RR] 1.2, 95% confidence interval [CI] 1.1–1.3, absolute risk difference [ARD] 10.7%, NNT 10) and lower recurrence rates (RR 0.6, 95% CI 0.5–0.8, ARD 10.6%, NNT 10) compared to vancomycin. There was no difference in clinical cure between fidaxomicin and vancomycin. There were no significant differences for adverse events between fidaxomicin and vancomycin.

Caveats

Treatment of CDI can be challenging due to the risk of recurrence. While this meta-analysis suggests oral fidaxomicin may be more effective than vancomycin for treatment of CDI, there are several limitations. Industry-sponsored studies represent >90% of the data in this review. Industry funding is a key risk factor for both visible and invisible bias favoring the company's product.12 These RCTs were also structured as noninferiority, not superiority, trials. Two trials were open-label, further increasing the risk of bias. The definitions of outcomes in each study were not identical, which introduces clinical heterogeneity. Outcome definitions varied across the trials, making the final measurement less reliable. Studies also differed in follow-up period and time point for evaluation of outcomes. The inclusion of patients with first-time CDI versus recurrence is another limitation, as these may reflect different populations with different potential risk for recurrence. Most of the studies were performed in western countries, and regional differences may affect the results due to diet, antibiotic use, and C. difficile strains.

Current IDSA guidelines recommend fidaxomicin over vancomycin for treatment of an initial or recurrent episode of CDI.8 Both medications are preferred over metronidazole. However, the IDSA guidelines recommend vancomycin administered orally or by nasogastric tube rather than fidaxomicin for patients with fulminant CDI.8

Based on the available evidence, we have assigned a color recommendation of yellow (unclear if benefit) for oral fidaxomicin compared with oral vancomycin for treatment of CDI. The risk of bias and heterogeneity suggest a need for larger, blinded trials that are not industry-supported. Further data are also needed regarding fidaxomicin for different severities and strains of CDI.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

August 4, 2022

Source

Ford B, Anderson A, Nichols C. Membrane sweeping at term to induce labor. afp. 2022;106(1):21A-21B.

Study Population: Patients who are pregnant with a live fetus at or near term (36 weeks or greater estimated gestational age) without indication for induction or urgent delivery

Efficacy Endpoints

Primary: spontaneous onset of labor, need for induction of labor, need for cesarean delivery, spontaneous vaginal delivery; secondary: instrument-assisted vaginal birth, epidural analgesia

Harm Endpoints

Primary: maternal death or serious morbidity, uterine hyperstimulation with or without fetal heart rate deceleration, neonatal death, serious perinatal morbidity; secondary: postpartum hemorrhage, uterine rupture, augmentation of labor

Narrative

Approximately 20% to 25% of deliveries occur after labor induction in middle-to high-income countries.1 Membrane sweeping is a common outpatient intervention associated with cervical dilatation assessment. Membrane sweeping is the advancement of one or two fingers through the internal cervical os to the lower uterine segment, followed by a circular sweeping movement between the amniotic membrane and the lower uterine segment. Successful performance of this procedure can decrease the need for formal induction with pharmacologic or mechanical methods or the need for hospitalization or advanced monitoring. The process of membrane sweeping leads to the release of phospholipase A2 and prostaglandin F_2α, which directly contribute to cervical ripening through an inflammatory cascade.2

A 2020 Cochrane review included 44 studies and 6,940 patients in 19 countries (14 from the United States).2 The trials compared membrane sweeping with expectant management, sham membrane sweeps, and several induction methods, including vaginal and intracervical prostaglandins and intravenous oxytocin (Pitocin). Seventeen randomized controlled trials (RCTs) with 3,170 patients compared membrane sweeping with expectant management, demonstrating a relative risk of 1.21 (95% CI, 1.08 to 1.34), an absolute risk difference of 12.5%, and a number needed to treat of 8 for spontaneous onset of labor. Sixteen RCTs with 3,224 patients demonstrated a relative risk of 0.73 (95% CI, 0.56 to 0.94) for requiring labor induction. This corresponds to an absolute risk reduction of 8.5% and a number needed to treat of 12 to prevent the need for further mechanical or pharmacologic induction of labor. Other primary outcomes, including the likelihood of spontaneous vaginal delivery, were not significantly different.

Evidence certainty was low to moderate for all primary outcomes. Patient perception of membrane sweeping was positive. In one study in the Netherlands (n = 742), 88% of patients noted they would opt for membrane sweeping in subsequent pregnancies if it were offered; 31% characterized the procedure as not painful, 51% as somewhat painful, and 17% as painful or very painful. Even among those who had pain, 88% indicated they would opt for the procedure again.3

Caveats

Despite the good numbers needed to treat and large sample sizes, the studies included in the Cochrane review had low certainty of evidence as assessed using the GRADE approach. The authors performed sensitivity analyses and excluded studies at high or unclear risk of bias for sequence generation or allocation concealment and studies with high or unclear risk of attrition. When sensitivity analyses were performed, including 12 of the 40 identified trials with low risk of bias, the result for spontaneous onset of labor was no longer statistically significant. Moderate heterogeneity was present in these included trials (Tau² = 0.00; I² = 37%; P = .16).

This Cochrane review predates the ARRIVE trial, which examined the safety of induction vs. expectant management at 39 weeks’ estimated gestational age. The results of ARRIVE are expected to increase the overall percentage of induced deliveries.4^, 5^, 6^, 7 Therefore, a safe, evidence-based modality such as membrane sweeping, which minimizes the need for formal induction, could have strong clinical utility.

The Cochrane review included data from 19 countries, a breadth of resource settings, and urban and rural locations. There was significant procedural heterogeneity across studies in the number of revolutions and standard depth of digital advancement in the membrane sweeping procedure. Sweeps can also be associated with bleeding from undiagnosed placenta previa or a low-lying placenta, which were not reported in this review because they were exclusion criteria.

There are concerns about the safety of membrane sweeping in carriers of group B streptococci. A prospective trial of 542 patients who underwent membrane sweeping demonstrated a nonsignificant difference in all study outcomes between those who were positive for group B streptococci and those who were negative.8

This Cochrane review did not address the possibility of artificial rupture of membranes after membrane sweeping. In one RCT of 300 patients, there was no significant difference in rates of prelabor rupture of membranes when directly comparing those who underwent membrane sweeping with those who did not. In subgroup analyses, patients with cervical dilatation of 1 cm or greater had a relative risk of 1.10 (95% CI, 1.03 to 1.18), with a number needed to harm of 378 for prelabor rupture of membranes.9 Studies comparing membrane sweeping with amniotomy and oxytocin demonstrated no statistically significant differences in outcomes or safety across the study groups. The same was true of one-time membrane sweeping vs. recurrent membrane sweeping.2

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Brian Ford, MD, FAAFP; Andrew Anderson, MD; Carrie Nichols, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

May 3, 2022

Source

Long B, Gottlieb M. Tranexamic acid for aneurysmal subarachnoid hemorrhage. Academic Emergency Medicine. Published online April 5, 2022:acem.14482.

Study Population: 13 trials comprising 2888 patients diagnosed acutely with aneurysmal SAH and awaiting surgical management

Efficacy Endpoints

All-cause mortality, poor functional recovery, risk of rebleeding

Harm Endpoints

Adverse events

Narrative

Subarachnoid hemorrhage (SAH) is a deadly cerebrovascular emergency that accounts for approximately 5% of strokes.1 It most commonly is the result of a ruptured intracranial aneurysm, which releases blood into the subarachnoid space.2^, 3 Rebleeding can occur in up to 20% of SAH cases, with the highest risk in the first 24 h following the initial hemorrhage,4^, 5^, 6^, 7 suggesting interventions that reduce the risk of rebleeding may improve outcomes. Securing the culprit aneurysm is the most effective means, but immediate control may not be possible in all patients.8 Tranexamic acid (TXA) is an antifibrinolytic agent that inhibits the conversion of plasminogen to plasmin and has been studied for a variety of uses.9 While current evidence supports the early repair of aneurysmal SAH,8 there is no clear consensus on administration of TXA if rapid repair is not possible.10^, 11^, 12 The American Heart Association 2012 guidelines state that if there is an unavoidable delay in surgical treatment, risk of rebleeding, and no contraindications, short-term therapy with TXA or aminocaproic acid is reasonable to reduce risk of early rebleeding.13 Despite this recommendation use of TXA remains controversial in those with aneurysmal SAH.10^, 11^, 12

The systematic review summarized here included randomized controlled trials (RCTs) of patients diagnosed with SAH with suspected or confirmed ruptured aneurysm randomized to TXA or placebo.14 The primary outcome used to evaluate the efficacy of TXA included all-cause mortality at the end of follow-up. Causes of mortality included rebleeding, hydrocephalus, cerebral ischemia, extracranial causes, and any complication from an operation or anesthesia. Secondary outcomes included poor functional outcome and rebleeding risk. Poor functional outcomes were evaluated using the modified Rankin Scale (mRS ≥4) or Glasgow Outcome Scale (GOS ≤3). Rebleeding was defined as definite (e.g., imaging) or possible (e.g., sudden deterioration). Safety endpoints included serious adverse events defined as hydrocephalus, cerebral ischemia, or venous thromboembolism (VTE).

The systematic review included 13 RCTs (n = 2888 patients). Of the included patients, 1451 received TXA. Sample size ranged from 39 to 954. Twelve of the trials were conducted in Europe. Five were multicenter RCTs. Three trials were published after 2000, and the remainder were published prior to 2000. Initial dosing was 1 g intravenously in nine trials, followed by maintenance dosing. One study administered TXA orally.15 The duration of therapy for TXA ranged from <72 h to 6 weeks.

TXA had no significant effect on mortality compared with control treatment (risk ratio [RR]: 1.0; 95% confidence interval [CI]: 0.8–1.1; 12 trials, 2426 patients). Analysis based on treatment duration (≤3 days versus >3 days), follow up period, and long-term mortality (>90 days) also demonstrated no subgroup benefit with TXA. TXA had no significant effect on poor functional outcome (RR: 1.0; 95% CI: 0.9–1.2; 5 RCTs, 2491 patients). TXA reduced risk of rebleeding compared with control (RR: 0.6; 95% CI: 0.4–0.8; absolute risk reduction 8.7%; NNT 12; 12 RCTs, 2851 patients). Adverse events did not differ between those receiving TXA versus control, specifically cerebral ischemia (RR: 1.2; 95% CI: 1.0–1.5; 8 RCTs, 2646 patients), hydrocephalus (RR: 1.1; 95% CI: 1.0–1.2; 7 RCTs, 2180 patients), or deep venous thrombosis (RR: 1.2; 95% CI: 0.8–1.8; 7 RCTs, 2151 patients). Cerebral ischemia was the most common side effect and second most common cause of death.

Caveats

This systematic review has several limitations. Importantly, there was clinical heterogeneity among studies, including SAH severity and patient comorbidities. Some studies lacked blinding and allocation concealment, which may introduce bias, and the means of detecting SAH may have varied over time due to differences in sensitivity and availability of different imaging modalities, as 10 studies predated the year 2000. Current management including critical care interventions, intracranial pressure monitoring, and surgical care for patients with SAH has evolved over time and may not reflect the management performed at the time of these studies. Importantly, the meta-analysis was not restricted to only studies in which aneurysm repair was delayed. While several studies included patients in whom surgical treatment was delayed, several did not involve delayed surgery. In the largest and most recent study, treatment was not specifically delayed and occurred on average 14 h after diagnosis.12 The authors also did not separate out aneurysmal versus non-aneurysmal SAH when studies included mixed populations. Of note, it is unclear whether the meta-analysis authors evaluated subgroup data from the included trials. Additionally, there were differences in the administration of TXA by dose, route, and maintenance therapy. The duration of TXA treatment also varied significantly from <72 h to 6 weeks. Although adverse outcomes were not statistically significant between patients receiving TXA compared to placebo, it is possible there was not sufficient power to detect a difference. Given the potential impact of the major adverse events (e.g., cerebral ischemia, hydrocephalus, deep venous thrombosis), TXA should not be considered as having “no risk”. Finally, follow up period varied from 3 weeks to 43 months.

Based on the available evidence, while there was a clear reduction in the secondary outcome of rebleeding with TXA, there was no measurable clinical benefit, suggesting TXA does not benefit patients with SAH awaiting surgery. This seems even clearer in the three largest, most methodologically rigorous trials, which all demonstrated a clear lack of benefit in reducing all-cause mortality and poor functional outcome. Due to the heterogeneity of this evidence and absence of improved patient-oriented outcomes we have assigned a color recommendation of Red (No benefit) for use of TXA in patients with aneurysmal SAH.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 26, 2022

Source

Moullaali TJ, Wang X, Sandset EC, et al. Blood pressure in acute stroke (BASC) investigators. Early lowering of blood pressure after acute intracerebral haemorrhage: a systematic review and meta-analysis of individual patient data. J Neurol Neurosurg Psychiatry. 2022;93(1):6-13.

Study Population: 16 studies with individual participant data from 6221 patients receiving a blood pressure lowering strategy within 7 days of acute spontaneous intracerebral hemorrhage

Efficacy Endpoints

Functional status based on distribution of modified Rankin scale scores, death or dependency, death or severe dependency, death

Harm Endpoints

Early deterioration, symptomatic hypotension, and other serious adverse events

Narrative

Acute spontaneous intracerebral hemorrhage (ICH) is associated with poor clinical outcomes, particularly in the setting of elevated blood pressure.1^, 2^, 3 Some trial data have suggested early aggressive lowering of blood pressure (BP) may improve patient outcomes,4 theoretically by reducing hematoma growth. However, rigorous randomized controlled trials (RCTs) have in many cases failed to show a benefit with a target systolic BP <140 mm Hg compared with a less stringent target.5^, 6 While American Heart Association (AHA) guidelines suggest it is safe to decrease systolic BP to 140 mm Hg during ICH,7 there is uncertainty about the effects of BP lowering in this setting.8^, 9^, 10

The systematic review summarized here included individual participant data (IPD) from trials evaluating BP management after acute ICH.11 In an IPD, the data points from each enrolled subject are collected from the study investigators and are used for performing the meta-analysis. This is different from a trial level data meta-analysis where only the final findings of trials are included in the analysis.

The authors of the review included trials of patients >18 years with acute primary spontaneous ICH <7 days from onset, and which randomized patients to more intensive or less intensive blood pressure lowering. In some cases, this meant drug versus placebo, while in others it meant lower BP targets, such as a systolic BP of 140 mm Hg versus 180 mm Hg. Primary outcomes included functional status defined by the ordinal distribution of modified Rankin Scale (mRS) scores ranging from 0 (no symptoms) to 6 (death) at the end of follow-up. Secondary outcomes included death or dependency (mRS, 3–6), death or severe dependency (mRS, 4–6), and death. Safety outcomes included early neurological deterioration, symptomatic hypotension, and other serious adverse events, with all outcomes defined by their respective trial.

The authors contacted investigators of 50 studies and obtained IPD from 16 (n = 6221 patients). Mean patient age from available studies was 64 years, and 36% of patients were female. Patients had a median National Institutes of Health Stroke Scale score of 11 on presentation. Mean systolic and diastolic BP at randomization were 177 and 100 mm Hg, respectively. Median hematoma volume was 11 ml. Median time to randomization from symptom onset was 4 h. One trial (n = 274) evaluated renin-angiotensin system blockers, three evaluated α- and β-blockers (n = 3310), four evaluated calcium channel blockers (n = 1319), five evaluated nitrates (n = 789), and one evaluated magnesium (n = 387), while the remainder evaluated multi-drug regimens.

Blood pressure was significantly lower in patients receiving more intensive management: mean difference in the systolic BP 8 mm Hg (95% CI: 6–9) at 1 h, 12 mm Hg (95% CI: 11–13) at 1–24 h, and 7 mm Hg (95% CI: 6–8) at 2–7 days. However, intensive interventions were not associated with improved recovery or death. Intensive BP-lowering interventions also had equivocal effects on hematoma growth, with a nonsignificant mean difference of roughly 1 ml at 24 h, while no effect was seen on rates of neurologic deterioration, severe hypotension, and cardiac or renal events.

Caveats

The most obvious weakness in these data is that they represent only half (54%) of patients randomized into the 50 relevant studies identified by the authors. Additionally, only 36% of patients were female, which may reflect underrepresentation. It is unclear how these issues might have introduced bias and impacted the results. However, just as publication bias tends to favor intervention groups, missing data (particularly those based on subjective decisions by investigators about whether to share data) are likely to have a similar effect.

In this sense, it is reassuring that two other systematic reviews arrived at the same conclusion.8^, 12

The systematic review did not report baseline mRS scores. Therefore, it is not clear how baseline imbalances between patients were addressed. The degree of clinical heterogeneity in the 16 included studies is a major limitation, raising concerns about combining data for a meta-analysis. The clinical heterogeneity could be attributed to various settings; differences in subject inclusion/exclusion criteria; and variations in agents, classes, and routes of administration for BP lowering. The meta-analysis also included multiple lower-quality RCTs but did not conduct a subgroup analysis of high-quality studies only.

The latest AHA ICH guideline takes a favorable stand on BP lowering, concluding: “surviving patients show modestly better functional recovery, with a favorable trend seen toward a reduction in the conventional clinical endpoint of death and major disability.”7 This recommendation is primarily based on INTERACT-2, but additional trials including ATACH-2 have failed to find evidence of benefit with aggressive BP lowering.5^, 6 Given the lack of statistical benefit in the data we reviewed (odds ratios at roughly 1.0 indicating equal odds of outcomes in control and study groups), the current data do not support a “favorable trend.” Moreover, bias from studies cited in the guideline7 (e.g., selective data inclusion, unblinded observational designs) favoring BP lowering, suggests that these results may actually overestimate potential benefits and underestimate harms.8^, 10^, 11

In summary, the existing evidence currently offers no support for aggressive attempts to lower BP in acute ICH patients. ICH remains a heterogenous condition, and the absolute difference in BP achieved in this meta-analysis was modest. Because of the heterogeneity of this evidence and its limited ability to represent published studies, we have assigned a color recommendation of yellow (unclear if benefit) for active/intensive BP-lowering interventions in the setting of acute ICH. Further high-quality data are needed.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

March 15, 2022

Source

Echeverría-Esnal D, Martin-Ontiyuelo C, Navarrete-Rouco ME, et al. Azithromycin in the treatment of COVID-19: a review. Expert Rev Anti Infect Ther. 2021;19(2):147-163.

Study Population: 9,049 patients hospitalized with COVID-19 (inpatient setting) and 1,917 patients not hospitalized with COVID-19 (outpatient setting)

Efficacy Endpoints

For patients who were hospitalized: all-cause mortality, clinical status (worsening or improvement), and death within 28 days of diagnosis; for patients not hospitalized: all-cause mortality, hospital admission, death within 28 days of diagnosis, and symptom resolution within 14 days

Harm Endpoints

Any adverse events and serious adverse events

Narrative

COVID-19 is a rapidly spreading disease that has resulted in more than 400 million cases and more than 5.7 million deaths worldwide as of February 10, 2022.1 Although most patients have mild symptoms and do not require hospitalization, some develop severe or critical disease requiring hospital admission, oxygenation, and ventilatory support.2^, 3^, 4^, 5 Vaccination has proven to be effective at reducing transmission, severe illness, and death, but vaccine hesitancy and unequal access to vaccinations pose significant global threats.6^, 7 Several therapies have been proposed for inpatient and outpatient treatment of COVID-19, including antibiotics such as azithromycin (Zithromax).

A Cochrane review including 10 randomized controlled trials (RCTs) evaluated azithromycin vs. standard of care or placebo in 10,966 inpatients with moderate to severe COVID-19 or outpatients with asymptomatic or mild COVID-19.8 Patients with suspected or confirmed COVID-19 (using polymerase chain reaction or antigen testing) were enrolled in these trials. Among inpatients, primary outcomes included all-cause mortality, clinical status (worsening or improvement), death within 28 days, any adverse events, and serious adverse events. Among outpatients, primary outcomes included all-cause mortality, hospitalization, death within 28 days, symptom resolution within 14 days, adverse events, and serious adverse events. Azithromycin dosages included the usual dose (250 mg to 500 mg orally once per day; 500 mg intravenously once per day), low dose (less than 250 mg orally once per day; less than 500 mg intravenously once per day), or high dose (greater than 500 mg orally once per day; greater than 500 mg intravenously once per day). Standard-of-care and placebo arms were considered the control groups. The Cochrane review excluded studies that compared antibiotics with other treatments.

The meta-analysis included seven RCTs with 9,049 inpatients who had moderate to severe COVID-19 and three RCTs with 1,917 ambulatory outpatients who had asymptomatic or mild COVID-19.8 In the outpatient setting, all medications were administered orally. In the inpatient setting, medications were administered orally, intravenously, or by nasogastric tube.

Compared with standard of care alone, azithromycin in the inpatient setting did not affect all-cause mortality within 28 days (high-certainty evidence) or any adverse events (low-certainty evidence). Compared with standard of care or placebo, azithromycin in the inpatient setting did not affect clinical worsening or death within 28 days (moderate-certainty evidence), clinical improvement within 28 days (moderate-certainty evidence), serious adverse events (moderate-certainty evidence), or cardiac arrhythmias (moderate-certainty evidence).

Compared with standard of care or placebo, azithromycin in the outpatient setting did not affect all-cause mortality within 28 days (low-certainty evidence), inpatient admission within 28 days, death within 28 days (low-certainty evidence), or symptom resolution within 14 days (low-certainty evidence). There was no difference in serious adverse events compared with placebo or standard of care (very low-certainty evidence), and no study reported adverse events or cardiac arrhythmias.

Caveats

This meta-analysis had limitations. Most of the patients who were hospitalized with COVID-19 had moderate disease. The sample size for patients with severe disease was small. Several trials included patients with suspected, but not confirmed, COVID-19. Although the evidence for inpatients demonstrated a high certainty of evidence and low risk of bias, only three RCTs of outpatients were included, with a relatively small number of patients having confirmed COVID-19. For total and serious adverse events, there was a high risk of bias and imprecision of results due to the limited number of studies reporting adverse events, and most of the studies were underpowered to adequately address harm outcomes.

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Chaudhuri D, Sasaki K, Karkar A, et al. Corticosteroids in COVID-19 and non-COVID-19 ARDS: a systematic review and meta-analysis. Intensive Care Med. 2021;47(5):521-537

Study Population: 18 trials comprising 2,826 critically ill patients with ARDS (COVID-19 and non-COVID-19)

Efficacy Endpoints

Mortality

Duration of mechanical ventilation

Intensive care unit and hospital length of stay

Harm Endpoints

Gastrointestinal bleeding

Neuromuscular weakness

Nosocomial infection

Narrative

Acute respiratory distress syndrome (ARDS) is a disabling lung condition associated with noncardiogenic pulmonary edema and capillary endothelial injury.1 Commonly accepted definitions require an acute condition with bilateral pulmonary infiltrates, severe hypoxemia, and noncardiogenic pulmonary edema.2^, 3 Up to 10% of patients admitted to an intensive care unit and 23% of mechanically ventilated patients are affected by ARDS, with a mortality rate approaching 45%.1^, 4^, 5 Current guidelines recommend corticosteroids in early, moderate-to-severe ARDS,6^, 7 and recent randomized controlled trials (RCTs) have suggested that corticosteroids may benefit patients with ARDS,8^, 9 including those suffering from COVID.10 The review summarized here, by Chaudhuri et al.,11 included RCTs evaluating corticosteroids for critically ill adults with non-COVID– or COVID-associated ARDS. The authors included studies of ARDS where the condition was strictly defined as well as subgroups of mechanically ventilated subjects with COVID on the assumption such patients are highly likely to be experiencing ARDS. Outcomes included mortality, duration of mechanical ventilation, intensive care unit length of stay, and hospital length of stay. Harm outcomes included opportunistic infections, muscle weakness, and gastrointestinal bleeding. The authors also conducted trial sequential analysis for the primary outcome of mortality.

The review identified 18 relevant RCTs (n = 2,826).11 Of these, 10 utilized strict ARDS criteria (eight non-COVID and two COVID). Eight trials used methylprednisolone, six hydrocortisone, and four dexamethasone. Two started corticosteroids >7 days after the diagnosis, whereas the other 16 initiated corticosteroids during the first 7 days. Eight trials administered ≤7 days of therapy, while 10 provided ≥10 days. Ten RCTs were double-blinded and utilized a placebo control, while eight were nonblinded and used standard care as a control group.

Overall, corticosteroids were associated with statistically lower mortality in adult patients with presumed ARDS (relative risk [RR] = 0.8; 95% confidence interval [CI] = 0.7 to 0.95, absolute risk reduction [ARR] = 8.8%, NNT = 12, n = 2,740), reduced length of mechanical ventilation (mean difference 4 days fewer; 95% CI = 2.5 to 5.5 days), and reduced hospital length of stay (8 days fewer; 95% CI = 3.1–13 days). In subgroup analysis of studies using a strict definition of ARDS corticosteroids were associated with lower mortality in non-COVID illnesses (RR = 0.7, 95% CI = 0.5 to 0.9, ARR = 12.6%, NNT = 8, n = 1,317). Patients receiving corticosteroids for >7 days also demonstrated lower mortality (RR = 0.7, 95% CI = 0.6 to 0.9, ARD = 11.2%, NNT = 9, n = 2,044). No statistical difference was seen, however, in the subgroup of patients with COVID (RR = 0.9, 95% CI = 0.8 to 1.1). No clinically relevant harms were noted.

The authors’ trial sequential analyses found data are inadequate to establish a benefit concerning mortality with corticosteroids. They calculate 4,690 subjects would be required for the difference found here to be considered significant (the meta-analysis included 2,740 subjects).11^, 12

Caveats

Beyond inadequate power there are important limitations to any findings suggested by these data. First, while this analysis included 18 RCTs, only 10 used a strict definition for ARDS, although findings were similar for those that did and did not. Second, the etiology of ARDS in the included studies varied significantly, including trauma, sepsis, pneumonia, aspiration, blast injury, post-operation, multiple transfusions, and drug reactions. While the statistical effect on mortality seemed somewhat consistent across trials, these are major sources of clinical heterogeneity. Moreover, corticosteroid types, doses, timing of initiation, and duration of therapy varied substantially. Unfortunately, while the authors performed many subgroup and sensitivity analyses in an effort to address these issues, the overwhelming heterogeneity in these data means (1) different disease states are being analyzed together and (2) determining an ideal agent, dose, timing, or duration is not possible. It should also be noted just one of eight COVID studies and two of eight non-COVID studies found a benefit with corticosteroids, while 13 were negative.

Finally, major sources of bias affect a preponderance of included trials. Perhaps most importantly only half used a placebo and doubleblinding, and two more were rated “high risk of bias” for other reasons.

Based on the presented evidence, we have assigned a color recommendation of yellow (data inadequate) for use of corticosteroids in critically ill patients with ARDS. Further data preferably from larger, more rigorous trials, performed in well-defined populations, are needed.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

January 18, 2022

Source

Wagner C, Griesel M, Mikolajewska A, et al. Systemic corticosteroids for the treatment of COVID-19. Cochrane Database Syst Rev. 2021;2021(8):CD014963.

Study Population: 8,075 participants in 11 randomized controlled trials that evaluated systemic corticosteroids for people with COVID-19, irrespective of disease severity

Efficacy Endpoints

All-cause mortality (at longest follow-up available); ventilator-free days; new need for invasive ventilation

Harm Endpoints

Adverse events; hospital-acquired infections

Narrative

The novel coronavirus-19 emerged in late 2019 and has resulted in a global pandemic with over 5.2 million deaths to date.1 A systemic inflammatory response with excessive release of cytokines and other inflammatory mediators can lead to lung injury and low oxygen levels (hypoxia).2 The RECOVERY trial, the world's largest clinical trial evaluating treatments of COVID-19, was conducted in 185 sites in the United Kingdom. It was the first trial to suggest that systemic corticosteroids, such as dexamethasone, may decrease the risk of death and decrease the progression of disease in nonventilated patients (from ventilator-free to requiring invasive mechanical ventilation).3 Since publication of this trial, several other trials have explored the use of corticosteroids in COVID-19.

The Cochrane systematic review and meta-analysis discussed here evaluates the effect of systemic corticosteroids in COVID-19 patients.4 It included 11 randomized controlled trials (RCTs) comprising 8,075 participants diagnosed with COVID-19. Of these participants, 3,072 were randomized to receive systemic corticosteroid plus standard care and the remainder received standard care alone. A majority of patients receiving corticosteroids were treated with dexamethasone (n = 2,322). The outcomes of interest included all-cause mortality, ventilator-free days, a new need for invasive mechanical ventilation, serious adverse events, all adverse events, and hospital-acquired infections.

The meta-analysis found that all-cause mortality was decreased in the systemic corticosteroids plus standard care arm compared to standard care alone (relative risk [RR] = 0.89, 95% confidence interval [CI] = 0.8 to 1.00, absolute risk difference [ARD] = 3%, NNT = 33, nine RCTs, n = 7,930, median follow-up = 28 days, quality of evidence = moderate).

Only one RCT showed that the number of ventilator-free days was larger in corticosteroid arms (mean difference = 2.6, 95% CI = 0.67 to 3.43, quality of evidence = low). There was not enough evidence to determine whether corticosteroids affected the need for invasive mechanical ventilation, liberation from mechanical ventilation, quality-of-life measures, viral clearance, and need for dialysis.

Only two of the included RCTs measured serious adverse events and only five RCTs reported the risk of hospital-acquired infections. Because of the high risk of bias, heterogeneous definitions, and underreporting, the authors of the meta-analysis were uncertain about the size and direction of the effect when assessing harm outcomes.

Caveats

Given that the upper limit of the 95% CI for the reported relative risk of death (all-cause mortality) reaches 1 (possibility of equal risk in treatment and placebo groups), the survival benefit form systemic corticosteroids is not clearly established. This explains why we labeled this benefit “probable.” Future high-quality trials could swing the pendulum toward more robust evidence of benefit or lack thereof.

There is no universal definition of “standard therapy” for the treatment of COVID-19, so patients in the different studies received a variety of other drugs such as antibiotics, hydroxychloroquine, convalescent plasma, and other treatments that might have confounded the results.

A subgroup analysis in the RECOVERY trial suggested that dexamethasone, when given to nonventilated patients, reduced the progression to invasive mechanical ventilation. However, the Cochrane meta-analysis did not reach the same conclusion. Moreover, because none of the trials enrolled asymptomatic, mildly ill patients, or those who were not hospitalized, there was not enough evidence to conclude that corticosteroids decreased the progression of mild or moderate COVID-19 disease to severe disease.

Corticosteroids have potential side effects such as immunosuppression, increased risk of infection, hyperglycemia, increased risk of gastrointestinal hemorrhage, and impaired wound healing.5 However, most of the adverse events from corticosteroids are associated with chronic use. Regardless, most of the included trials either did not measure or did not report adverse events. It is thus not possible to comment on the magnitude of harm associated with this treatment in COVID-19 patients.

A subgroup analysis from the RECOVERY trial suggested that mortality was higher in the subgroup of patients not on oxygen who received dexamethasone. This difference was not statistically significant. Similarly, the Cochrane meta-analysis also performed a subgroup analysis that stratified patients based on the need for respiratory support at randomization. This subgroup analysis also showed a negative effect on mortality from administration of dexamethasone in the participants who did not require respiratory support. The difference was also not statistically significant. The findings of such subgroup analyses simply generate hypothesis for future trials and drawing conclusions from them is not appropriate.

Over 80% of the study participants were from high-income countries and there were no studies from low-income countries, which may limit the generalizability of these results. Lower-income countries may have limited hospital resources, such as hospital beds, supplies of medications and oxygen, equipment for respiratory support, and medical personnel, and therefore may not experience the same benefits of systemic corticosteroid treatment delivered in well-resourced hospital settings.

Since most of the studies included in the meta-analysis used dexamethasone, the benefits may not be generalizable to all systemic corticosteroids. Only one study compared different types of systemic corticosteroids (methylprednisolone vs. dexamethasone). There was no analysis comparing different doses (range = 150 to 5,000 mg daily hydrocortisone equivalents) or duration of corticosteroids (range = 0 to 20 days). It is thus unclear whether other systemic corticosteroids have the same impact on mortality as dexamethasone or if there is an ideal dose or duration of treatment. Dozens of ongoing and recently completed trials will likely generate additional evidence to shed light on these issues as well as additional data on safety, long-term survival, and other patient-centered outcomes.

All the trials included in the Cochrane systematic review enrolled patients from the beginning of the pandemic and prior to the widespread availability of vaccinations and the emergence of variants such as the delta variant, which has been shown to have increased transmissibility and decreased host immunity.6 It is thus unclear whether systemic corticosteroids demonstrate the same benefits in vaccinated patients, in those who are infected with different variants, or in those with breakthrough infections despite vaccination or prior infection.

Some uncertainty remains in regard to the survival benefit of systemic corticosteroids in COVID patients. In addition, while harms are unlikely from a short course of corticosteroids in patients with severe COVID (the current common practice), we are concerned about the lack of proper tracking and reporting of the harms. Therefore, we have assigned a color recommendation of yellow (unclear if benefits) for systemic corticosteroids in hospitalized COVID-19 patients.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Kenneth Lu, MD; Richard W. Leno, MD; Selwena R. Brewster, MD
Supervising Editors: Kabir Yadav, MD

Published/Updated

December 28, 2021

Source

McKenzie BJ, Wechalekar MD, Johnston RV, Schlesinger N, Buchbinder R. Colchicine for acute gout. Cochrane Musculoskeletal Group, ed. Cochrane Database of Systematic Reviews. 2021;2021(8).


Study Population: 803 patients with acute gout in four trials

Efficacy Endpoints

Global assessment of treatment success (self-reported pain control ≥50%)

Harm Endpoints

Serious adverse events and total adverse events

Narrative

Gout is a common inflammatory condition of the joints. Episodes often begin rapidly and have a duration of days to weeks. The prevalence increases with age, occurring in up to 7% of men over the age of 65 years.1 Gout occurs due to an increase in the serum uric acid, with deposition in the joints eventually leading to inflammation that may present with pain, swelling, and redness (erythema) of the affected joint.2 Gout can be precipitated by toxins (e.g., alcohol), cellular breakdown (e.g., chemotherapy) or joint trauma.2 Although gout is self-resolving, medications can expedite resolution, reduce pain, and help patients return to their normal daily activities. Gout is commonly treated with anti-inflammatory medications, which can include non-steroidal anti-inflammatory drugs (NSAIDs), colchicine, or glucocorticoids. The American College of Rheumatology guidelines recommend that NSAIDs, oral colchicine, or glucocorticoids may all be considered for the treatment of an acute gout flare.3

The systematic review and meta-analysis summarized here identified four trials 4^, 5^, 6^, 7 (n = 803 participants) comparing colchicine with placebo, NSAIDs, or higher doses of colchicine.8 The authors included randomized and quasi-randomized controlled trials of adults (age ≥18 years) with a diagnosis of acute gout. The primary outcome for the review was pain control (defined as the proportion of patients reporting ≥ 50% decrease in pain within the first 32–36h of presentation).8 Total adverse events and serious adverse events were also reported.

Lower-dose colchicine (1.8 mg over 1h) improved pain control compared to placebo (one study, 103 participants, relative risk [RR] = 2.43, 95% confidence interval [CI] = 1.05 to 5.64, absolute risk difference [ARD] = 24.7%, number needed to treat [NNT] = 5, quality of evidence = low) with no difference in adverse events between groups.4 The quality of evidence was downgraded one level due to the risk of selection bias from unclear reporting of the method of randomization and risk of reporting bias. The level of evidence was downgraded a second level for imprecision as the number of events was small (<200) and the 95% CI included a low value with no appreciable benefit.

Higher-dose colchicine (4.8 mg over 6h or 1 mg followed by 0.5 mg every 2h until complete response or adverse events) also improved pain control compared to placebo (two studies, 124 participants, RR = 2.2, 95% CI = 1.3 to 3.7, ARD = 23%, NNT = 4, quality of evidence = low) and increased adverse events (84% in the higher-dose colchicine group vs. 26% in the control group; two studies, 124 participants, RR = 3.2, 95% CI = 2.0 to 5.1, ARD = 58%, number needed to harm [NNH] = 2, quality of evidence = low).4^, 5 The quality of evidence was downgraded one level due to the risk of selection bias from unclear reporting of the method of randomization and risk of reporting bias in one study. The level of evidence was downgraded a second level for imprecision because the total number of participants was small and the total number of events was small. Adverse events primarily comprised nausea, vomiting, diarrhea, and abdominal pain. There were no serious adverse events reported in either group.

Compared with lower-dose colchicine (1.8 mg over 1h), higher-dose colchicine (4.8 mg over 6h) did not improve pain control. However, higher-dose colchicine was associated with an increased risk of adverse events (one study, 126 participants, RR = 2.1, 95% CI = 1.5 to 3.0, ARD= 40%, NNH = 3, quality of evidence = low).4 The quality of evidence was downgraded one level due to the risk of selection bias from unclear reporting of the method of randomization and risk of reporting bias. The level of evidence was downgraded a second level for imprecision because the number of events was small (<200) and the 95% CI included a low value with no appreciable benefit. Adverse events primarily comprised nausea, vomiting, diarrhea, and abdominal pain. There were no serious adverse events reported in either group.

Lower-dose colchicine (0.5 mg taken three times per day for 4 days) did not improve pain compared with NSAIDs (naproxen 750 mg at baseline followed by 250 mg every 8h for 7 days) and there was no difference in adverse events (one study, 399 patients).6

Caveats

There are several important limitations to the data reported in this systematic review. First, there were differences in the diagnostic criteria used for gout, with some using joint aspirate while others used the American College of Radiology criteria. There was heterogeneity in doses and timing of the medication among studies. The primary outcome for which the individual studies were powered varied slightly with two mirroring the primary outcome for the review,4^, 5 while the other primary outcomes included worst pain in the first 24h6 and recurrence rate within 3 months.7 Additionally, trials used different scales for pain and different time periods for the assessment of outcomes. Only one NSAID (naproxen) was evaluated, so it is unclear if other NSAIDs would be more effective. Moreover, the sample sizes were small, the trials were not powered for adverse events, and comparisons were from only one or two studies, meaning none of these results can be considered broadly reliable. One of the studies did not blind the participants, personnel, or outcome assessors, which may have introduced bias.6 Another study did not report the pain scores despite the methods stating that these data were collected, raising concern for selective reporting bias.4 Further, colchicine and its metabolites are renally and hepatically cleared, but the included studies did not provide data on participant renal or hepatic function. This is important to note because colchicine has a much higher risk of complications (e.g., rhabdomyolysis, seizures, disseminated intravascular coagulation) in the case of overdose or inadequate excretion than NSAIDs. Finally, despite the common use of corticosteroids in gout, there were no RCTs comparing the effectiveness and safety between these two interventions, limiting the translation of this evidence into practice. Similarly, multimodal therapy (e.g., NSAIDs with colchicine vs. colchicine alone) was also not studied.

Colchicine was associated with better pain control and more adverse events than placebo for acute gouty arthritis in small studies. Given the limitations of these studies, we have assigned a color recommendation of yellow (unclear if benefits) to this intervention particularly in light of the lack of benefit over common NSAIDs. Future studies should be performed with larger sample sizes that factor in patient comorbidities.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Michael Gottlieb, MD; Willeed Rabah, MD; Brit Long, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

December 6, 2021

Source

Ortega-Paz L, Galli M, Capodanno D, et al. Safety and efficacy of different prophylactic anticoagulation dosing regimens in critically and non-critically ill patients with COVID-19: A systematic review and meta-analysis of randomized controlled trials. Eur Heart J Cardiovasc Pharmacother. Published online September 14, 2021:pvab070.

Study Population: 7 trials of 5,145 hospitalized patients with COVID-19

Efficacy Endpoints

All-cause mortality, venous thromboembolism, stroke, myocardial infarction, systemic arterial embolism

Harm Endpoints

Major bleeding, any bleeding

Narrative

Coronavirus disease 2019 (COVID-19) is a global pandemic, which has resulted in over 230 million cases and 4.7 million deaths as of September 25, 2021.1^, 2 Literature suggests those with moderate or severe disease have increased endothelial activation and inflammation, coagulopathy, and elevated D-dimer levels, which may increase thromboembolic events.2^, 3^, 4 Breakthrough thromboembolic events in hospitalized COVID-19 patients receiving preventive anticoagulation and observations of heparin resistance have raised the question of whether higher-dose anticoagulation may be beneficial.5^, 6^, 7

The systematic review summarized here included randomized controlled trials (RCTs) comparing higher-dose versus standard-dose preventive anticoagulation in hospitalized patients with COVID-19.8 Authors included all types of anticoagulants. Dosing was defined by the individual trials. Authors pooled therapeutic and intermediate dosing regimens into the escalated-dose group. Primary outcomes included all-cause death at the longest follow-up available and major bleeding. Secondary outcomes included venous thromboembolism (VTE), myocardial infarction (MI), stroke, systemic arterial embolism, any bleeding, and minor bleeding.

The meta-analysis identified seven RCTs of 5,154 hospitalized patients with COVID-19.8 Six RCTs used unfractionated heparin and low molecular weight heparin, with one study using rivaroxaban.9 There were 1,893 critically ill patients and 3,261 non-critically ill patients, with follow up ranging from 14 to 90 days. Authors also performed a pre-planned subgroup analysis comparing critically ill and non-critically ill patients.

All-cause mortality did not differ between groups (17.8% versus 18.6%), but higher-dose anticoagulation increased major bleeding (2.4% vs 1.4%; risk ratio [RR] 1.7; 95% confidence interval [CI]: 1.2–2.6; absolute risk increase [ARI] 1%, number needed to harm [NNH] 101), and bleeding overall (RR: 2.0; 95% CI: 1.1–3.7; ARI 5.3%, NNH 18). Higher-dose anticoagulation was associated with less VTE (2.5% versus 4.7%; RR: 0.6; 95% CI: 0.4–0.7; absolute risk reduction 2.2%, number needed to treat [NNT] 45), but did not reduce MI, stroke, or arterial embolism. Results for subgroup analyses were consistent with overall results except for increased bleeding with higher-dose anticoagulation in non-critically ill patients that was not present in the critically ill patients.

Caveats

COVID-19 is a complex disease, and patients at the beginning of the disease may be prothrombotic, while in later or more severe forms they can develop an increasing bleeding risk.10 Thus, timing of anticoagulation may be crucial. It is possible at earlier stages of disease a higher dose of anticoagulation may be beneficial, while in later stages, this may be harmful. In the present meta-analysis, median time from symptom onset to randomization was approximately 10 days, suggesting future studies should assess the impact of timing.

There are several other important limitations. There was significant heterogeneity for all-cause death; major bleeding and VTE occurred more often in critically ill patients; and some subgroups were too small for useful analysis. Importantly, each trial included in this meta-analysis individually defined major bleeding, resulting in significant heterogeneity concerning this outcome. Duration of follow-up also varied, ranging from 14 days to 90 days. Only the ACTION trial used a direct-acting oral anticoagulant (DOAC) as anticoagulation,9 limiting the ability to assess these agents. Moreover, there was heterogeneity in dosing with only two trials using an intermediate dose of anticoagulation rather than full dose.11^, 12

Based on the evidence, we have assigned a color recommendation of Yellow (Unclear if benefit) for higher-dose prophylactic anticoagulation compared to standard-dose in hospitalized patients with COVID-19. The lack of mortality benefit, increase in bleeding, and reduction in VTE suggest a complicated array of effects requiring larger, more rigorous trials and careful subgroup assessments. There are over 30 RCTs currently enrolling patients to evaluate the role of anticoagulation in patients with COVID-19, and we await further data assessing timing, specific patient populations (e.g., elderly, ventilated, pediatric), dosing, and agent.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

October 26, 2021

Source

Mastoris I, Kokkinidis DG, Bikakis I, et al. Catheter-directed thrombolysis vs. anticoagulation for the prevention and treatment of post-thrombotic syndrome in deep vein thrombosis: an updated systematic review and meta-analysis of randomized trials. Phlebology. 2019; 34(10): 675-682.

Study Population: 4 trials of 1005 patients with deep vein thrombosis

Efficacy Endpoints

Postthrombotic syndrome, patent iliofemoral veins, future venous thromboembolism

Harm Endpoints

Bleeding, all-cause mortality

Narrative

Clot formation in deep veins (deep vein thrombosis [DVT]), such as large veins in the lower extremities and pelvis, is common with an annual incidence approximating 900,000 in the United States.1^, 2 Some patients with DVT (20%–50%) develop long-term complications such as postthrombotic syndrome (PTS).1^, 2^, 3 PTS is manifested by signs and symptoms including swelling of the affected limb, pain, cramps, burning or prickling sensations (paresthesias), itching, redness, skin color change, and varicose vein formation.4 The severe form of PTS can cause skin ulcers from venous stasis.4 These chronic symptoms can reduce patient mobility, affect quality of life, and increase health care costs, particularly with moderate and severe disease. The severity of PTS is determined by grading systems such as the Villalta score.1^, 2^, 3^, 5 This scoring system is composed of signs and symptoms of PTS, with scores of 5^, 6^, 7^, 8^, 9 defined as mild and scores of ≥10 defined as moderate to severe.5 Drugs that prevent clot formation (anticoagulants) can treat DVT but do not prevent PTS in all patients, and residual clots and venous valvular incompetence due to the dilatation of the veins can increase the risk of PTS and recurrent VTE.6^, 7^, 8^, 9 Dissolving the existing clot (thrombolysis) may reduce the risk of PTS but may also increase the risk of bleeding complications.10 To reduce the risk of bleeding, a clot-dissolving medication can be applied directly next to the clot, a strategy called catheter-directed thrombolysis (CDT).11^, 12

The systematic review and meta-analysis summarized here included randomized controlled trials (RCTs) comparing CDT plus anticoagulation versus anticoagulation alone for adult patients with DVT.13 Prespecified outcomes included overall occurrence of PTS, iliofemoral vein patency, overall rates of recurrent VTE, bleeding rates, and all-cause mortality. The authors also preplanned a subgroup analysis based on severity of PTS (mild vs moderate to severe). The severity of PTS was determined by the Villalta score.5

The meta-analysis identified four RCTs comprising 1005 patients with DVT in aggregate. The trials were published between 2002 and 2017.13 Of the included patients, 491 underwent CDT plus anticoagulation and 514 were treated with anticoagulation only. Mean patient age was 53 years, and women accounted for 38% of patients. CDT devices varied. Three trials used alteplase for clot lysis via CDT and one used streptokinase. The anticoagulation regimens included unfractionated heparin or low-molecular-weight heparin with warfarin, although the largest trial (ATTRACT) utilized direct oral anticoagulants (DOACs).12

CDT reduced the risk of overall PTS (odds ratio [OR] = 0.32, 95% confidence interval [CI] = 0.12 to 0.85, absolute risk difference [ARD] = 9.9%, number needed to treat [NNT] = 10). When only data for patients with moderate-to-severe PTS were analyzed, CDT did not offer any preventive benefit and also did not reduce the risk of future VTE. However, patients undergoing CDT were more likely to have patent iliofemoral veins (OR = 2.69, 95% CI = 1.07 to 6.75, ARD = 20.8%, NNT = 5). The rates of death or bleeding were not different between the groups in the overall study population.

The largest trial included in the systematic review (ATTRACT), composed of 692 patients (365 with mild PTS and 327 with moderate-to-severe PTS), showed no difference in PTS or recurrent VTE between the CDT and standard anticoagulation groups.12 However, this trial reported a greater bleeding risk in the CDT group compared to the anticoagulation-only group (1.7% vs 0.3%).12 Bleeding sites included gastrointestinal (two patients), retroperitoneal (two patients), and the device access site (two patients).

A recent Cochrane review published in 2021 evaluated systemic, locoregional, CDT, and pharmacomechanical thrombolysis for treatment of DVT.14 While the Cochrane review found that thrombolysis in general reduced the risk of PTS (risk ratio [RR] = 0.78, 95% CI = 0.66 to 0.93, ARD = 3.2%, NNT = 32, 6 months’ to 5 years’ follow-up, six trials), no reduction in PTS was detected in the CDT subgroup. Combined data from both systemic and local thrombolysis with CDT also suggested that this intervention was associated with increased risk of bleeding compared to standard anticoagulation (RR = 2.45,95% CI = 1.58 to 3.78, ARD = 4.4, number needed to harm = 22). This Cochrane review included only two of the studies from the systematic review by Mastoris et al. which we have summarized here.12^, 13^, 14^, 15

Caveats

There was clinical heterogeneity with regard to devices, thrombolytic agents, and anticoagulation regimens among the different studies. Time to CDT may be an important variable in preventing PTS, as recently formed clots may respond better to thrombolysis.16 Thus, some have recommended that thrombolysis optimally should be performed within 10 days of the onset.16 However, included trials varied concerning time to treatment initiation (within 10 days in ATTRACT trial and 21 days in CaVenT).12^, 15 Duration of follow-up also varied, ranging from 6 months to 5 years. Importantly, the ATTRACT trial utilized DOACs, compared to the other included trials that utilized unfractionated or low-molecular-weight heparin and warfarin.12 The ATTRACT trial also accounted for over half of the included patients, and it did not find a benefit for using CDT to reduce PTS in the overall population.12 Interestingly, this trial found no difference in thrombus removal success rate with CDT compared with anticoagulation alone. Therefore, it is possible the lack of benefit for preventing PTS may have been due to issues with the CDT approach and modality.12 Additional limitations for the meta-analysis include variations in PTS classification, which may have further influenced the outcomes and subgroup analyses. Of note, none of the included trials separated upper versus lower DVT and PTS, and this distinction may impact therapeutic approach and patient morbidity. Finally, the majority of patients analyzed in the systematic review suffered from mild PTS, and the overall number of patients with moderate-to-severe PTS was small. Therefore, the absence of benefit from CDT might be due to a type II error.

Based on the existing evidence, we have assigned a color recommendation of yellow (data Inadequate) for use of CDT to reduce risk of PTS in patients with DVT. This was based on several factors. Overall, CDT was associated with reduced risk of PTS, but there was a lack of benefit for moderate-to-severe PTS in the meta-analysis by Mastoris et al.13 Moreover, data were conflicting in the 2021 Cochrane review, which demonstrated reduced risk of PTS but increased risk of bleeding. The potential benefits and harms of CDT likely vary based on several factors, such as severity of DVT (patient symptoms, comorbidities, venous system involvement), patient activity levels, duration of DVT, and risk of bleeding. Further data from larger trials are needed to assess available devices, more current anticoagulation strategies with DOACs, and patient with various DVT severities.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

October 12, 2021

Source

Ansems K, Grundeis F, Dahms K, et al. Remdesivir for the treatment of COVID-19. Cochrane Database Syst Rev. 2021;8:CD014962. Published 2021 Aug 5. doi:10.1002/14651858.CD014962.

Study Population: 7,452 adult patients with SARS-CoV-2 (hospitalized, moderately to severely ill) enrolled from 5 RCTs comparing remdesivir use to placebo or standard of care

Efficacy Endpoints

All-cause mortality up to day 28 after treatment, need for respiratory support (high-flow oxygen, non-invasive or invasive ventilation), duration to liberation from invasive mechanical ventilation at up to day 28, quality of life, duration of hospitalization, Intensive Care Unit (ICU) admission, and ICU stay duration

Harm Endpoints

Any adverse events up to day 28 after treatment

Narrative

The COVID-19 pandemic has, to date, infected over 230 million people and caused over 4.7 million deaths worldwide.1 There has been great urgency to identify effective treatments, including antiviral medications, to decrease morbidity and mortality. Remdesivir is an antiviral drug that has been hypothesized to inhibit viral replication of RNA viruses2 such as SARS-CoV-2.

The systematic review by Ansems et al3 that we summarize here identified five randomized controlled trials including 7,452 participants (mean age 59 years; hospitalized with moderate to severe illness, most requiring low-flow oxygen or mechanical ventilation at baseline; conducted in high or upper-middle income countries). Pooled data across the five studies found moderate certainty4 evidence that remdesivir treatment of COVID-19 patients resulted in no statistically significant difference in all-cause mortality at up to 28 days after treatment when compared to placebo or standard of care.

Remdesivir did not reduce the need for overall respiratory support (high-flow oxygen, invasive ventilation, or non-invasive ventilation) and had no effect on the duration of invasive mechanical ventilation. However, the certainty of evidence was low to very low. Although remdesivir reduced the need for new invasive mechanical ventilation within 28 days when compared to placebo or standard care (relative risk: 0.56, 95% CI 0.41 to 0.77; Absolute risk difference: 6.7%; number-needed-to treat: 15), the certainty of evidence was low. Therefore, we did not report this finding in the summary table. In general, the certainty of the data for the outcomes of need for any specific type of respiratory support and duration of invasive mechanical ventilation was low or very low. No studies measured quality of life, need for ICU admission, or duration of ICU stay.

The safety data were inconsistently reported in the original trials. Appropriately, the systematic review reported that the data on whether remdesivir increases or decreases adverse events was inconclusive, being comprised of very low-certainty evidence.

Caveats

The existing evidence has significant limitations. Ansems et al downgraded the certainty of evidence for the outcome of mortality from high to moderate due to concern for the failure of randomization in one of the included studies.5 For the outcomes of clinical improvement and adverse events, the authors downgraded the level of certainty due to limitations in the reported data. They cited missing data regarding the competing risks of death. For example, duration of mechanical ventilation would be reduced if a patient died. Another limitation was open-label design that was employed by one of the as included trials.6

Included RCTs used different scales for determining the disease severity and progression, as there are no standardized guidelines for determining 'moderate' and 'severe' severity of disease related to SARS-CoV-2. The need for respiratory support, and specifically, the need for each method of respiratory support (high-flow oxygen, non-invasive ventilation, and invasive ventilation) is also determined by clinical judgment and subject to significant variability.

Despite the significant limitations in regards to the study population, relative homogeneity of institutions, and risk of bias, the Cochrane review by Ansems et al3 is the largest and most recent review that included only randomized controlled trials comparing remdesivir to placebo or standard of care.

In summary, we assign a color recommendation of Yellow (unclear if benefits/more data needed) for the use of remdesivir in patients hospitalized for SARS-CoV-2. Additional data are needed to clarify the efficacy and safety of remdesivir for SARS-CoV-2 treatment. Further studies are also needed to address the efficacy and safety in different population subgroups (e.g. age, disease severity, timing of treatment administration, etc.) and socioeconomic settings.

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Robert Allen, MD; Matthew Turner, MD; Ian S. deSouza, MD
Supervising Editors: Shahriar Zehtabchi, MD; Dan Runde, MD

Published/Updated

Source

Wikkelsø A, Wetterslev J, Møller AM, Afshari A. Thromboelastography (Teg) or thromboelastometry (Rotem) to monitor haemostatic treatment versus usual care in adults or children with bleeding. Cochrane Database Syst Rev. 2016 Aug 22;2016(8):CD007871.

Study Population: 1493 total elective surgery patients included those with bleeding from cardiac surgery (1435 patients), excision of burn wounds (30 patients) and liver transplantation (28 patients)

Efficacy Endpoints

All-cause mortality, requiring transfusion (PRBC, FFP, platelets), surgical re-intervention for bleeding control

Harm Endpoints

Any adverse events

Narrative

Severe bleeding and coagulopathy contribute to high morbidity and mortality in patients undergoing surgery. While conventional coagulation testing is commonly used to guide transfusion, these assays do not reveal the complexities of hemostatic derangements and were not designed to predict bleeding or guide coagulation management. Point of care thromboelastography (TEG) and rotational thromboelastometry (ROTEM) assays measure fibrinolysis and were developed partly for the purpose of guiding hemostatic resuscitation.1 TEG/ROTEM are currently used in many trauma centers to guide acute resuscitation, within minutes of a patient's arrival to the emergency department (ED). We examined the evidence pertaining to the benefit and safety of these assays for ED patients with acute hemorrhage.

The systematic review and meta-analysis2 included 17 randomized controlled trials with 1493 subjects, most of whom were adults undergoing elective cardiac surgery. The sample size of the 16 trials ranged from 28 to 224 per trial (median of 96 subjects per trial). The trials compared TEG- or ROTEM-guided transfusion strategies to care guided by clinical judgment or conventional testing. The systematic review’s primary outcome was overall mortality; secondary outcomes included the proportion of patients requiring transfusion of packed red blood cells (PRBC), platelets, and plasma, the amount of blood products transfused, rate of surgical re-intervention, hospital length of stay, and complications from bleeding or transfusion.2

All-cause mortality was lower with TEG/ROTEM (7.4% control vs 3.9% with Accepted Article TEG/ROTEM, Relative Risk [RR]: 0.5; 95%CI, CI 0.3 to 0.9; Absolute Risk Difference [ARD]: 3.6%; Number-needed-to-treat [NNT]: 28; low quality evidence, 8 trials, n=717). The proportion of patients receiving transfusion of packed red blood cells (PRBC) was lower in patients assigned to TEG/ROTEM (RR 0.9; CI 0.8 to 0.9; ARD: 10%; NNT: 10; low quality evidence, 10 trials, n=832). Similarly, the need for plasma (RR 0.6; CI 0.3 to 0.9; ARD: 20%; NNT: 5; low quality evidence, 8 trials, n=761), and platelets (RR 0.7; CI 0.6 to 0.9; ARD: 9%; NNT: 11; low quality evidence, 10 trials, n=832) was also lower. The systematic review found no difference in rate of surgical re-intervention to control bleeding. Some trials reported the amount of blood products used in each group as a continuous variable. However, because of the variability of the blood product measures (some trials reported them in units, others in milliliters, or milliliters/kg) and methods of summarizing data (medians and quartiles versus means and standard deviation), as well as the significant heterogeneity between these trials, the systematic review did not summarize these data.2

Adverse outcomes and complications were inconsistently reported, thereby making estimates unreliable.

Caveats

Given the many limitations of the data, the authors of this systematic review offer guarded support for their findings and rate the quality of evidence as low. In addition, due to the wide array of transfusion algorithms and varied thresholds for administering blood products in each trial, there was significant clinical and statistical heterogeneity among the trials.2 Furthermore, no trial used a 1:1:1 hemorrhage transfusion strategy of red cells, plasma, and platelets; an approach currently favored by many experts.11^, 12

Mortality data was missing or incompletely reported in many trials. In addition, statistical significance was lost for mortality when comparing fixed-effect and random-effect models, thereby suggesting the results were unreliable. Therefore, we have not reported mortality outcomes in the summary table.

Methodological weaknesses were rife throughout the included trials. Only two of the 17 trials (12%) were appropriately blinded and had low risk of bias. Additionally, the dataset was too heterogenous and biased to allow for meaningful appraisal of most outcomes.

Lastly, 91% of study participants were adults undergoing elective cardiac surgery. Data from these participants are not generalizable to other clinical settings of hemorrhage management, such as the ED or intensive care units. It may be premature to extrapolate the findings to acute trauma resuscitations, or spontaneously bleeding medical patients, until trials are conducted in these specific populations. Severely injured patients or those with gastrointestinal bleeding, for example, may arrive in the ED already coagulopathic, with ongoing blood loss that continues until source control of bleeding is accomplished. In the two randomized trials comparing TEG/ROTEM-guided strategies to conventional management in acute care environments, no meaningful benefits were seen.9^, 10

In summary, TEG/ROTEM-guided hemostatic resuscitation requires further study. While early, low quality data from elective surgical settings suggests a possible benefit, more high-grade evidence from ED-relevant resuscitations are needed. Therefore, we have assigned a color recommendation of Yellow (Unclear if benefits, more data needed).

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Suvam Neupane, MD; David Warshaw, MD; Elias Youssef, MD, MBA; Bonny J. Baron, MD
Supervising Editor: Kabir Yadav, MD

Published/Updated

August 19, 2021

Source

Lewis K, Piticaru J, Chaudhuri D, Basmaji J, Fan E, Møller MH, et al. Safety and Efficacy of Dexmedetomidine in Acutely Ill Adults Requiring Noninvasive Ventilation: A Systematic Review and Meta-analysis of Randomized Trials. Chest. 2021 Jun;159(6):2274-2288.

Study Population: 738 critically ill patients requiring noninvasive ventilation

Efficacy Endpoints

Need for intubation, delirium, intensive care unit length of stay, pneumonia, mortality

Harm Endpoints

Bradycardia, hypotension

Narrative

Noninvasive ventilation (NIV) is an effective therapy for hypercapnic and hypoxemic respiratory failure and can reduce the need for intubation and mechanical ventilation.1 It may also reduce intensive care unit (ICU) length of stay, pneumonia, and mortality.2^, 3^, 4^, 5^, 6 However, NIV can be uncomfortable for patients due to the mask interface and respiratory pressures delivered, and over one-third of patients placed on NIV will experience agitation.7^, 8 Intolerance to NIV typically requires intubation. A variety of interventions can be utilized to improve compliance with NIV, including medications such as dexmedetomidine, an α-2 agonist with sedative and analgesic effects.9 Current guidelines recommend the use of a non-benzodiazepine sedative such as propofol or dexmedetomidine in critically ill, mechanically ventilated adults, as these medications may improve delirium, ICU length of stay, and duration of mechanical ventilation.10

The systematic review summarized here included randomized trials (RCTs) of adults > 18 years admitted to the ICU with acute respiratory failure treated with NIV.11 Authors included studies in which the intervention group received dexmedetomidine in the ICU (any dose, initiation day, route, frequency, formulation, administration, and duration) and the control group received a different form of pharmacologic sedation or placebo. Studies of patients treated chronically at home with NIV who were admitted to the ICU, patients supported with NIV for post-extubation respiratory failure, and patients with alcohol withdrawal were excluded. Outcomes included need for intubation and mechanical ventilation, delirium, ICU length of stay, mortality, duration of NIV, pneumonia, bradycardia, and hypotension.

The authors of the meta-analysis identified 12 RCTs comprising 738 ICU patients.11 The mean age was 61.5 years and 36% of patients were women. Four trials included 200 patients with baseline agitation or delirium. Four trials provided an intravenous (IV) loading dose of dexmedetomidine, 2 trials did not report a loading dose, and 6 trials used IV infusion only. Most studies used a dosing range of 0.2-0.7 micrograms/kilogram/hour IV, though 3 trials used a wider range of dosing (0.2-2 micrograms/kilogram/hour). Six RCTs used placebo as the comparator, while 2 used haloperidol, 3 used midazolam, 1 used propofol, and 1 utilized sedation according to the ICU team.

Dexmedetomidine reduced the need for intubation and mechanical ventilation (Relative risk [RR]: 0.54; 95% confidence interval [CI]: 0.41-0.71; absolute risk difference [ARD]: 17%; number needed to treat [NNT]: 6; moderate certainty), delirium (RR: 0.34; 95% CI: 0.22-0.54; ARD: 19%; NNT: 6; moderate certainty), ICU length of stay (mean difference [MD]: 2.40 days fewer ICU days; 95% CI: 3.51-1.29, low certainty), and pneumonia (RR: 0.30; 95% CI: 0.17- 0.52; ARR: 16.7%; NNT: 6; moderate certainty). Dexmedetomidine did not impact survival (low certainty). However, it increased the risk of bradycardia (RR: 2.80; 95% CI: 1.92-4.07; ARD: 23%; number needed to harm [NNH]: 4; moderate certainty) and hypotension (RR: 1.98; 95% CI: 1.32-2.98; ARD: 18%; NNH: 5; moderate certainty).11

Caveats

While 12 RCTs were included, not all trials reported all the outcomes. In particular, the data pertaining to mortality were limited. The low number of patients who died in the ICU resulted in imprecision in reporting this outcome. Second, due to the relatively small population of 738 patients, results may be associated with a type I or type II error due to overestimation or underestimation of the statistical significance of the results, respectively. Third, authors were unable to complete analysis of several prespecified outcomes such as hypertension and subgroup analyses by age and dose of dexmedetomidine due to limited patient-level data. This also limited subgroup analyses including patients with and without agitation or delirium at the time of enrollment. Fourth, the systematic review was not able to generate a funnel plot for publication bias due to the small number of trials. Of note, dexmedetomidine was associated with an increased risk of bradycardia and hypotension. Overall, bradycardia occurred in 78 of 236 patients and hypotension in 75 of 232 patients receiving dexmedetomidine. However, treatment for bradycardia and hypotension varied significantly in the included studies, including decreased dexmedetomidine infusion, vasopressor or inotrope infusion, or no intervention, with not all studies reporting treatment for bradycardia and hypotension. Finally, there was significant statistical heterogeneity concerning ICU length of stay, as well as clinical heterogeneity concerning the other outcomes. While the other outcomes demonstrated little to no statistical heterogeneity, the underlying etiology requiring the use of NIV differed in the included studies, raising concerns of significant clinical heterogeneity and threatening the validity of the results.

Based on the presented evidence, we have assigned a color recommendation of Yellow (Data Inadequate) for use of dexmedetomidine in critically ill patients undergoing NIV. Further data from larger trials are needed to provide a more accurate effect size and to evaluate the ideal dosing, as well as studies comparing dexmedetomidine with other agents. Additional research is also needed to assess the effects on important subgroups such as patients with pre-existing delirium or agitation, elderly patients, and those separated by etiology of acute respiratory failure (e.g., hypoxemic, hypercarbic).

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

June 23, 2021

Source

Tan E, Braithwaite I, McKinlay CJD, et al. Comparison of acetaminophen (paracetamol) with ibuprofen for treatment of fever or pain in children younger than 2 years: a systematic review and meta-analysis. JAMA Netw Open. 2020;3(10):e2022398.

Study Population: 241,138 patients younger than two years from 18 studies who received acetaminophen or ibuprofen for fever or pain

Efficacy Endpoints

Reduction in temperature or pain within four hours and four to 24 hours

Harm Endpoints

Serious adverse events, renal impairment, gastrointestinal bleeding, liver injury, severe soft tissue infection, empyema, and asthma or wheezing

Narrative

Fever and pain are common in childhood. Acetaminophen and ibuprofen are the most widely used medications to treat these conditions, with up to 95% of children receiving acetaminophen by nine months of age.1 The decision to use ibuprofen or acetaminophen for fever or pain typically varies based on purported benefits and harms, and recommendations regarding these medications differ in terms of age and dosing.2^, 3

Several older studies suggest that acetaminophen is safer because ibuprofen was associated with an increased risk of acute kidney injury and serious bacterial infection or soft tissue infection.4^, 5^, 6^, 7^, 8^, 9 However, more recent evidence suggests early acetaminophen exposure may be associated with childhood asthma.10^, 11 Previous meta-analyses have suggested that ibuprofen is likely more effective than acetaminophen for fever and at least as effective as acetaminophen for pain relief, with no difference in adverse events.12^, 13 These earlier reviews had significant heterogeneity and included ages one month to 18 years, limiting the applicability to children younger than two years.

The meta-analysis in this review included studies of any design and children younger than two years; directly compared acetaminophen with ibuprofen; and reported antipyretic, analgesic, and/or safety outcomes.14 Primary outcomes were fever (continuous variable) or pain within four hours of treatment. Secondary outcomes included the proportion of patients who were fever-free or pain-free within four hours of treatment and at four to 24 hours (categorical variable). Safety outcomes comprised serious adverse events, renal impairment, gastrointestinal (GI) bleeding, liver injury, severe soft tissue infection, empyema, and asthma or wheezing. The authors did not place limits on total duration of follow-up.

The meta-analysis included 18 studies with 241,138 patients from seven countries and multiple health care settings. The studies consisted of 11 randomized studies (n = 28,450) and seven nonrandomized studies (n = 212,688). Of these nonrandomized studies, two were prospective cohort, two were retrospective cohort, one was case-control, one was crosssectional, and one was retrospective cross-sectional. Based on data from four randomized controlled trials only, children who received ibuprofen had lower temperatures within four hours compared with those who received acetaminophen (standardized mean difference = 0.38°C [0.21°F]; 95% CI, 0.08 to 0.67; moderate-quality evidence). Based on data from five randomized controlled trials, ibuprofen was associated with a higher likelihood of the child being afebrile at four hours (odds ratio = 1.86; 95% CI, 1.01 to 3.44; absolute risk reduction = 12.5%; number needed to treat [NNT] = 8) and at four to 24 hours (odds ratio = 2.22; 95% CI, 1.55 to 3.17; absolute risk reduction = 18.5%; NNT = 6).14

No study reported pain outcomes within four hours of treatment. Two randomized studies found reduced pain based on a variety of scales (standardized mean difference = 0.20; 95% CI, 0.03 to 0.37) and higher likelihood of the child being pain-free (absolute risk reduction = 25.2%; NNT = 4) at four to 24 hours. There were low rates of adverse events reported in the randomized studies, with most studies reporting no adverse events during the follow-up period. Seven studies with moderate-quality evidence reported similar potential harms within 28 days, including renal impairment, liver injury, and asthma or wheezing, whereas low-quality evidence suggested similar rates of GI bleeding. No severe soft tissue infections or empyema was reported. Two randomized trials reported similar rates of asthma or wheezing and no serious adverse events in the long term (greater than 28 days).

Caveats

There are several limitations associated with this meta-analysis. Studies differed in setting, drug dosages, sample size, and duration of therapy. Only four studies with 796 patients evaluated analgesia, and there were no studies reporting on pain within four hours of therapy. Pain scales were heterogeneous and included the Children’s Hospital of Eastern Ontario Pain Scale, facial expression, a discomfort scale, an irritability score, and the Non-communicating Children’s Pain Checklist. Sample sizes were small in most studies, limiting the comparison of adverse events. Only nine studies evaluated medication safety as a primary outcome, and measurement bias may have affected evaluation of adverse events. Most randomized studies provided adverse event data for less than 28 days, with most of the long-term adverse events reported only by observational studies.

Conclusion: The American Academy of Pediatrics recommends against the use of acetaminophen in infants younger than three months or ibuprofen in those younger than six months without clinical evaluation, although it states that both medications are safe and effective when used appropriately.2 Based on the available evidence, the meta-analysis found that ibuprofen is more effective in reducing temperature and pain at various follow-up periods compared with acetaminophen.14

These findings are consistent with previous studies, but the meta-analysis found no differences in adverse events. Thus, we have assigned a color recommendation of green (benefits outweigh harms) for the use of ibuprofen compared with acetaminophen in children younger than two years. Further study is needed to evaluate combination therapy, specific dosing, and use in those younger than six months.

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

June 16, 2021

Source

Klowak JA, Hewitt M, Catenacci V, et al. Levetiracetam versus phenytoin or fosphenytoin for second-line treatment of pediatric status epilepticus: a meta-analysis. Pediatr Crit Care Med. Published online March 10, 2021.

Study Population: 1,575 children (<18 years of age) with status epilepticus who failed benzodiazepine therapy for seizure cessation

Efficacy Endpoints

Time from intervention to seizure cessation, early seizure cessation (within 20-40 minutes of intervention), and late seizure cessation (within 1-3 hours)

Harm Endpoints

Failure of response (seizure recurrence or need for third-line antiepileptic medication), pediatric intensive care unit admission, intubation, and development/neurocognitive assessment after hospital discharge, hypotension, respiratory depression, arrhythmia, acute behavioral change, extravasation, drug-related severe adverse event, and death

Narrative

Status epilepticus, defined by continuous seizure activity for at least 5 minutes or recurrent seizures without a return to baseline, is a neurologic emergency.1^, 2^, 3 The annual incidence in children approximates 20 per 100,000 population.4 Unfortunately, a significant portion of children presenting with status epilepticus can experience long-term neurologic, cognitive, or behavior abnormalities;5 therefore, rapid treatment is essential.3 While benzodiazepines are considered first-line treatment, they fail to stop seizure activity in approximately one-third of patients.6^, 7 Common second-line agents include phenytoin, its prodrug fosphenytoin, and levetiracetam.6^, 7^, 8^, 9

The meta-analysis summarized here examined data from randomized controlled trials (RCTs) comparing levetiracetam with phenytoin or fosphenytoin for status epilepticus refractory to at least one dose of benzodiazepine.10 Authors included data from trials enrolling primarily patients under age 18 years, or else reported the results of a pediatric subgroup from trials enrolling patients of all ages. The three main endpoints included overall time from intervention to seizure cessation, early seizure cessation (within 20-40 minutes), and late seizure cessation (within 1-3 hours). Additional outcomes included medication failure (seizure recurrence or need for thirdline medication), requirement for pediatric intensive care unit admission, and neurocognitive function after hospital discharge. Major safety outcomes included hypotension, intubation, arrhythmia, and death.

Seven relevant RCTs (n=1,575) were identified and their data pooled. Mean patient age ranged from 2.3 to 6.1 years. Four studies used phenytoin and 3 studies used fosphenytoin. There was no difference between levetiracetam and the comparator for any measure of seizure cessation. There was also no difference in failure of response, pediatric intensive care unit admission, intubation, and drug-related severe adverse events.

Caveats

There are several limitations associated with this meta-analysis.10 First, seizure etiology varied, including central nervous system infection, cryptogenic seizure, epilepsy, and febrile seizure. Over 75% of patients experienced generalized seizures, the remainder experienced focal seizures. While the meta-analysis protocol was pre-registered, three included RCTs were not. Trials varied in reporting and definition of primary outcomes. Moreover, evaluating timing of seizure cessation is complicated by the different infusion durations of the agents. Dosing of levetiracetam also differed between studies, ranging from 20 mg/kg to 60 mg/kg, with infusion durations of 5-20 minutes. Authors of this meta-analysis defined a minimal important difference as 5 minutes for seizure cessation, and thus deemed a 3.1-minute difference between groups as clinically nonsignificant, a threshold that could be debated.

In addition, analysis of safety outcomes such as respiratory depression suffered from heterogeneity, imprecision, and low certainty evidence. This resulted in the outcome ‘intubation’ being no different between groups in 5 trials, while ‘respiratory depression’, theoretically defined by the need for intubation, was slightly (about 4%) more common with phenytoin—a difference driven by results from a single trial.11 Therefore, we did not present the number-needed-to-harm for respiratory depression/intubation reported by the meta-analysis.

The Established Status Epilepticus Treatment Trial (ESETT), included in the meta-analysis, may be the most reliable evidence on second-line anticonvulsants for generalized convulsive status epilepticus.11 Its inclusion in pooled data is a bit fraught, however, as the trial compared three agents, not two, and was then extended, by over a year, for additional pediatric enrollments. Of the three large multicenter efforts that together contributed nearly 80% of the data in the metaanalyses, ESETT and one other trial utilized a double-blinded design.11^, 12 This other trial also evaluated three arms, including phenytoin, valproate, and levetiracetam in 110 patients age 3 months to 12 years.12 All other included trials were open label with two arms. ESETT compared levetiracetam, fosphenytoin, and valproate for patients > 2 years of age with status epilepticus who continued to seize despite receiving benzodiazepines.11 Approximately half of the 462 subjects in each arm achieved the primary endpoint (seizure cessation <60 minutes), with no difference between groups. The safety outcomes of life-threatening hypotension and cardiac arrhythmia were rare and did not differ by treatment group. However, in children (n=225), endotracheal intubation appeared to occur more in the fosphenytoin group (33%) than the levetiracetam (8%) or valproate group (11%).11^, 13 As the authors note, this was a secondary outcome measured in a subgroup, and suggests a safety concern of unclear significance, not found in any other setting or by any other research group.

Based on the available evidence, the review summarized here found levetiracetam compared to phenytoin or fosphenytoin was associated with similar time to seizure cessation and a similar safety profile. Because of similarity in efficacy and safety profiles, we have assigned a color recommendation of Yellow (similar efficacy and safety profiles). One reasonable conclusion from the data is that levetiracetam is not superior to phenytoin/fosphenytoin in the acute setting. Whether they are truly equivalent would require equivalence (non-inferiority) trials, which have yet to be done. However no clinically important safety or efficacy differences were found in the trials data reviewed. For patients with benzodiazepine-refractory status epilepticus, existing data suggest that levetiracetam and phenytoin are both reasonable agents.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 30, 2021

Source

Ducharme FM, Ni Chroinin M, Greenstone I, Lasserson TJ. Addition of long-acting beta2-agonists to inhaled steroids versus higher dose inhaled steroids in adults and children with persistent asthma. Cochrane Database Syst Rev. 2010;(4):CD005533.

Salpeter SR, Buckley NS, Ormiston TM, Salpeter EE. Meta-analysis: effect of long-acting beta-agonists on severe asthma exacerbations and asthma-related deaths. Ann Intern Med. 2006;144(12):904-912.

Cates CJ, Wieland LS, Oleszczuk M, Kew KM. Safety of regular formoterol or salmeterol in adults with asthma: an overview of Cochrane reviews. Cochrane Database Syst Rev. 2014;(2):CD010314.

Study Population: Adult patients with asthma of differing severities

Efficacy Endpoints

Asthma-related quality of life, asthma exacerbation requiring oral steroids, exacerbation requiring hospitalization, life threatening asthma-related event (death or intubation), all-cause death

Harm Endpoints

Death, asthma-related death, life threatening asthma event

Narrative

Long-acting beta-agonists (LABAs) relax smooth muscle in asthmatic lungs, potentially preventing attacks and improving symptoms. While inhaled corticosteroids (ICS) are first-line ‘controllers’, adding LABAs to the regimen is common. Unfortunately, the 2006 ‘SMART’ trial showed that LABAs increase severe attacks and asthma-related deaths.1 However there remains debate on whether partnering LABAs with ICS neutralizes this risk. The FDA recently removed its black box warning2 for LABAs on the basis of a meta-analysis, and most guidelines suggest adding them for improved control. The common alternative involves increasing the dose of ICS. This summary examines adding LABAs to ICS versus increasing ICS dose for improved control.

The Cochrane review discussed here examined 48 trials of >33,000 asthmatics on ICS randomized to add a LABA or increase the ICS dose.3 Adding LABAs did not reduce hospitalizations, deaths, or severe attacks. However, it did reduce asthma exacerbations requiring 3-5 days of oral corticosteroids (relative risk 0.9, 0.95CI: 0.8-0.98; absolute risk difference [ARD] 1.5%; NNT: 73). For those at highest risk (patients with an estimated 20% risk of an exacerbation during a 12-week study period) the ARD was 2.2% (NNT: 45), while for those at lower risk (1% risk of exacerbation during the study period) the ARD was just 0.15% (NNT: 673). In other words, those with more severe asthma were more likely to benefit. LABAs also improved symptoms slightly, however asthma-related quality of life was unchanged.2

The issue of whether LABAs increase asthma deaths or near fatal asthma events when partnered with ICS is complex. Different topic reviews and meta-analyses designed to answer this question offer different answers, although the differences are mostly based on statistical significance (relative risks of 1.4 or higher for fatal and near fatal events with the presence of a LABA are seen across virtually all reviews). The review by Salpeter et al., for instance, notes statistically increased fatalities and near fatalities even when combined with steroids,4 while a 2014 Cochrane review finds a non-statistically significant odds ratio of 1.4 (95% CI: 0.6-3.4).1^, 5

Because of this uncertainty, in 2010 the FDA commissioned four LABA manufacturers to conduct safety trials pitting ICS alone against LABA plus ICS. In a meta-analysis of 36,000 subjects in these four trials there were just 2 asthma deaths (LABA group) and 3 intubations (1 LABA, 2 control), yielding a relative risk of 1.5 (95% CI: 0.3-9.0) for this endpoint.6

Caveats

There are many reviews of LABA safety and efficacy. We focus on one clinical choice, increasing the steroid versus adding a LABA. However other comparisons may be more useful at times, for instance when steroid doses are maximized, or other options (e.g. anticholinergics, leukotriene inhibitors) fail.7^, 8

Available data are imperfect for the answers we seek. The 2006 SMART trial was stopped early (both because of harms and slow enrollment) and was designed to address dangers due to LABAs generally. It was not designed specifically to test whether those dangers remain when LABAs are partnered with ICS. Moreover, many reviews fail to control for ICS dose when showing this comparison. This makes interpretation tricky since ICS improves outcomes in a dose-dependent fashion—thus when a group on LABAs + ICS suffers more life-threatening events than a group taking double-dose ICS, is this because LABAs increased events or because the higher ICS dose decreased them?

One large 2012 review catalogs prior reviews and performs meta-analyses that control for ICS and other confounders. This ‘meta-review’ finds higher death and intubation rates with LABA plus ICS than ICS-alone (Table 3),9 but the differences are statistically nonsignificant. Citing mostly unpublished drug company trials, in which few deaths or intubations occurred in either group, the authors conclude partnering LABAs with ICS may neutralize the dangers.9

Importantly, the benefits of LABAs for symptomatic relief were statistically significant but clinically questionable compared to increasing ICS dose, and conferred no advantage in asthma-related quality of life. However, one potential advantage of using LABAs remains: less ICS exposure, which could lead to less inhibition of growth in children.

On the safety of LABAs when partnered with ICS, however, we believe the studies commissioned by the FDA enrolled the wrong cohort. The purpose of commissioning the trials was to evaluate if combining LABAs with ICS prevents the established increase in fatal and near fatal events caused by LABAs. The 2006 SMART trial (similarly commissioned to address concerns about LABA dangers) saw 86 fatal or near fatal events among 26,000 subjects. This robust number of events made it possible to detect an important difference between groups.1 The 2018 meta-analysis of four FDA studies, however, reports just 5 fatal or near fatal events in 36,000 trial subjects, suggesting a near zero risk cohort.5 The FDA-commissioned trials therefore studied a population in whom it was not possible to answer the question they were commissioned to answer—in a cohort where virtually no one dies, it is not possible to reduce deaths.

In addition, and ironically, neither an important benefit nor an important harm was possible in a cohort at such a low risk. Adding LABAs in this cohort resulted in a less than 2% reduction in minor exacerbations requiring 3-5 days of oral steroids.5 However the ICS dose was kept the same in both groups, whether LABAs were added or not. Therefore, even the small benefit would almost certainly have been smaller, or disappeared, if the ICS dose had been increased in the control group. Meanwhile, the higher risk asthmatics enrolled in the SMART trial are the group most likely to see meaningful benefits since, as data consistently show, the higher the risk, the greater the benefit from LABAs. The finding of little or no meaningful benefit in the FDA trials is consistent with a voluminous body of literature, summarized above, showing that patients at low risk see virtually no meaningful benefit.

Of note, the many trials performed on subjects at near zero risk for asthma-related death or intubation may help explain the differing conclusions from safety reviews. Earlier reviews found statistically significant differences while later reviews often did not, despite relative risks remaining the same (typically 1.4 to 1.5). However instead of 1 in every 350 subjects experiencing a life-threatening event (as in the SMART trial),1 in the more recent trials life-threatening events were 20-fold less common, or roughly 1 in 7000 subjects.5 In other words, while higher rates of fatal and near fatal events persisted, they became statistically nonsignificant as participants in later trials virtually never had an event. If, for comparison, the FDA-commissioned trials had enrolled subjects similar to those in the SMART trial, most likely more than 100 events would have occurred—enough for a meaningful statistical analysis.

Meanwhile, and perhaps more concerning, it is the higher risk group who benefits most from LABAs, for whom guidelines recommend LABAs, and who are almost universally prescribed LABAs. This issue is critical for real world application: one commonly cited study suggests <5% of asthmatics in the community would meet eligibility criteria for most randomized trials of interventions for asthma.9 The FDA trial cohorts support this notion: patients at higher risk, i.e. those typically prescribed LABAs in the community, were not enrolled. Instead the researchers excluded any patient with ‘a history of life-threatening asthma’, ‘unstable asthma’, and anyone with multiple hospitalizations. These are the patients most consistently prescribed LABAs in practice, and the 2006 SMART trial made no such exclusions.

An editorial accompanying the 2018 meta-analysis of FDA trials declares LABAs safe. The editorialists cite equal hospitalizations between groups, and note the FDA removed its black box warning based on this finding.2 However, in the 2006 SMART trial hospitalizations were also the same between groups.1 The more recent trials existed solely to address dangers established in the SMART trial, but increased hospitalization was not one of them. Declaring the drugs safe on this basis, without resolving whether LABAs increase deaths in the population most commonly taking them, seems unfounded.

The widespread use of LABAs today is based entirely on a theory that when paired with ICS the increase in fatal and near fatal events caused by LABAs disappear. But in the only trial with adequate power to address this theory LABA dangers were present whether or not patients were also prescribed an ICS inhaler (SMART trial, see Table 8).1 While adherence to both inhalers was not tracked, leaving much room for uncertainty, this finding highlights a concerning fact: the only study of LABAs adequately powered to detect life threatening harms presents a direct challenge to the theory that ICS is protective.

Finally, all-cause mortality is a better outcome than asthma-related deaths but is often not reported. Moreover, using different endpoints from these trials is fraught, and selecting the right review for reporting the estimates is difficult, which is why we offer a range. Our NNH of 1430 is based on one estimate. Which estimate is the truth is anyone’s guess.

We have chosen to designate the color recommendation of RED because of a small potential benefit of questionable clinical utility (<2%, avoiding a brief course of oral steroids), no quality of life benefit, and the lingering possibility of a fatal harm. We considered BLACK (harms>benefits) however there is uncertainty on harms, and it remains possible that ICS have a protective effect against the harms that has not yet been found. A proper trial would require enrollment of real world patients at genuine risk for a fatal asthma event, and we hope to see such a trial in the future. We recognize this color designation and interpretation are at odds with some clinical guidelines, and the FDA’s recent decision, as well as common practice. We hope providers and patients can use this to discuss these issues and come to helpful, informed conclusions.

Author

Shahriar Zehtabchi, MD
Supervising Editors: Allan Wolfson, MD

Published/Updated

Source

Donaldson L, Fitzgerald E, Flower O, Delaney A. Review article: Why is there still a debate regarding the safety and efficacy of intravenous thrombolysis in the management of presumed acute ischaemic stroke? A systematic review and meta-analysis. Emerg Med Australas 2016;28(5):496–510.

Study Population: 10,431 patients in 26 randomized trials comparing intravenous thrombolysis with placebo or standard care in acute ischemic stroke

Efficacy Endpoints

Good functional outcome, defined as a modified Rankin Score of 3 or less, i.e. some residual disability requiring assistance but able to walk and care for personal needs independently

Harm Endpoints

Symptomatic intracranial hemorrhage (as defined by individual trials) and overall mortality

Narrative

This is the third NNT summary of thrombolytics for acute ischemic stroke. The first gave thrombolytics, as a class, a "red color recommendation: no benefit." The second gave alteplase, a single agent, a "green color recommendation: benefit>harm." As no relevant trials were published between the two, both author groups examined essentially the same data and arrived at opposing conclusions. We believe it would be hubris to presume this third summary will arrive at the one true answer. We focus, therefore, on the uncertainty we believe leads to conflicting interpretations.

The systematic review we chose to summarize includes 26 randomized trials of more than 10,000 participants, assessing the benefits of thrombolysis for acute ischemic stroke.1 The authors report a 3.2% improvement in good neurologic outcome, a 5.4% increase in symptomatic intracranial hemorrhage, and a 2.5% increase in mortality. However, we question the certainty implied by these summary numbers.

There are multiple relevant systematic reviews with varying methods but similar findings. In one widely cited review, Emberson and colleagues reported only on alteplase (a problem we discuss below) and find a 5% improvement in neurologic outcomes, a 5.5% increase in intracranial hemorrhage, and a 1.4% increase in 90 day mortality that was not statistically significant.2 The 2014 Cochrane review by Wardlaw and colleagues arrives at similar conclusions with significant improvement in neurologic outcomes, increased intracranial hemorrhage, and increased mortality.3 Thus our conclusions and discussion are unchanged by choice of review, and reflect our belief that pooling data on this topic is overly simplistic and masks profound uncertainty.

Caveats

One of the great conceptual difficulties of summary statistics like the number-needed-to-treat (NNT) is the implication of certainty. A major strength is its simplicity, making complex research easier to understand. A weakness, however, is also simplicity, hiding the complexity of research, ignoring confidence intervals, and obscuring biases. For most topics, these details are more important than any individual number.

Sources of uncertainty:

Conflicting individual trial results: A major source of uncertainty is the differing results of trials. Among 26 trials in this systematic review, 24 research groups found no benefit in their selected primary outcome.1 Moreover, the two that claim a benefit (NINDS part 2 and ECASS III)4^, 5 had baseline imbalances that may explain the difference. In re-analyses adjusting for imbalances in both these trials the benefits disappear.7^, 8 However, in some re-analyses of NINDS-2 the benefit is maintained, adding to the uncertainty.9^, 10

Clinical heterogeneity of individual trials: The 26 trials are clinically heterogeneous, enrolling stroke patients of differing demographics, treatment times, stroke severities, anatomic territories, and thrombolytic agents. The author of the first NNT summary felt this was too much heterogeneity for appropriate pooling, a position supported by the major differences in conclusions drawn depending on which studies an author group chooses to include.

Selective emphasis on trials claiming benefit: It is circular and erroneous logic to claim efficacy for thrombolytics based on the trial characteristics of the two positive trials. First, there is legitimate debate about whether they were truly positive. Second, selectively highlighting positive results is a form of the "Texas sharpshooter fallacy".11 For instance, because both NINDS and ECASS III used alteplase, some have suggested alteplase is a superior agent.4^, 5 On close inspection this logic falters: few trials have compared thrombolytic agents head to head, nine additional trials of alteplase are negative, and systematic reviews consistently find no heterogeneity of effect between agents.3^, 5^, 6 Moreover, in evaluating drug efficacy, establishing a class effect is generally a prerequisite for debating or comparing individual agents.12 Therefore, while it may increase complexity, we believe it is a mistake to exclusively examine data from the agent used in the two trials that claimed benefit.

Likewise, while there are theoretical reasons to think early treatment is better, this has not been directly tested and is not strongly supported by data. Neither Donaldson et al. nor the Cochrane review find an interaction between time to treatment and effect.1^, 3 IST-3, the largest trial of thrombolytics for stroke, found better outcomes among those treated after 4.5 hours than in patients treated at 3-4.5 hours from onset. Again, we feel it is best to consider this literature as a whole rather than using time windows selected based on outlying (i.e. positive) results.

Individual trial bias: Bias is a major source of uncertainty in all scientific research. Importantly, using the GRADE tool,13 Donaldson et al. rate the risk of bias as “serious” for all outcomes.1 One notable source is the outcome scales used, for instance the modified Rankin score. This score is known to be subjective with poor inter-rater reliability and questionable validity. When trained neurologists examine the same patients there is substantial variability in Rankin score assignments.14^, 15 Compounding the problem, some trials assessed patients by phone or mail, a choice certain to increase variability and imprecision. For example in IST-3, which contributes nearly 40% of subjects in the Donaldson meta-analysis,16 results were obtained using telephone and mail follow-up, and non-blinded. This subjectivity is important, because removing this trial from the pooled analysis removes any finding of benefit.

Early stoppage and low power: Because larger trials are weighted more heavily in a meta-analysis, early termination (which reduces trial size) can significantly affect results. Five thrombolytic trials were stopped early for harm or futility.17^, 18^, 19^, 20^, 21 Together these would have enrolled more than 2,000 additional subjects who, had they been included, may have neutralized or even reversed findings from the two trials claiming benefit, NINDS2 and ECASS III (combined n=1,445).4^, 5 Furthermore, while over 10,000 subjects were enrolled in stroke trials, some individual trials for acute myocardial infarction enrolled far more, and in aggregate those trials included more than 60,000.22^, 23 The comparatively small number of participants in stroke trials means chance findings like baseline imbalances are both more likely and more influential, furthering uncertainty.

Harms and overlapping outcomes: In contrast to the heterogeneous data on benefits, the data on harms are more certain. Exact numbers vary based on definitions and whether one focuses on fatal, symptomatic, or any hemorrhage, but an increase in intracranial hemorrhage is certain. More importantly, there is also an increase in mortality with thrombolytics.1^, 3

Prior NNT summaries directly compared the number of patients with intracranial hemorrhage to the number with a good neurologic outcome, which can be misleading. Any harms due to intracranial hemorrhage are incorporated into a final neurologic outcome assessment, and therefore good functional outcome is far more important than intracranial hemorrhage.

There is, however, a difficult comparison to consider between the chance of neurologic benefit and the increased risk of death. Thrombolytics appear to increase mortality;1^, 3 however many patients may be willing to accept this risk for an increased chance of functional improvement. Shared decision making for this already very difficult decision is made even harder by the fact that a mortality harm is more certain than any benefit.

Financial conflicts: A final important source of uncertainty in this research arises from financial conflicts of interest, which are unfortunately common in medical research. Although they do not always invalidate findings, they complicate interpretation because studies with financial conflicts are more likely to report positive findings.24^, 25^, 26^, 27 Such conflicts are well documented in both the original thrombolysis research and the subsequent guidelines28 and are known to affect the quality and conclusions of systematic reviews. This may be an unconscious contributor to the divergent conclusions seen for this topic.29

Like all such work, our summary has inherent limitations. Not included in this review, for instance, are newer trials using advanced imaging to select patients for thrombolysis,30^, 31^, 32^, 33 trials we felt should not be combined with studies using thrombolytics based largely on clinical criteria. Also of note, the Donaldson review uses a different cut-off on the modified Rankin Score (≤3) than others. The authors argue being able to walk and care for one’s self independently is the threshold most would consider a good outcome. While this cut-off may seem unconventional, other reviews find similar harms and benefits using different cutoffs, suggesting to us that cutoff points are not an important differentiator of conclusions.2^, 3

The uncertainty in these data may help explain differing views of thrombolysis for stroke: trials are heterogenous with many sources of bias; rare positive findings are marred by baseline imbalances but continue to have outsize influence; outcomes were assessed using imprecise, subjective scales; early terminations tilt the data; outcomes overlap; and conflicts abound. There is no simple summary. We find the only appropriate conclusion is uncertainty.

Summaries of the literature cannot answer questions the literature cannot answer. Thrombolytics for stroke has been among the most debated topics in medicine for three decades, perhaps due to underlying data deficiencies. The Donaldson et al. meta-analysis suggests a potential functional outcome benefit which must be balanced against increased mortality. However, the authors warn us that the findings are at serious risk of bias, and pooling results introduces more bias. Three NNT summaries have now arrived at three conclusions. This question will not be answered by re-analyzing the same data. We need additional, carefully executed trials.

In the meantime, it is difficult to offer clear guidance. Clinicians should be aware of the weaknesses in these data, and patients deserve to know as well. We believe the data are not compelling enough for specific recommendations. Shared decision making is essential, and physicians are left with the difficult task of guiding patients through a decision without a clear answer.

See theNNT.com's previous reviews of this topic:
Thrombolytics for Acute Ischemic Stroke, March 25, 2013
Tissue Plasminogen Activator (tPA) For Acute Ischemic Stroke, January 11, 2019

Author

Justin Morgenstern, MD; W. Ken Milne, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 21, 2021

Source

Venekamp RP, Burton MJ, van Dongen TMA, van der Heijden GJ, van Zon A, Schilder AGM. Antibiotics for otitis media with effusion in children. Cochrane Database Syst Rev. 2016;(6):CD009163.

Study Population: 3663 children with otitis media with effusion seen in primary care and specialty offices and enrolled in 25 randomized controlled trials

Efficacy Endpoints

Resolution of OME by tympanometry or otoscopy

Harm Endpoints

Medication adverse events (e.g. diarrhea, vomiting, rash)

Narrative

Otitis media with effusion (OME), one of the most common diseases in early childhood, is a self-limited illness1 with an estimated recurrence rate of 50% within 24 months.2 OME is characterized by an accumulation of fluid in the middle ear without symptoms or signs of acute infection.3 In most cases OME causes mild hearing impairment for a short period. OME in the pediatric population can lead to diminished hearing which in theory could hinder early learning. Some studies have shown an association with adult hearing loss as well.4 In one study, for instance, OME and recurrent otitis media before the age of 3 were weakly associated with worse classroom concentration, mathematical skills, oral performance and reading skills.5

OME can be managed surgically (ventilation tubes, adenoidectomy) or medically (antibiotics, antihistamines, decongestants). The American Academy of Pediatrics 2019 advises against the use steroids, antihistamines and antibiotics in the management of OME but recommends surgical management for children with OME for more than 3 months and hearing deficits.6

The systematic review discussed here7 is an update from a previous Cochrane review in 2012,8 and reports meta-analyzed data collected from 3258 children from 23 trials. Seven were open-label, three were investigator-blinded, and 15 were double-blinded randomized controlled trials. The subjects’ mean age was 4.7 years, and 54% were boys.

Tympanometry, alone or in combination with otoscopy, was used to define OME and to monitor its resolution. Several antibiotics were used in the study but the commonly used included amoxicillin, amoxicillin/clavulanate potassium, trimethoprim-sulfamethoxazole. Duration of treatment varied, but was 10-14 days or four weeks in the majority of studies.

The primary outcome was complete resolution of OME at 2-3 months post-randomization. Data from six trials (523 children, with 484 (93%) included in the analysis) showed children treated with oral antibiotics were more likely to have complete resolution of OME at 2-3 months, relative risk [RR] 2.0, 95% confidence interval [CI], 1.6-2.5; absolute risk difference [ARD]: 24.6%; Number-needed-to-treat [NNT] 5; moderate quality evidence. When resolution of symptoms at 2-4 weeks was examined in data from 14 trials (n=2253, of whom 2091, or 93%, were included in the analysis), results were similar: RR 2.0, 95% CI, 1.5-2.7; ARD: 20%; NNT 5; low quality evidence.

A second primary outcome was adverse medication effects, specifically diarrhea, vomiting, or rash. Data from five trials (n=775, of whom 742, or 96%, were included in the analysis) showed children treated with oral antibiotics were more likely to experience adverse effects, RR 2.2, 95% CI 1.3-3.6; ARD: 20%; NNH 5; low quality evidence). None of the trials reported on long-term adverse effects of antibiotics such as bacterial resistance.

Two studies reported on hearing level based on speech recognition threshold with no differences at 4 weeks. No trials reported on language, speech, or cognitive development.

Caveats

Several caveats must be considered. Some of the studies included were limited by potential risk of bias, and there was inconsistency of effects across trials. Therefore, the Cochrane systematic review rates the quality of evidence for the primary outcome (resolution of OME at 2-3 months) as moderate and for the secondary outcomes of resolution of OME at 2-4 weeks as low. Critically, no trials reported data on language and speech development or cognitive development, theoretically the long term complications treatment of OME aims to prevent.

Similarly, several studies show the use of tympanostomy tubes in children with persistent middle-ear effusion results in short-term improvements in hearing, when compared with watchful waiting.9 However, as with antibiotics there is no evidence of a sustained or long term benefit on hearing., and no data showing improvements in cognitive or developmental outcomes.

Based on this evidence we rate the administration of antibiotics for OME as Red ( harm outweigh benefits). Based on the apparent success of antibiotics in reducing effusion this may seem incongruous., however, the AAP 2019 guideline also recommends against antibiotic use for OME, using logic we agree with: existing data demonstrate no long-term or patient oriented benefit, despite finding clinically meaningful harms. In short, while more patients in the antibiotic group may experience short term resolution of OME, the absence of any data showing impact on the target outcomes of language, speech, and cognitive goals points to a need for further research focusing on patient-centered endpoints.

Finally, long-term considerations such as antibiotic compliance, cost of treatment, recurrence of OME ( 50% recuurence within the first 24 months)2, and development of antibiotic resistance10 should be considered before prescribing antibiotics for this or any other self-limiting disease.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Kelvin Kwofie, MD; Allan B. Wolfson, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 15, 2021

Source

Parry Smith WR, Papadopoulou A, Thomas E, et al. Uterotonic agents for first-line treatment of postpartum haemorrhage: a network meta-analysis. Cochrane Database Syst Rev. 2020;11:CD012754.

Study Population: 6 RCTs including women giving birth vaginally (n = 3674), and 1 RCT where women gave birth either vaginally or by caesarean (n = 64) in low resource settings

Efficacy Endpoints

Additional blood loss of at least 500 mL, death

Harm Endpoints

Transfusion or maternal mortality. Adverse effects including fever, hypothermia, nausea, vomiting, hypertension, headache, shivering, tachycardia, arrhythmia, diarrhea, abdominal pain

Narrative

Postpartum hemorrhage (PPH) may occur in 15% of women giving birth and is the leading cause of peripartum maternal death, with most cases occurring in low income countries.1^, 2 PPH is commonly defined as greater than 500 mL of blood loss after birth. Uterine atony is the most common cause of PPH,3 and oxytocin is recommended as first-line medical therapy by the American College of Obstetricians and Gynecologists and the World Health Organization.3^, 4 However, there are many effective (i.e. better than placebo) ‘uterotonic’ agents for the treatment of PPH,1 including misoprostol which may be used alone or in combination with oxytocin.5

The Cochrane Review summarized here aimed to compare the efficacy of different agents for PPH, and included randomized controlled trials (RCTs) or cluster randomized trials evaluating the benefits and harms of uterotonic agents in women with PPH after vaginal or caesarean birth.6 Trials were eligible if they compared systemically administered uterotonic agents of any dosage, route, or regimen.

The primary outcomes of the Cochrane review included blood loss of >500 mL after enrollment and a composite outcome of maternal death or severe morbidity (hysterectomy, organ dysfunction, transfer to higher level of care, coagulopathy, or shock). From this composite outcome, we report only maternal mortality, a patient-centered outcome reported consistently in the original trials.

The systematic review identified 7 RCTs (n = 3738) that met inclusion criteria. One trial included women giving birth vaginally or by cesarean section, while the others included only vaginal births. Agents evaluated included oxytocin (6 trial arms), misoprostol plus oxytocin (4 trials arms), misoprostol (3 trial arms), and fixed-dose oxytocin/ergometrine plus oxytocin infusion (1 trial arm). Data using this last regimen were limited, of low certainty, and showed unclear effects, therefore we have not summarized this comparison.

Two trials (n = 1787) found no difference in maternal mortality for misoprostol compared to oxytocin. These trials did suggest, however, misoprostol may increase blood transfusions (relative risk [RR]: 1.5; 95% confidence interval [CI]: 1.02-2.1; absolute risk increase: 2.3%; number needed to harm [NNH]: 43). Misoprostol also increased vomiting (RR: 2.5; 95% CI: 1.4-4.5; absolute risk increase: 2.9%; NNH: 34) and shivering (RR: 2.7; 95% CI: 2.3-3.2; absolute risk increase: 26.8%; NNH: 3).

Four trials (n = 1873) of misoprostol plus oxytocin versus oxytocin alone found no primary outcome benefit with the addition of misoprostol for maternal mortality, but did find an increase in adverse effects. These included fever (RR: 3.0; 95% CI: 2.6-3.6; absolute risk increase: 32.1%; NNH: 3) and vomiting (RR: 1.9; 95% CI: 1.2-3.0; absolute risk increase: 2.9%; NNH: 34).

Caveats

The quality of evidence for these analyses ranged from very low to high, with most data rated low or moderate certainty. No studies including injectable prostaglandins, ergometrine, or oxytocin/ergometrine as first line agents were available. Most subjects were women with a singleton term vaginal birth in a low-resource setting. Women with significant comorbidities were excluded from trials, limiting generalizability. There were also differences in dosing and route across interventions. Finally, blood loss can be challenging to quantify based on visual assessment, a measurement method used in some studies, which may have influenced the accuracy of this outcome.

Based on the existing evidence, misoprostol alone or in combination with oxytocin did not improve outcomes and is associated with more adverse effects, a finding supporting current recommendations from ACOG and WHO.3^, 4 Therefore, we have assigned a color recommendation of Black (harms > benefits) for misoprostol. Further study in other settings on the benefits and harms of uterotonic agents is needed.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Linde M, Mulleners WM, Chronicle EP, McCrory DC. Topiramate for the prophylaxis of episodic migraine in adults. Cochrane Database Syst Rev. 2013;(6):CD010610.

Study Population: Adults with episodic migraines who took daily prophylactic topiramate

Efficacy Endpoints

Frequency of episodic migraines in a 4-week period; response to prophylaxis (subjects reporting ≥50% reduction in headache frequency)

Harm Endpoints

Anorexia, paresthesia, fatigue, nausea, weight loss, taste disturbances

Narrative

The global prevalence of migraine headaches is estimated to be about 15%, consisting mostly of young and middle-aged adults.1 Migraine headaches not only cause a burden of pain and suffering, but also result in significant loss of productivity. For example, in the European Union migraines are estimated to cost nearly 50 billion Euros in lost productivity in terms of aggregate direct and indirect costs to society annually.2 A reliable method of preventing symptoms would be beneficial in terms of quality of life as well as societal productivity.3

There are several options for medications that patients can take to mitigate the effects of migraines. One medication that has been recommended for migraine prophylaxis is topiramate.4 Topiramate is classified as an antiepileptic, and while the exact mechanism by which antiepileptics affect migraine headaches is unclear, it has been suggested that topiramate prevents migraine attacks by modulating a variety of neurotransmitters.5

Here we will discuss data from a Cochrane systematic review examining the efficacy of topiramate as a pharmaceutical first line agent for migraine prophylaxis. There are many reasons for a patient to prefer one agent over another, such as desire for pregnancy, comorbid conditions, and side effect tolerance. Therefore, we chose to report the data on the subset of patients in whom topiramate was compared to placebo rather than against alternate agents.

The Cochrane systematic review and meta-analysis discussed here included 17 randomized controlled trials involving a total of 2282 adult patients (>16 years old) who suffered from episodic migraines, which are defined to be discrete flares of migraine symptoms.6 The included studies were required to be prospective, controlled trials of self-administered topiramate taken on a regular basis for preventing migraine attacks. The study endpoints included headache frequency, the proportion of responders (subjects with ≥50% reduction in headache frequency), and adverse events.

Treatment with topiramate increased the number of responders (odds ratio: 3.18, 95% CI, 2.1 to 4.82; absolute risk difference: 24%; number needed to treat [NNT]: 4, 95% CI 3 to 6). The included studies used topiramate in doses of 50 mg, 100 mg, or 200 mg per day. The meta-analysis found that topiramate significantly reduced the frequency of headaches in the treatment group within 28 days, as compared to placebo (mean difference for headache frequency: -1.20, 95% confidence interval (CI) -1.59 to -0.80; 9 trials, 1737 subjects).

The results remained statistically significant when data were stratified by dose.5 Data from three studies that directly compared 50mg, 100 mg, and 200mg doses of topiramate favored the 100 mg. 100 mg resulted in an increased proportion of responders compared to the 50mg dose. Meanwhile, there was no significant superiority of the 200mg dose over the 100 mg dose.2^, 5

Harms: The method of reporting adverse events varied from trial to trial and not all trials reported adverse events.

Only 3 trials included in the systematic review explicitly reported adverse events.6 It is thus likely that adverse events were underreported. The lack of adequate information about harms should be discussed with patients as part of the shared decision-making process when this medication is being considered for prophylaxis.

In the studies where adverse events were adequately reported, the risk of adverse events was higher in the topiramate group when compared to placebo. The adverse events were dose-dependent and more common in patients on the 200 mg per day regimen than those on the 100 mg per day regimen.6 The adverse events included anorexia (number needed to harm [NNH]: 12-17), fatigue (NNH: 12-25), memory problems (NNH: 12-25), nausea (NNH: 17), paresthesia (NNH: 2-3), taste disturbance (NNH: 7-14), and weight loss (NNH: 11-25).6 In the group of patients who took less than 100 mg topiramate per day, the incidence of most adverse effects was not statistically different compared to placebo, with the exception of weight loss and taste disturbance.

Another systematic review published in 2015 assessed the comparative effectiveness of drugs used for prophylaxis of migraine headaches also reported an increased risk of nausea and paresthesia with daily use of topiramate compared to placebo.6

Because of the inconsistencies in reporting the adverse events and the small sample size of the few trials that reported them, interpreting the harm numbers warrants caution. For this reason, we have not reported the NNH values in the summary table.

Caveats

The most notable consideration before prescribing topiramate is that in addition to the patient, the patient's primary care physician and/or neurologist ought to be involved in decision-making. Topiramate is a maintenance medication for chronic use. The patient must be willing to take the medication daily, and their outpatient team must be willing to prescribe it. Coordinating care with a primary care physician or a neurologist will ensure that the treatment success and possible adverse events are monitored regularly, refills are ordered, etc. In addition, there are important considerations for women of childbearing age. There is an increased risk of cleft lip and cleft palate among infants exposed to topiramate during the first trimester (relative risk of 21.3, 95% CI:7.9 – 57.1), as compared to the risk in a background population of untreated women.7 It should be noted that palate formation often occurs before patients are even aware of early pregnancy.

While data from the Cochrane systematic review is largely in favor of Topiramate for migraine prophylaxis,6 biases in the individual studies, challenges in the clinical identification of migraine diagnosis, and underreporting and analysis of harms must be considered. FDA warnings regarding birth defects in association with topiramate use and drug costs are also discussed.

The review does not specifically rate the quality of evidence for each outcome. However, it states that the quality of the existing evidence is enough to allow a robust conclusion to be drawn regarding the efficacy of topiramate in preventing migraine attacks. Some of the included trials suffered from biases that could threaten the validity of the findings, of the 17 studies reviewed, nine were judged to have a high risk of bias for at least one item, including allocation concealment, blinding of participants, incomplete outcome data, and selective reporting. Some studies were also underpowered. Another threat to validity was the heterogeneity among the trials, originating from differences in study populations, treatment regimens, and study protocols. As noted in this systematic review, this resulted in increased uncertainty when evaluating the overall effect of topiramate in migraine prophylaxis.6^, 7

Another factor that adds to subjectivity and bias is that the diagnosis of migraine headaches is largely clinical and based on a thorough physical examination and a careful medical history.8^, 9 While some clinical criteria such as those proposed by the International Classification of Headache Disorders are frequently used, the inconsistencies and lack of agreement between clinicians and researchers point to the subjective quality of many of these criteria.8 Thus, in general, trials involving migraine headache tend to suffer from biases originating from imperfect case definition and differential misclassification of exposure. In addition, data from the clinical trials analyzed do not provide sufficient evidence for the use of topiramate in preventing other components of migraine attacks, such as aura symptoms, prodromal symptoms, or chronic migraines.

Summary: The 2012 guidelines from the American Academy of Neurology (AAN) and the American Headache Society supports the use of topiramate and recommends that it be offered for migraine prevention (Level A).4 The European Federation of Neurological Societies guideline has a similar recommendation.10

Considering data from this Cochrane review, a daily dose of 100mg is a reasonable target daily dose.6 Having a discussion of possible adverse events with patients prior to starting a prescription is encouraged. As noted above, a discussion with the patient’s primary care physician or neurologist is also prudent before starting the treatment.

Cost is another factor for many patients when considering a new therapy. As of November 2020, a one-month supply of topiramate (50 mg twice daily) costs $9.00-11.81.11^, 12

Current dosing recommendations of topiramate for migraine prophylaxis approved by the FDA is to start at 25 mg a night for week 1, 25 mg twice a day for week 2, 25 mg in the morning and 50 mg in the evening on week 3, and finally 50 mg both morning and evening starting week 4.12 In patients with significant side effects, rapid discontinuation without tapering is favored.12

In summary, the existing evidence supports the prophylactic use of topiramate to reduce the frequency of headaches in episodic migraine. Although the risk of adverse events was not reported fully, the adverse events that were reported with topiramate use appear to be relatively minor. Therefore, we have rated this treatment Green (Benefits > Harms). This recommendation is consistent with existing practice guidelines from specialty societies. Future trials should focus on the safety of topiramate and better delineate the side effect associated with the use of this drug.




The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Bing Li, MD; Paniz Johari, MS; Carolina Camacho Ruiz, MD
Supervising Editors: Kabir Yadav, MD

Published/Updated

April 5, 2021

Source

Jefferson T, Jones MA, Doshi P, et al. Neuraminidase inhibitors for preventing and treating influenza in healthy adults and children. Cochrane Database Syst Rev. 2014;(4):CD008965.

Study Population: Treatment: Adults and children with confirmed or suspected to have influenza and also those with confirmed or possible exposure to influenza

Efficacy Endpoints

Hospitalization and time to symptom alleviation

Harm Endpoints

Adverse events (e.g. nausea, vomiting, and neuropsychiatric events)

Narrative

Neuraminidase inhibitors (NAIs) are commonly used in the prevention and treatment of influenza. Previous studies and reviews have demonstrated a questionable and modest benefit of their use while demonstrating potential adverse effects.1^, 2

The most recent Cochrane review discussed here3 analyzes the data from randomized controlled trials evaluating the effects of NAIs in children and adults with confirmed or suspected exposure to influenza. The systematic review included trials testing the effectiveness and safety of two commonly used NAIs, oseltamivir (Tamiflu) and zanamivir (Relenza). Twenty-three oseltamivir trials were included, with a total of 9623 subjects (6574 in treatment trials and 3049 in prophylaxis trials) with ages ranging from 1 to 82 years. Twenty-eight zanamivir trials were included, with a total of 14,628 subjects (7678 in treatment trials and 6950 in prophylaxis trials) with ages ranging from 5 to “over 65.” The authors’ novel methodology for study selection was complicated. All of the included trials were industry-supported randomized control trials (RCT) comparing oseltamivir or zanamivir versus placebo, some of which were unpublished reports from manufacturers (hence not peer-reviewed). Studies were excluded if they were not placebo RCT’s, were pharmacokinetic studies, were dose comparison studies, or were ongoing trials.

The authors of the Cochrane systematic review state that they are not convinced they had access to all existing manufacturer data. They report higher risk of bias with frequent lack of reporting of random sequence generation methods, incomplete data (symptoms, complications, and safety data), and concern for lack of full blinding of participants and personnel.3 The Cochrane review included trials that enrolled previously healthy children or adults diagnosed with influenza (with or without symptoms) for the treatment analysis, and similar populations who took NAIs for prophylaxis (with or without exposure).

The primary outcome measures for treatment analysis were: 1. symptom relief, 2. hospitalization, and 3. harms. The primary outcome measures for prophylaxis analysis were: 1. influenza (symptomatic and asymptomatic, with laboratory confirmation) and influenza-like illness (ILI), and 2. hospitalization.

Treatment: In adults, oseltamivir reduced time to first alleviation of symptoms by 16.8 hours (95% CI 8.4 to 25.1). Zanamivir reduced the time to first alleviation of symptoms in adults by 14.4 hours (95% CI, 9.4 to 19.2 hours). In healthy children, based on a single study, oseltamivir reduced the time to first alleviation of symptoms by 29 hours. However, in asthmatic children, there was an increase in time to first alleviation of symptoms by 5.2 hours (95% CI 11.1 hours lower to 21.4 hours higher). Standard asthma medical care and close follow-up may overshadow the incremental benefit of oseltamivir in this specific population. The effect of zanamivir on time to first alleviation of symptoms in children was not statistically significant. Data for asthmatic adults was not reported. Oseltamivir treatment did not reduce the risk of hospitalization in children or adults.

Serious adverse events or those leading to withdrawal from the study were higher in treatment groups. Oseltamivir treatment in adults was associated with nausea (Relative risk [RR]: 1.57, 95% CI 1.14 to 2.15; Absolute risk difference [ARD] 3.7%; Number needed to harm [NNH]: 28) and vomiting (RR: 2.43, 95% CI 1.75 to 3.38%; ARD: 4.56%; NNH: 22). Vomiting was also seen in pediatric studies (RR: 1.70, 95% CI 1.23 to 2.35; ARD: 5.3%; NNH: 19). There was no significant increase in the risk of neuropsychiatric events during treatment.

Oseltamivir reduced self-reported, unverified pneumonia (RR: 0.55, 95% CL 0.33-0.90). The lack of clear pneumonia definitions across studies makes this outcome unreliable. There was no reduction in pneumonia with zanamivir. Neither oseltamivir nor zanamivir reduced incidence of sinusitis, bronchitis, or otitis media.

Prophylaxis: Oseltamivir prophylaxis was initiated on the basis of “local outbreaks” which were not well elucidated. Patients generally took the medication for 6 weeks. Zanamivir prophylaxis was initiated based on population characteristics (e.g. patients in nursing homes) without an indication based on direct exposure or local outbreaks. Patients were treated for 28 days. There were no pediatric prophylaxis studies.

The studies reporting the benefits of NAIs for prophylaxis were of low quality (significant biases), had small sample sizes, and did not clearly define the indications for giving prophylaxis. Therefore, we are refraining from reporting the data for this particular indication here.

Caveats

The use of NAIs for the treatment of influenza in adults confers a small decrease in time to symptom alleviation (17 hours). The statistical heterogeneity for the treatment analysis was reported to be low. The Cochrane review employed the use of clinical study reports from national drug regulators, in addition to studies published in biomedical journals. This resulted in a comprehensive study that better represents the effects of these medications. Although the meta-analysis only included RCT’s, limitations such as selective publication, inclusion of non-peer-reviewed data, and high attrition rates negatively impacted the quality of the evidence. A subgroup analysis on the treatment effect of zanamivir between influenza-positive and influenza-negative patients revealed that both populations had an identical duration of reduction in symptoms. A recent open-label RCT published in January 2020 (the ALIC4E trial4) also showed that patients with symptomatic influenza treated with oseltamivir plus usual care recovered 1 day (95% CI 0.74 to 1.31) sooner than those who received usual care alone. As demonstrated in the Cochrane review, this reduction occurred equally in influenza positive and influenza negative groups, irrespective of laboratory confirmed influenza. This suggests that NAIs most likely do not have an influenza-specific effect in symptom alleviation, or as likely that this was a placebo effect given that the study was unblinded and patients knew if they were receiving an NAI. The use of NAIs in healthy children may reduce time to symptom alleviation by about a day, but carries the risk of adverse events.

The systematic review did not report outcomes where incidence was less than 0.5%, and this included mortality. Lack of clarity regarding mortality is unfortunate, given the estimated 200,000-600,000 estimated annual deaths worldwide due to influenza.5 It is important to note that the Infectious Diseases Society of America (IDSA)6 and the Centers for Disease Control and Prevention (CDC)7 endorse the consideration of NAIs for all people with influenza-like illness within 48 hours of onset. These groups recommend NAIs for all persons at higher risk of dying from influenza (i.e. immunocompromised status, advanced age, multiple comorbidities) and for all persons hospitalized with influenza. These recommendations are based on observational data and expert opinion. Unfortunately, we do not have strong data to support these recommendations, and given their widespread use, it is hard to imagine we will see a randomized control trial in the future, answering the questions “Do NAIs save lives?” or “Is there a benefit with NAIs in people who are very sick (hospitalized)?”

The data for NAIs are further confounded by the financial conflicts of interest that are present in many of the studies on NAI use. Dunn et al8 demonstrated that among NAI studies associated with a FCOI, 88% of the studies were classified as favorable. For those without FCOI (among which was this Cochrane review), 17% of the studies were classified as favorable. Thus, much of the positive data regarding NAI is heavily biased.

In summary, the existing data indicate that NAIs reduce the duration of symptoms by less than a day in patients with confirmed or suspected influenza. The use of NAIs to treat influenza does not prevent hospitalization and is associated with adverse events. Therefore, we have assigned a color recommendation of yellow (unclear if it provides benefit, more data needed) to this treatment.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

See theNNT.com's previous reviews of this topic:
Neuraminidase Inhibitors Given for Influenza, January 26, 2010

Author

Marc Zwillenberg, MD; Eric Tang, MD; Joshua Quaas, MD
Supervising Editors: Kabir Yadav, MD

Published/Updated

Source

Smaill FM, Vazquez JC. Antibiotics for asymptomatic bacteriuria in pregnancy. Cochrane Database Syst Rev. 2019;2019(11):CD000490

Study Population: Pregnant women in all three trimesters of pregnancy in 12 studies (N=2,017)

Efficacy Endpoints

Rates of pyelonephritis, preterm birth, low birth weight

Harm Endpoints

Maternal side effects

Narrative

Asymptomatic bacteriuria, occurring in 2-15% of pregnancies, is generally defined as at least one urine culture showing >100,000 colony-forming units (CFUs)/mL in the absence of fever or symptoms of urinary tract infection.1 Escherichia coli is the most commonly associated pathogen, comprising up to 80% of isolates.2 While asymptomatic bacteriuria in non-pregnant women is generally benign, in pregnant women there is an increased likelihood for progression to pyelonephritis likely due to mechanical compression of the ureters by an enlarged uterus, as well as smooth muscle relaxation induced by progesterone.3 Some studies suggest if asymptomatic bacteriuria is left untreated, up to 30% of pregnant women will develop acute pyelonephritis,4 which may be associated with potentially serious maternal complications such as sepsis,5 and pregnancy outcomes such as low birth weight and preterm birth.6

The systematic review summarized here included 2,017 pregnant women from twelve randomized controlled trials that compared antibiotic treatment of asymptomatic bacteriuria to placebo or no treatment.7 Most of the women were enrolled through prenatal screening in hospital-based clinics, and the studies included women in all stages of pregnancy. The definition of pyelonephritis varied across studies but mainly encompassed women with flank tenderness, fever, with or without urinary symptoms (such as frequency, dysuria, and hematuria), and >100,000 CFUs/mL of urine. Incidence of pyelonephritis ranged from 2.2% to 36%. Included studies were performed in the 1960s through the 1980s, and study antibiotics included sulfa drugs, tetracycline, methenamine, nalidixic acid, nitrofurantoin, and ampicillin. Duration of treatment varied including single dose, short course (three to seven days), intermediate course (three to six weeks), and continuous antibiotic until delivery.

As compared to placebo or no treatment, [in 11 studies of 1932 subjects] antibiotic treatment reduced the incidence of pyelonephritis (Relative risk [RR]: 0.2, 95% confidence interval [CI] 0.1 to 0.4; Absolute Risk Difference [ARD]: 15%; Number-Needed-To-Treat [NNT]: 7). Data from three studies (n=327) also found antibiotic treatment reduced the incidence of preterm birth (gestational age <37 weeks) [RR 0.3, 95% CI 0.1 to 0.9; ARD: 11%; NNT: 9]. Finally, data from six studies (N=1437) found antibiotics also decreased the incidence of birth weight <2500g [RR 0.6, 95% CI 0.5 to 0.9; ARD: 5%; NNT: 20].

Caveats

The findings of this review suggest a potentially important, clinically meaningful benefit with the use of antibiotics for asymptomatic bacteriuria in pregnancy. The underlying data, however, are heterogeneous, rife with potential systematic bias, and mostly generated 40-60 years ago.

The wide variation in pyelonephritis incidence among studies (2-36%), for instance, may partially originate from the variability in definitions of pyelonephritis, or patient level characteristics such as infecting organisms, socioeconomic status, and prenatal care. Regardless of source, this variation in baseline risk introduces major clinical heterogeneity and reduces the applicability of results. Moreover, the lack of consistent blinding (four studies were doubleblinded) and the significant heterogeneity across studies led the review authors to judge the quality of evidence as low. It is, in addition, difficult to confidently attribute the development of pyelonephritis to asymptomatic bacteriuria. Similarly, data on the incidence of preterm birth and low birth weight were also deemed low quality, again due to lack of blinding, heterogeneity, and small sample size.

Another important weakness is the lack of reporting on maternal side effects, making it impossible to characterize the potential for adverse events in the treatment group.

In summary, although antibiotics may reduce the risk of pyelonephritis in pregnancy, as well as preterm birth and low birth weight, the evidence for treatment of asymptomatic bacteruria is low quality. More robust research is needed, particularly in a contemporary milieu, using standardized treatment protocols, and stratification for baseline risk. Despite limitations, however, the data suggest significant benefit in preventing pyelonephritis, preterm birth, and low birth weight. Based on these findings it seems likely benefits outweigh harms. It should be emphasized that this signal of benefit is based upon a definition of asymptomatic bacteriuria that requires a positive urine culture, not simply a suggestive urinalysis. We would expose many more pregnant women to unclear harms for reduced benefit if we do not adhere to the clinical definition of asymptomatic bacteriuria. Therefore, we have rated antibiotic treatment for culturepositive asymptomatic bacteriuria during pregnancy Green (Benefit>Harms).

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Kelvin Kwofie, MD; Allan B. Wolfson, MD
Supervising Editors: Kabir Yadav, MD

Published/Updated

Source

Guay J, Kopp S. Peripheral nerve blocks for hip fractures in adults. Cochrane Database Syst Rev. 2020;11:CD001159.

Study Population: 2,750 adults admitted to the hospital with acute hip fracture awaiting operative repair in 23 countries, mean age range 59-89

Efficacy Endpoints

Pain on movement 30 minutes after block, time to first mobilization after surgery, delirium, myocardial infarction, chest infections, and all-cause mortality.

Harm Endpoints

No harm endpoints reported

Narrative

Between 2000 and 2010, there were 2.6 million hospitalizations for hip fractures in the United States among women over 55 years of age alone.1 This is nearly as many hospitalizations as those for myocardial infarction or stroke among that same population and outnumbered hospitalizations for all other fractures combined.1 Studies have also demonstrated an increased risk for mortality in the months after hip fracture.2 While pain from hip fractures is often treated with intravenous opioids, these agents can lead to respiratory depression and delirium.3^, 4 Peripheral nerve blockade (PNB) is a method of injecting local anesthetic into nerves that supply the hip region in an attempt to reduce pain. This alternative method of pain control may reduce the quantity of opioids needed,5^, 6 improve pain and mobility,7 and reduce complications of immobilization.6

The Cochrane review summarized here included all parallel randomized controlled trials (RCTs) and cluster RCTs comparing PNB versus sham block or no block among adult patients (age >16) with hip fracture.8 PNBs were mostly femoral nerve (22 studies) or fascia iliaca compartment blocks (21 studies) and could be performed preoperatively, intraoperatively, or postoperatively. Outcomes summarized here included pain on movement 30 minutes after block placement, delirium within 30 days, time to mobilization after surgery, and chest infection (as defined by the individual study authors).

The review identified 49 trials, 43 of which (n=2,750) were pooled for meta-analysis. Only 15 used ultrasound guidance, while 14 used a nerve stimulator. The remaining studies used a landmark-based technique or did not describe the technique. Sixteen studies were performed in the Emergency Department. PNB reduced pain within 30 minutes of block placement (standardized mean difference: -1.0; 95% confidence interval [CI]: -1.3 to -0.9; equivalent to -2.5 on a scale from 0 to 10; quality of evidence high) when compared to no block or sham. PNB also reduced delirium (risk ratio [RR]: 0.7; 95% CI: 0.5 to 0.9; absolute risk reduction [ARR]: 7.3%; number-needed-to-treat [NNT]: 14; quality of evidence high). Three studies reporting on chest infections found a reduced incidence with PNB (RR: 0.4; 95% CI: 0.2 to 0.9; ARR: 15.9%; NNT: 7; quality of evidence moderate). Three studies reporting time to mobilization also found a reduction (mean difference: -10.8 hours; 95% CI: -12.8 to -8.8 hours).

Caveats

There are several limitations associated with this meta-analysis. First, studies varied with regard to PNB technique and location, with the majority using either the fascia iliaca or femoral nerve block. Additionally, there were differences in choice of agent (e.g., bupivacaine, ropivacaine, lidocaine), concomitant use of epinephrine or corticosteroids, and single injection versus continuous infusion. Most blocks were performed by anesthesiologists in the peri-operative setting. Some studies used a sham block, while others did not use a placebo. There were differences in definitions for delirium and chest infections, and pain scoring differences between studies. There were also limited data for chest infections with only 131 total patients for this outcome. Moreover, they did not assess adverse events which is a critical gap. Data suggest some adverse effects are rare, with permanent nerve injury occurring in 0.03% of PNBs,9 and local toxicity occurring in 1.3 per 10,000 PNBs.10 However, local reactions, tissue infections, and other adverse effects were not reported. Finally, it is worth noting that most studies were small and performed predominately by experts in regional analgesia, which may not fully reflect the average user.

The American Academy of Orthopedic Surgeons recommends regional analgesia for preoperative pain control in patients with hip fracture (strong evidence).11 Based on the available evidence, the review summarized here found PNB reduced pain on movement, shortened time to first mobilization, and resulted in lower rates of delirium and chest infections. Thus, we have assigned a color recommendation of Green (benefit>harm) for PNB for hip fracture. Further study is needed to evaluate PNB in periods and settings other than the inpatient peri-operative period, and the potential benefit of continuous infusion versus single injection. However, current data support that this would be a valuable intervention for hip fractures that could be utilized in the Emergency Department.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 1, 2021

Source

Appelen D, van Loo E, Prins MH, Neumann MH, Kolbach DN. Compression therapy for prevention of post‐thrombotic syndrome. Cochrane Database Syst Rev. 2017; 9(9): CD004174.

Azirar S, Appelen D, Prins MH, Neumann MH, de Feiter AN, Kolbach DN. Compression therapy for treating post‐thrombotic syndrome. Cochrane Database Syst Rev. 2019; 9(9): CD004177.

Study Population: Prevention: five trials comprising 1,393 patients with deep vein thrombosis. Treatment: four trials comprising 116 patients with post‐thrombotic symptoms

Efficacy Endpoints

Prevention of postthrombotic syndrome (prevention endpoint); reduction in severity of postthrombotic symptoms (treatment endpoint)

Harm Endpoints

Discomfort, pain, and pressure sores (prevention harm endpoint); discomfort, skin damage, pain, arterial leg ulceration, recurrence of thrombosis (treatment harm endpoint)

Narrative

Postthrombotic syndrome (PTS) arises from venous outflow restriction.1^, 2 It may occur in 20% to 50% of patients after a deep vein thrombosis (DVT) and is characterized by edema, skin changes, pruritus, paresthesias, and pain, which can adversely affect quality of life (QOL).1^, 2 Adequate dosing and duration of anticoagulation for DVTs may lower the likelihood of PTS3; however, compression therapy has also been used for prevention and treatment.4^, 5^, 6 While this therapy, which includes bandaging or compression stockings, has been proposed to reduce edema and improve QOL,6^, 7 clinicians, articles, and guidelines differ in their recommendations concerning the use of these devices for prevention and treatment of PTS.1^, 2^, 3^, 4^, 5^, 6 This article discusses two Cochrane reviews evaluating prevention and treatment of PTS.5^, 8

Prevention: The systematic review evaluating prevention of PTS included randomized controlled trials (RCTs) investigating compression therapy in patients with DVT, diagnosed by ultrasonography or venography.5 The primary outcome was the incidence of PTS up to 2 years after diagnosis, which was defined by the individual study authors. Secondary outcomes included venous thromboembolism, adverse effects (discomfort, pain, and pressures sores), patient satisfaction and QOL, and compliance.

The authors initially included 10 RCTs (n = 2,361 patients) with mean follow‐up ranging from 2 to 6.3 years. However, due to differences in intervention types and inadequate data, they were only able to pool five trials (n = 1,393 patients) for meta‐analysis. All trials compared elastic compression stockings (pressure 20–40 or 30–40 mm Hg at the ankle) versus no intervention. The stockings were not statistically associated with reduction in incidence of PTS (relative risk [RR] = 0.62, 95% confidence interval [CI] = 0.38 to 1.01), reduction in incidence of severe PTS (RR = 0.78, 95% CI = 0.53 to 1.15), or difference in DVT recurrence (RR = 0.94, 95% CI = 0.69 to 1.28). There was no difference between knee‐high and thigh‐high stockings (RR = 0.92, 95% CI = 0.66 to 1.28). Three studies reported patient satisfaction and QOL, with one of these trials finding more rapid and pronounced improvement in QOL with compression compared to bed rest. No trials described serious adverse events.

Treatment: The systematic review evaluating compression dressings for reducing the severity of PTS in patients with objectively diagnosed DVT included four trials (n = 116 patients).8 The primary outcome was severity of PTS as defined in each study (based on a systematic clinical history and scoring of examinations) and adverse effects including recurrence of thrombosis, discomfort, skin damage, pain, and arterial leg ulceration.

Because of the presence of methodologic and endpoint heterogeneity among the studies, the authors of the systematic review elected not to pool data to perform a meta‐analysis. Among the four trials reviewed, two evaluated gradual elastic compression stockings,9^, 10 and two evaluated intermittent compression devices.11^, 12

Individual Trials:
Gradual elastic compression
One trial compared gradual elastic compression stockings (40 mm Hg pressure at the ankles) to placebo stockings in patients with PTS 1 year after DVT diagnosis.9 The endpoint was treatment failure, defined as pain and swelling that did not improve, or worsened, after the first 3 months; worsening of symptoms during further follow‐up; symptoms preventing participants from performing their daily activities for 5 or more days in any 3‐month period; or developing a leg ulcer. This study did not show any significant difference in treatment failure between groups, and it did not assess satisfaction or QOL. Since the prevalence of PTS was very low the investigators concluded PTS is rare and compression dressings were not necessary.9 Another trial evaluated the hemodynamic performance of medical elastic compression devices based on a symptom scale and air plethysmography, with each patient acting as his/her own control.10 The measured endpoints were venous filling index, venous filling time, and venous volume. In this study, all types of compression dressings significantly improved the hemodynamic endpoints compared with no compression.10 This study found that 21 patients changed their preference of leg stocking, with eight of these 21 patients preferring above‐the‐knee stockings.10 Neither trial reported adverse events.9^, 10

Intermittent compression
One trial evaluated the Jobst extremity pump (15 mm Hg in the control group vs. 50 mm Hg pressure in the active treatment group) in 15 patients, with treatment considered successful if patients had improved symptom score and preferred the therapeutic pressure.11 The second trial evaluated the Venowave device, a lower‐limb venous return device.12 They performed a placebo‐controlled, double‐blind, crossover RCT, with a primary outcome of at least moderate improvement in symptoms, and willingness to continue the device. The first trial found 80% of patients (11/15) preferred higher pressure compression,11 with symptom scores at therapeutic levels of 14.3 (standard deviation [SD] ±5.4) compared to 16.5 (SD±6.1) for placebo levels. The second study found improved scores and QOL based on VEINES‐QOL score (mean difference = 2.9 points) in patients utilizing the Venowave device.12 This latter trial also found 9% of patients experienced an adverse event such as leg swelling, irritation, superficial skin bleeding, or pruritus.12

Caveats

Both of these systematic reviews contain limitations.5^, 8 In the Cochrane review evaluating prevention of PTS, included studies demonstrated no significant effect although considerable heterogeneity was present due to variation in follow‐up, time following DVT, blinding, size of compression stockings, pressure and length of stockings, and scoring systems utilized for PTS assessment. It is notable that pooled results found endpoints suggesting the potential for clinically important effects, although the trials may have been underpowered to detect them.

The Cochrane review evaluating treatment of PTS found very‐low‐certainty evidence concerning graduated elastic compression stockings, with one study reporting beneficial hemodynamic effects and one finding no benefit on severity of PTS.10^, 11 Authors of the Cochrane review did not perform quantitative meta‐analysis due to significant heterogeneity, instead presenting the evidence narratively.8 We applaud this choice, which is why we elected to provide a summary of each trial above. The authors’ narrative description of each study and their findings seems far more likely to be helpful to patients and clinicians than any single numerical result hoping to represent such a wide variety of methods and interventions.8 Included studies utilized different methodologies and evaluated different devices, populations, symptom severity scales, and outcomes.9^, 10^, 11^, 12 Symptom scoring systems and classification systems varied in the included studies. This lack of standardization is a significant limitation in comparing and contrasting study results, and several studies did not assess patient‐centered outcomes such as QOL. Most also evaluated primary outcomes at less than 5 months. Because PTS is a chronic condition, evaluation of therapies for longer time periods may be needed to demonstrate sustained efficacy. Sample sizes were also small, with 116 subjects in total. One included trial used a non–patient‐centered endpoint.10

The two systematic reviews summarized here highlight the paucity of reliable data in supporting or refuting the effectiveness of compression therapy for preventing or reducing the symptoms of PTS.5^, 8 Based on the presented data, authors rate the certainty of evidence for prevention as low, and the evidence for treatment of PTS symptoms as very low or low. We agree and have thus assigned a color of yellow (unclear if benefits) for the use of elastic compression stockings for prevention and treatment of PTS symptoms. Further studies are needed using validated and uniform symptom scales, which also evaluate long‐term benefits and adverse events.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

February 3, 2021

Source

Derry CJ, Derry S, Moore RA. Sumatriptan (all routes of administration) for acute migraine attacks in adults - overview of Cochrane reviews. Cochrane Database Syst Rev. 2014;2014(5):CD009108.

Study Population: 9365 adults with active migraine

Efficacy Endpoints

Pain-free or headache relief at one and two hours

Harm Endpoints

Fatigue, malaise, dizziness, vertigo, nausea, vomiting, taste disturbance, chest pain, sweating, numbness, paresthesias, drowsiness, flushing, back pain

Narrative

Migraines are common and most often described as painful pulsatile headaches with nausea and sensory sensitivities. Sumatriptan is one of many medications available for treatment of acute migraine attacks.

The Cochrane systematic review discussed here1 evaluated sumatriptan for acute migraines administered subcutaneously (the most studied and commonly utilized route of administration). Studies included adults with migraine, used single-dose sumatriptan, and were randomized, double-blinded, and placebo-controlled. The overall pain and headache pain were assessed using standard pain scales. The total number of participants in 35 included studies was 9365.

Outcomes included ‘pain-free’, ‘headache relief’, and adverse events within 24-hours. Pain-free indicated participants initially with moderate or severe pain experiencing complete resolution. Headache relief indicated moderate or severe pain with subsequent reduction to mild or no pain.

More patients were pain-free at one hour with 4mg subcutaneous sumatriptan than placebo (relative risk [RR] 4.7; 95%CI, 2.8-7.7; absolute risk difference [ARD] 27%; Numberneeded-to-treat [NNT] 4. For 6mg sumatriptan dosing RR was 5.6; 95%CI, 4.6-6.8; ARD 34%; NNT 3). Similar results were found at two hours. Sumatriptan was also superior to placebo for headache relief at one hour with 4mg (RR 2.6; 95%CI, 2.0-3.2; ARD 41%; NNT 3) and 6mg dosing (RR 2.7; 95%CI, 2.5-2.9; ARD 45%; NNT 3). Again, two hour results were similar.

Subcutaneous sumatriptan administration increased adverse events (RR 1.8, 95%CI, 1.6 to 2.2; ARD 30%; number-needed-to-harm [NNH] 3 for a single subcutaneous 4mg dose, and NNH 4 for 6mg dose). Treatments were however described as well-tolerated with most adverse events mild or moderate in severity and self-limiting. Events included fatigue, malaise, dizziness, vertigo, nausea, vomiting, taste disturbance, chest pain, sweating, numbness, paresthesias, drowsiness, flushing, and back pain.

Caveats

Authors of the Cochrane reviews deemed the quality of the evidence largely high and the results reliable. Moreover, additional reviews suggest similar benefits using other routes of administration including oral and intranasal, helping to confirm the consistency of these findings.1 However, other routes also were associated with increased adverse events (except for oral 25mg dose) with NNH of 13 for 50 mg oral, 5 for 100 mg oral, and 4 for 20 mg intranasal.

One important source of potential bias is publication bias, in that unpublished data could theoretically alter the outcomes and conclusions of this review, a concern unfortunately confirmed in the setting of other pharmaceuticals developed prior to the advent and common use of trial registries.2 In regards to adverse events, the review found high variability and often poor quality in assessing and reporting.

Of note, in the current milieu it is unusual to see placebo controls in studies of active migraine. For instance, rigorous comparison finds metoclopramide, a common, safe, inexpensive antiemetic, equivalent or superior to subcutaneous sumatriptan for acute migraine.3 In addition, while apparently safe in the setting of carefully screened trial populations, in higher risk patients triptan medications may cause rare serious, potentially fatal adverse events such as coronary vasospasm, myocardial infarction, and ischemic colitis, though much of the evidence is derived from animal models and case reports.4^, 5^, 6 For this reason, sumatriptans are generally contraindicated in patients with known ischemic heart disease, angina pectoris, or poorly controlled hypertension or renal disease.7 The widespread availability of safer, less costly, equally effective agents may help explain why sumatriptan is rarely used in acute care settings where acute migraine patients are typically treated.8 For instance, rigorous comparison finds metoclopramide, a common, safe, inexpensive antiemetic, to be equivalent or superior to subcutaneous sumatriptan for acute migraine.3 Sumatriptan may still hold benefit in ease of administration (subcutaneous vs intravenous).

In summary, existing evidence supports subcutaneous sumatriptan over placebo for treatment of migraine headache. Adverse events occur at a similar rate to benefits, but appear mild and well-tolerated. Therefore, we have assigned a color recommendation of Green (Benefits > Harms) for subcutaneous sumatriptan compared to placebo. However, given the availability of other abortive medications for migraines with even safer side effect profiles, sumatriptan does not necessarily serve as first-line treatment for some clinicians.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Christopher Lim, MD; Manpreet Singh, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Naqvi IA, Kamal AK, Rehman H. Multiple versus fewer antiplatelet agents for preventing early recurrence after ischaemic stroke or transient ischaemic attack. Cochrane Database Syst Rev. 2020;8(8):CD009716.

Study Population: 10739 patients with atherothrombotic ischemic stroke or transient ischemic attack from 5 trials

Efficacy Endpoints

Stroke recurrence and vascular death. Secondary outcomes included myocardial infarction, vascular death, and death from all causes

Harm Endpoints

Intracranial and extracranial hemorrhage

Narrative

Ischemic strokes range in severity from minor to debilitating. Minor strokes and transient ischemic attacks (TIA) may be followed by recurrent strokes, with the highest risk in the first 48 hours.1 Approximately 30% of strokes are recurrent.2 Antiplatelet agents may reduce the risk of recurrence and prevent disability, but may also increase the risk of hemorrhage.2^, 3^, 4 The purpose of this summary is to update and refine a prior summary examining the effect of adding clopidogrel to aspirin after atherothrombotic acute ischemic stroke or TIA.

The Cochrane review from which we extracted this information included randomized controlled trials (RCTs) evaluating patients taking any combination of multiple antiplatelet agents versus fewer agents within 72 hours of an atherothrombotic acute ischemic stroke or TIA and followed up for at least one month.5 The primary outcome was stroke during follow up of at least 3 months. Secondary outcomes included myocardial infarction, intracranial hemorrhage, extracranial hemorrhage, and death. When there was more than one follow-up period, the authors included outcomes at one week, one month, three months, and six months. Here we focus on the comparison of aspirin plus clopidogrel versus aspirin alone.

The meta-analysis authors included 5 RCTs of 10739 subjects comparing aspirin plus clopidogrel versus aspirin alone.3^, 6^, 7^, 8^, 9 All medications were administered orally, and most utilized a loading dose of clopidogrel 300 mg. Dosing of aspirin ranged from 75-300 mg. Follow-up time points ranged from 30 days to 1 year.

Dual antiplatelet therapy with aspirin plus clopidogrel was associated with less recurrent stroke when compared to aspirin alone (6.5% versus 9%; absolute risk reduction [ARR]: 2.5%; number needed to treat [NNT]: 40; risk ratio [RR]: 0.7; 95% confidence interval [CI]: 0.6-0.8). There was no difference in vascular death (RR 1.4; 95% CI: 0.6-2.9) or myocardial infarction (RR 1.5; 95 CI: 0.6-3.4). Extracranial hemorrhage was higher with aspirin plus clopidogrel (1.4% versus 0.3%; absolute risk increase: 1.1%; number needed to harm [NNH]: 91; RR 4.8; 95% CI: 2.2-10.6), while intracranial hemorrhage was not statistically different (RR: 1.3; 95% CI: 0.6-2.9). The risk of hemorrhage significantly increased after 3 months of therapy.

Caveats

There are several limitations associated with this meta-analysis. A single trial conducted in China (CHANCE) accounted for the majority of data concerning aspirin plus clopidogrel versus aspirin alone.3 However, exclusion of this study did not significantly change the primary or secondary outcomes. Several studies report data from before 2010, and stroke care has significantly changed since this period.4 Only three studies reported intracranial hemorrhage data, and just two reported extracranial hemorrhage data.3^, 6^, 7^, 8^, 9 Four trials included patients with TIA or non-disabling stroke (defined by a The National Institutes of Health Stroke Scale ≤3), while one included strokes regardless of severity.3^, 6^, 7^, 8^, 9 Other adverse outcomes such as myocardial infarction were also reported in only a few studies. Another significant consideration is that the results apply to atherothrombotic and not cardioembolic strokes. Finally, duration of therapy is an important consideration. Included patients received at least 1 month of antiplatelet therapy, and follow up was 3 months in most studies.

Broadly, the Cochrane review summarized here found a strategy of multiple antiplatelet agents initiated within 72 hours of the event compared to a single agent reduced stroke in the short term, while increasing major hemorrhage.5 This finding is consistent with results from trials adding clopidogrel to aspirin.3^, 6^, 7^, 8^, 9 Based on this evidence, we have assigned a color recommendation of Green (benefit >harm) for the use of aspirin plus clopidogrel versus aspirin alone after atherothrombotic ischemic stroke or TIA. While further study is needed, the CHANCE trial suggests a duration of 21 days to 1 month is appropriate.3 Further data are needed to better evaluate adverse events, cardioembolic events, and specific durations of therapy.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Desborough MJ, Oakland K, Brierley C, et al. Desmopressin use for minimising perioperative blood transfusion. Cochrane Database Syst Rev. 2017;7(7):CD001884.

Study Population: Patients with normal platelets: 25 trials (1806 patients) of DDAVP versus placebo evaluating red cell transfusion, 22 trials (1631 patients) evaluating mortality, and 29 trials (1984 patients) evaluating thrombotic events

Patients with abnormal platelets: 5 trials (258 patients) of DDAVP versus placebo evaluating red cell transfusion, 7 trials (422 patients) evaluating mortality, and 7 trials (422 patients) evaluating thrombotic events

Efficacy Endpoints

Transfusion requirement and all-cause mortality

Harm Endpoints

Thrombotic events

Narrative

Desmopressin (DDAVP) is a synthetic analogue of vasopressin that stimulates release of von Willebrand factor and promotes platelet adhesion and aggregation, and is therefore typically used for treatment of von Willebrand disease and hemophilia1^, 2^, 3 Some professional societies, however, recommend perioperative DDAVP for the prevention of blood loss and blood product transfusion in patients with bleeding and platelet dysfunction.4^, 5^, 6

The Cochrane Review discussed here included randomized controlled trials (RCTs) evaluating adult and pediatric patients with or without platelet dysfunction who received subcutaneous or intravenous perioperative DDAVP.7 Platelet function was defined as prolonged bleeding time, abnormal Platelet Function Analyzer closure times, or taking antiplatelet medications. Studies including patients with known hemophilia or von Willebrand disease were excluded. Primary outcomes included the number of patients receiving red blood cell transfusion during the procedure or within 30 days, volume transfused, and blood loss in milliliters. Secondary outcomes included all-cause mortality within 30 days, thrombotic events, reoperations due to bleeding, bleeding events during or within 30 days, hypotension within 30 days, and quality of life.

The authors identified 65 RCTs (n = 3874) that met inclusion criteria. For this summary, we focused on studies evaluating DDAVP versus placebo or no treatment in patients with and without platelet dysfunction. Included studies examined DDAVP in one or two doses of 0.2-0.4 micrograms/kilogram, 20 micrograms once, or 15-45 micrograms based on weight. Timing of administration varied, including preoperative, operative, or postoperative. Settings included adult and pediatric cardiac surgery, orthopedic surgery, vascular surgery, plastic surgery, hepatic surgery, kidney biopsy, and maxillofacial surgery.

Overall, there were no significant differences in efficacy endpoints such as transfusion requirements, all-cause mortality, or thrombotic events (quality of evidence: Low). These results remained the same when analysis was repeated in the subgroup of patients with platelet dysfunction.

In addition to thrombotic events, the Cochrane review reports on what the authors term ‘clinically important hypotension’, which was more common in patients receiving DDAVP (RR: 2.9; 95% CI: 1.3 to 6.3; ARD: 5.9%, NNH: 17). Unfortunately, the Cochrane review did not define clinically important hypotension, and definitions were variable and often arbitrary in the original trials. Therefore, we did not list this endpoint in the summary table.

Caveats

There was variable quality of evidence for the included studies, preventing strong conclusions. Evidence quality was moderate for blood transfusion requirement in patients with normal platelet function, but low to very low for all other outcomes and populations. As a result, the authors of the review were unable to pool the data for volume of blood transfused and blood loss. Additionally, mortality and thrombotic events were rare, and studies were therefore underpowered to assess these outcomes. The patient populations were heterogeneous with regard to age, comorbidities, and DDAVP dosing. The most common studied groups were adult cardiac surgery and orthopedic surgery, accounting for 51 trials, with under-representation of other surgeries. The majority of trials were conducted over two decades ago and may not reflect current settings. Less than one third of trials included patients receiving DDAVP preoperatively, and there were no trials evaluating patients with thrombocytopenia or coagulopathy. Finally, most of the included trials did not provide sufficient information for assessment of risk of bias.

When compared to placebo, DDAVP is associated with no clinical difference in patients with normal or abnormal platelet function in red blood cell transfusion, all-cause mortality, or thrombotic complications. Based on this evidence, we have assigned a color recommendation of Red (No benefit). Further data, however, are needed to better evaluate the role of DDAVP in current settings, those with coagulopathies or on antiplatelet medications, and administration based on platelet function or viscoelastic tests.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

CRASH-3 trial collaborators. Effects of tranexamic acid on death, disability, vascular occlusive events and other morbidities in patients with acute traumatic brain injury (CRASH3): a randomised, placebo controlled trial. Lancet. 2019;394(10210):1713-1723.

Yokobori S, Yatabe T, Kondo Y, Kinoshita K; Japan Resuscitation Council (JRC) Neuroresuscitation Task Force and the Guidelines Editorial Committee. Efficacy and safety of tranexamic acid administration in traumatic brain injury patients: a systematic review and meta-analysis. J Intensive Care. 2020;8:46.

Study Population: 7 trials comprising 10,044 patients with traumatic brain injury

Efficacy Endpoints

Mortality, poor neurologic outcome, hemorrhagic complications

Harm Endpoints

Ischemic or thromboembolic complications

Narrative

Traumatic brain injury (TBI) is a significant cause of morbidity and mortality. In 2014, there were approximately 2.5 million emergency department visits for TBI in the United States, with 56,800 deaths attributed directly to TBI.1 Among survivors it is estimated millions more live with long-term disability.2 While one large study demonstrates a mortality benefit (NNT=67) for tranexamic acid (TXA) in multi-system trauma,3 no primary outcome benefit was demonstrated for TXA in the WOMAN trial4 of post-partum hemorrhage or in the HALT-IT trial5 of gastrointestinal bleeding. Here we summarize results from the CRASH-3 trial,6 as well as from a systematic review combining this trial with data from 6 smaller trials of TXA for TBI.7

CRASH-3 was a multinational randomized placebo-controlled trial enrolling 12737 subjects with TBI, with a mean age of 42 years.6 Included patients suffered TBI <8 hours prior to potential administration of study drug, and had either a Glasgow Coma Scale (GCS) score ≤12 or intracranial hemorrhage on brain imaging, with no major extracranial bleeding. Patients were randomly assigned to either TXA 1g intravenously over 10 minutes followed by 1g over 8 hours, or placebo infusion. The author-reported primary outcome was head injury-related in-hospital mortality within 28 days of injury among the subgroup receiving treatment within 3 hours. Secondary outcomes included, among others, all-cause mortality, disability, vascular occlusive events, seizures, neurosurgical interventions, and other complications.

For the outcome of head injury-related mortality in patients treated within 3 hours, there was no significant difference, with a mortality rate of 18.5% in the tranexamic acid group and 19.8% in the placebo group (Relative risk [RR]: 0.94 95% CI 0.86 to 1.02). There was also no difference in the key secondary outcomes of neurologic or functional disability and all-cause mortality. For subjects receiving TXA within 3 hours the authors do, however, report a lower risk of head injury-related death among those with mild or moderate head injury. This is based on post-hoc subgroup analysis (5.8% versus 7.5%; RR: 0.8; 95% confidence interval [CI]: 0.6 to 0.95, Absolute risk difference [ARD]: 1.7%; Number needed to treat [NNT]: 59). Risk of vascular occlusion was not different between groups.

A systematic review incorporating these data and six additional randomized trials (total n=10,044) further compares TXA with placebo or no intervention for patients with TBI.7 This systematic review included 9202 subjects from CRASH-3 who received TXA within 3 hours of injury and 842 subjects from additional trials administering TXA within 8 hours. The primary outcome was mortality, while secondary outcomes included poor neurologic outcome, ischemic or thromboembolic complications, and hemorrhagic complications.

The mean age of participants in the included studies ranged from 32-42 years, and 9-25% were women. There were no differences between TXA and control groups in mortality, neurologic outcome, ischemic or thromboembolic complications, or hemorrhagic complications.

Caveats

The CRASH-3 trial is the largest to date evaluating the utility of TXA in patients with TBI.6 CRASH-3 and a separate meta-analysis of seven trials both find no mortality or disability benefit from administering TXA in TBI patients.6^, 7 The authors of CRASH-3, however, assert a mortality benefit in subgroups based on TBI severity and recommend that patients receive TXA within 3 hours of injury to reduce “head injury-related death,” but there are several significant issues with this interpretation.6 First, the authors executed more than half of CRASH-3 with a primary outcome of all-cause mortality for those treated within 8 hours, but in 2016 they changed their protocol. This change resulted in head injury-related mortality replacing all-cause mortality as the primary endpoint, and the authors excluded those treated after 3 hours from their new primary outcome. Therefore, more than 3500 enrolled subjects were removed from their primary analysis.

This transition has other considerations as well. All-cause mortality is a more patient-centered outcome. Additionally, head injury-related death is not specifically defined, and deaths in the period immediately following an isolated TBI could all be considered head injury-related. Randomization and blinding should neutralize any biasing effect of ostensibly non-head injury deaths, though an exception would be if drug-related deaths occur. Thus, all-cause mortality is critical, because using head injury-related deaths may remove deaths potentially due to TXA, such as vaso-occlusive fatalities. By our calculations from the data provided in the supplementary online appendix of CRASH-3,6 we noted a statistically significant increase in combined fatal stroke, myocardial infarction, and pulmonary embolism (the adverse event measure reported in CRASH-2),3 in the TXA group (RR: 2.0, 95% CI: 1.04 to 3.7). These deaths, however, were removed from the CRASH-3 primary outcome analysis by the change to ‘head injury-related death’.

Third, CRASH 3 posit a subgroup benefit with TXA in patients with GCS 9-15 (i.e., mild-tomoderate head injury). Based on their 2018 statistical analysis plan paper,8 they planned to evaluate three discrete severity subgroups: mild, moderate, and severe. However, in their 2019 paper, they combine mild and moderate into one subgroup.6^, 8 Moreover, in their statistical analysis paper the investigators state they expected outcomes with TXA not to vary based on severity, and thus select a heterogeneity calculation threshold of p <0.001 for presenting their subgroup analysis results.8 The 2019 paper then reports a p-value of 0.03 for this calculation, suggesting that according to their protocol they should not have presented these subgroup analyses.6 Finally, subgroup analyses reduce external validity in general, create challenges in application, and should be considered hypothesis generating at best. In this case, it is particularly unreliable as the primary analysis found no benefit.

The systematic review and meta-analysis that incorporated CRASH-3 also has several important limitations.7 First, CRASH-3 represents 92% of the total number of patients included in the analysis, but uses a different enrollment window (0-3 hours) than the other 6 trials. Additionally, GCS score varied between studies, and there was limited description of severity or mechanism of injury. Moreover, many patients had polytrauma, and it is unclear the degree to which this contributed to death. Finally, the authors did not assess important complications such as seizures or cognitive deficits.

A more recent RCT released in September 2020 included out-of-hospital patients with TBI with GCS of 12 or less and a systolic blood pressure of greater than 90 mm Hg.9 Of the 966 patients included, TXA administration within 2 hours of injury compared with placebo did not improve 28-day mortality or 6-month neurologic outcome.9

Based on the above data, TXA does not show any significant benefit in patients with TBI. Administration of TXA after TBI also does not seem to offer any benefit in disability or neurologic functional outcomes. In addition, there may be a small increase in the rate of thromboembolic fatalities associated with this treatment. Therefore, we have assigned a color recommendation of Red (No benefit) to this intervention.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

December 1, 2020

Source

Andrade-Castellanos CA, Colunga-Lozano LE, Delgado-Figueroa N, GonzalezPadilla DA. Subcutaneous rapid-acting insulin analogues for diabetic ketoacidosis. Cochrane Database of Syst Rev. 2016;Issue 1. Art. No.: CD011281.

Study Population: 201 participants enrolled in five randomized controlled trials that compared continuous intravenous regular insulin with subcutaneous rapid-acting insulin analogs

Efficacy Endpoints

Time to resolution of ketoacidosis, episodes of hypoglycemia, and all-cause mortality

Harm Endpoints

Hypoglycemia

Narrative

Diabetic ketoacidosis (DKA), characterized by hyperglycemia, metabolic acidosis, and ketosis, is a serious complication of diabetes mellitus. Treatment focuses on fluid restoration, correction of hyperglycemia, and inhibition of ketogenesis. Insulin is considered a fundamental component of DKA treatment as it promotes the peripheral tissues’ utilization of glucose in peripheral tissues by reducing hepatic gluconeogenesis and suppressing ketogenesis.1

Early studies of DKA showed no difference in outcomes with regular insulin administered intravenously, intramuscularly, or subcutaneously. However, based on a preference for rapid onset of effects,2 historically DKA has been treated with continuous intravenous regular insulin in the emergency department or in the intensive care unit (ICU). In the last two decades, newly developed insulin analogs such as lispro and aspart have demonstrated rapid onset of effects in patients with DKA,3 suggesting these analogs given subcutaneously may attain effects commensurate with those of intravenous regular insulin for DKA.

The systematic review discussed here included 5 trials (201subjects in total) that compared subcutaneous rapid-acting insulin analogs with standard intravenous infusion of regular insulin in patients with DKA.4 The meta-analysis did not find any statistically significant difference between groups regarding the study primary endpoints of time to resolution of DKA, episodes of hypoglycemia, and all-cause mortality.4 Time to resolution of DKA was defined as time to reach blood glucose levels < 200 mg/dL and two of the following criteria: a serum bicarbonate level ≥ 15 mEq/L, a venous pH >7.3, and a calculated anion gap ≤12 mEq/L. Hypoglycemia was defined as a plasma glucose less than ≤70 mg/dL (3.9 mmol/L) or according to authors’ definition. Although other adverse events were mentioned as secondary outcomes in 2 trials, no events other than hypoglycemia were reported.

The meta-analysis also analyzed secondary outcomes such as length of hospital stay and costs. Three trials including 90 adult patients compared subcutaneous insulin lispro to intravenous regular insulin and found no difference in hospital length of stay.5^, 6^, 7 One study of 40 adults reported hospital costs were significantly lower in the non-critical care setting with subcutaneous lispro.7

Caveats

The authors of the systematic review rated the quality of evidence for the primary endpoints as mostly low or very low. Other limitations included poor definitions of outcomes, and small numbers of patients and eligible trials. All-cause mortality did not show any significant difference between treatment groups as no deaths were reported. Regarding potential harms, the trials only measured the risk of hypoglycemia and no other harm endpoints were reported.

The clinical implications for resource allocation and medical costs favor subcutaneous insulin as a treatment over continuous infusion of regular insulin for mild to moderate DKA. A retrospective comparison of DKA admissions to the ICU and to step-down units in two academic hospitals in Canada, reported that patients managed in step-down units incurred approximately one-third the cost of an ICU admission.8 The current admission disposition and treatment algorithms are largely institution dependent. However, in the context of increasing demand for limited ICU beds, similar outcomes comparing subcutaneous versus intravenous insulin, and the potential for healthcare cost savings, consideration of subcutaneous rapid insulin analogues as a replacement for intravenous insulin seems logical.9

Enthusiasm for treatment of DKA with subcutaneous should be tempered with caution. Patients with severe DKA and the elderly patients were not included in this review. The applicability in children must also be carefully evaluated since just one study enrolled pediatric patients.10 However, since the publication of the Cochrane review, Razavi et al reported on the use of subcutaneous aspart in 50 pediatric patients with mild-to-moderate DKA and noted decreased time to hospital discharge in the subcutaneous group without change in mortality or adverse events.11

In summary, the use of continuous intravenous regular insulin compared to subcutaneous insulin rapid-acting analogs did not offer increased clinical benefits for patients admitted with mild to moderate DKA. The data appears to be supportive of using this treatment option particularly in non-ICU settings. However, the existing data are severely limited by the small number of trials and small sample sizes, as well as the low quality. Further multi-centered randomized controlled trials detailing patient outcomes, patient satisfaction, and economic end-points are needed. Given the potential benefits in light of significant data limitations, we have assigned a color recommendation of yellow (Unclear If Benefits) to this intervention.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Lillian Chow, MD; Walter Valesky, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Smith SM, Fahey T, Smucny J, Becker LA. Antibiotics for acute bronchitis. Cochrane Database Syst Rev. 2017;6(6):CD000245.

Study Population: 5099 patients with acute bronchitis or acute productive cough with persistent cold or flulike illness

Efficacy Endpoints

Overall clinical improvement, patient-reported cough symptoms, feelings of illness, and limitations on activity, as assessed during follow-up at 2 to 14 days

Harm Endpoints

Adverse effects of antibiotic use (most commonly gastrointestinal symptoms), headache, skin rash, and vaginitis

Narrative

Acute bronchitis is a lower respiratory tract infection, most commonly viral, that accounts for a significant number of health care visits (100 million annually in the United States).1 It is characterized by acute onset and persistence of cough for one to three weeks. Reassurance and symptom control are the foundation of care for this self-limited condition; however, studies indicate that 50% to 90% of patients are prescribed antibiotics.2

The Cochrane review summarized here assessed the effect of antibiotics in patients with acute bronchitis.3 The review analyzed 17 randomized trials that compared any antibiotic therapy with placebo or with no treatment. The studies included 5,099 patients of either sex with acute bronchitis or cough, with persistent cold- or flulike illness, and who did not have preexisting pulmonary disease. Most trials excluded patients with clinical findings of pneumonia; four trials excluded patients based on radiographic findings. Eight of the studies enrolled adults, whereas the remainder included children and adolescents of varying ages. Antibiotics included doxycycline, erythromycin, trimethoprim/sulfamethoxazole, azithromycin (Zithromax), cefuroxime, amoxicillin, and amoxicillin/clavulanate (Augmentin). In most of the studies, there was a single follow-up reassessment at two to 14 days after the initiation of treatment.

Among 3,841 participants, there was no statistically significant difference in general improvement as assessed by clinicians at follow-up (relative risk [RR] = 1.1; 95% CI, 0.99 to 1.2). This outcome integrated a mix of patient-reported measures from the individual trials (e.g., reduction in severity scores, reported global improvement, decreased limitations, resolution of moderate to severe symptoms). Patients treated with antibiotics were less likely to report cough (RR = 0.6; 95% CI, 0.5 to 0.9; number needed to treat = 6; n = 275). There was also an overall reduction in days with impaired activity (0.5 days; 95% CI, 0.04 to 0.9; n = 767) and days of feeling ill (0.6 days; 95% CI, 0.1 to 1.2; n = 809).

Twelve of the 17 studies (n = 3,496) reported adverse effects. There was a statistically significant increase in adverse effects overall, including nausea, vomiting, diarrhea, headache, rash, and vaginitis (RR = 1.2; 95% CI, 1.1 to 1.4; number needed to harm = 24).

Caveats

The quality of trials was reported as "generally good". The 17 studies varied in outcomes measured and time of follow-up, and some reported only on the duration of symptoms. There appeared to be a modest statistical effect of antibiotics on some symptom-related outcomes, but these benefits were slight, raising the question of clinical significance.

No benefit was found for the global measure of improvement at follow-up (n = 3,481). The most recently conducted study included in the review (n = 2,061) showed that neither duration of "moderately bad" or "worse" symptoms, nor mean symptom severity, differed between those who received antibiotics and those who did not. Cough and combined-symptom outcomes were not evaluated separately in this trial, so the results could not be incorporated into the meta-analysis of cough. If the results had been incorporated, it seems likely that the demonstrated small benefits in resolution of cough might have disappeared.

In addition, the 17 studies were not uniform as to whether chest radiography was performed before enrollment in the study. Apparent benefits may have resulted from the inclusion of patients who had pneumonia. However, imaging is not commonly performed to confirm acute bronchitis, and limiting the meta-analysis to studies that required chest radiography would have reduced its generalizability.

This review demonstrates that antibiotics provide small benefits in cough and activity level in patients with acute bronchitis. If more recent trials had been included in the subgroup analyses, these benefits may have disappeared. The clinical significance of this is unclear, particularly in comparison to adverse effects, the medicalizing of a self-remitting condition, and broader threats such as antibiotic resistance. Therefore, we have assigned a color rating of red (no benefits) to this treatment.

The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.

See theNNT.com's previous reviews of this topic:
Antibiotics for the Treatment of Acute Bronchitis in Adults, January 30, 2012

Author

Brian M. Killeen, MD; Allan B. Wolfson, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Berbenetz N, Wang Y, Brown J, et al. Non-invasive positive pressure ventilation (CPAP or bilevel NPPV) for cardiogenic pulmonary oedema. Cochrane Database Syst Rev. 2019;4(4):CD005351.

Study Population: 2664 adult participants (>18 years of age) with respiratory distress due to acute cardiogenic pulmonary edema not requiring immediate mechanical ventilation treated either prehospital, in the ED, or in the intensive care unit

Efficacy Endpoints

Mortality, need for endotracheal intubation

Harm Endpoints

Acute myocardial infarction, increased length of hospital stay

Narrative

Acute cardiogenic pulmonary edema (ACPE) has traditionally been treated pharmacologically, with a combination of nitrates, diuretics, morphine, and inotropes. Since the introduction of noninvasive positive pressure ventilation (NIPPV), a method of providing mechanical ventilation that does not bypass the upper airway, this modality has been widely used as an important addition to the acute care of ACPE. NIPPV encompasses both CPAP (continuous positive airway pressure) and BiPAP (bilevel positive airway pressure), two slightly varied modalities of breathing support where oxygen and additional pressure are delivered most commonly through a facial mask. CPAP delivers a constant pressure throughout the respiratory cycle, while BiPAP delivers independent levels of pressure support set by the clinician for both inspiratory and expiratory phases.

Noninvasive positive pressure ventilation provides multiple benefits in ACPE including improving respiratory dynamics, hemodynamic effects, and decreasing work of breathing. Beneficial respiratory effects include recruitment of additional alveolar units for gas exchange, correction of hypoxemia, reduction in hypoxia-induced pulmonary vascular resistance, and improvement in lung compliance. Hemodynamic effects are complex, however, and ultimately lead to a reduction in afterload and increased LV emptying. Decreasing work of breathing will in turn reduce myocardial oxygen demand and allow for improvement in overall cardiopulmonary dynamics. A 2013 meta-analysis of patients with ACPE found no significant differences in clinical outcomes when comparing CPAP to BiPAP.1

Common contraindications to the use of NIPPV include cardiac or respiratory arrest, inability to cooperate or to remove own mask in an emergency, inability to tolerate or clear secretions, facial trauma, high aspiration risk, and anticipation of a prolonged duration of mechanical ventilation. Canadian and European guidelines differ in their recommendations regarding NIPPV; Canadian guidelines suggest initiating NIPPV only if hypoxia persists despite standard medical care, whereas European guidelines recommend early initiation of NIPPV for those presenting with tachypnea (RR > 25) and hypoxia (SpO2 < 90). American guidelines do not provide clear recommendations for the treatment of ACPE.2

This is an updated review utilizing the 2019 Cochrane review by Berbenetz et al.,3 which sought to evaluate the safety and effectiveness of NIPPV compared to standard medical care for adults with ACPE. The review included 24 studies with 2,664 adult participants (>18 years of age) with respiratory distress due to ACPE who did not require immediate mechanical ventilation. Patients in the included studies were randomized to standard medical care or standard medical care plus NIPPV. NIPPV included both continuous (CPAP) and bilevel (BiPAP) positive pressure support. Standard medical care was supplemental oxygen and pharmacologic treatments that included various combinations of loop diuretics, nitrates, opioids, and inotropes. The median follow-up for determining hospital mortality was 13 days and for endotracheal intubation was 1 day.

In the meta-analysis, NIPPV was associated with statistically significantly lower mortality than standard care (relative risk [RR] = 0.65, 95% confidence interval [CI] = 0.51 to 0.82, absolute risk difference [ARD] = 5.9%, number needed to treat [NNT] = 17, quality of evidence = low). NIPPV was also associated with lower intubation rates (RR = 0.49, 95% CI = 0.38 to 0.62, ARD = 7.7%, NNT = 13, quality of evidence = moderate).3

The risks of myocardial infarction and increased length of stay (potential harms) were not significantly different between the groups. Two RCTs reported a shortened ICU length of stay for NIPPV. However, the quality of the pooled evidence was very low. Other risks or limitations of NIPPV such as inability to tolerate the mask, aspiration, etc., were inconsistently defined and reported in the included trials.

Caveats

diuretics was not defined, and it is possible that the benefits of NIPPV might have been different (reduced or null), based on the clinician providing the care. In addition, the studies included in the analysis followed different protocols for intubation, and it is unclear if this affected the reduction in intubation rates appreciated in the NIPPV group. It is worth noting that patients presenting with need for immediate intubation, depressed GCS scores, hypotension, need for PCI, and evidence of infectious sources were broadly excluded from the studies. It remains to be seen if there is a certain phenotype of patient with ACPE (e.g., more severe ACC/ AHA heart failure grade) who might benefit more from the initiation of NIPPV. These are all potential areas for further research.

The evidence for reduced intubation was of moderate quality and for reduction in mortality and mean length of hospital stay was of low quality. Overall there was low statistical heterogeneity among studies.

Not all patients with ACPE are suitable for NIPPV and the contraindications must be considered prior to initiating therapy; these contraindications include obtunded and uncooperative patients, poor mask fit, facial trauma/burns, facial/esophageal/gastric surgeries, and vomiting or inability to handle secretions.

In conclusion, NIPPV in ACPE (for those not requiring immediate ETI) appears to provide beneficial effects, including reduced mortality and decreased intubation rates, without an apparent increase in clinically significant harms. We thus assign a color rating of green (benefits > harms) to this treatment.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

See theNNT.com's previous reviews of this topic:
Non-Invasive Positive Pressure Ventilation for Acute Pulmonary Edema, August 22, 2010

Author

Brian M. Killeen, MD; Allan B. Wolfson, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Carson JL, Stanworth SJ, Roubinian N, et al. Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion. Cochrane Database Syst Rev. 2016;10(10):CD002042.

Study Population: 12,587 hospitalized adults with anemia of varying severity in 31 trials

Efficacy Endpoints

30-day mortality

Harm Endpoints

Cardiac events, myocardial infarction, congestive heart failure, rebleeding, sepsis, pneumonia, thromboembolism, renal failure

Narrative

Blood transfusion is a common treatment of anemia due to chronic disease or acute blood loss.1 However, there continues to be uncertainty concerning the appropriate threshold for transfusion. A restrictive protocol could decrease blood administered, transmissible infections, transfusion reactions, volume overload, and utilization of a limited commodity. However, anemia may result in decreased oxygen delivery which could lead to metabolic dysfunction and increased cell death. Determining an appropriate transfusion threshold is thus an important objective. A previous Cochrane review2 found the use of restrictive thresholds decreased transfusions compared to liberal thresholds. No difference in secondary outcomes, including 30-day mortality, hospital length of stay, cardiac events, and myocardial infarction, was noted.

Here we summarize an updated Cochrane review of 31 trials and 12,587 adult participants.3 Included studies were randomized trials that assigned patients to either liberal or restrictive transfusion thresholds, typically defined as a hemoglobin level of 9-10 g/dL or 7-8 g/dL, respectively. The primary outcome was 30-day mortality. Secondary outcomes included the proportion of participants transfused, number of units transfused, cardiac events, non-fatal and fatal myocardial infarction, renal failure, and congestive heart failure.

The authors found no difference in 30-day mortality between liberal and restrictive transfusion protocols (Relative risk [RR] 0.97, 95% CI 0.8 to 1.2). However, 84% of the subjects assigned to liberal thresholds received a blood transfusion, compared to 48% of the restrictive group subjects, an absolute reduction of 36% (NNT of 3 for avoiding red cell transfusion). The liberal threshold also resulted in 1.3 more units being transfused per participant than the restrictive threshold. There was no difference between the two groups in cardiac events, myocardial infarction, congestive heart failure, rebleeding, sepsis, pneumonia, thromboembolism, or renal failure.

Subgroup analysis of 30-day mortality was also performed and two subgroup results are of particular note. The first consisted of patients with an acute myocardial infarction, for whom two studies found 30-day mortality was numerically (though not statistically) higher in the restrictive than in the liberal group, with low numbers of deaths (9/78 vs. 2/76) and low numbers overall in the analysis (n=154). On the other hand, participants with acute gastrointestinal bleeding in three studies had a significantly lower mortality with the restrictive protocol than with the liberal protocol [RR 0.65 CI 0.43-0.97, n=1522, NNT 37].

Caveats

This Cochrane review was a comprehensive and important update to the prior review, but several important caveats should be kept in mind. These include the potential for biases due to lack of blinding, statistical heterogeneity in the secondary outcomes, and differences among study methodologies. The intervention examined in this study, the administration of blood, makes blinding difficult, increasing risk of bias as only one trial blinded participants to treatment. Another concern is statistical heterogeneity. While the primary outcome had low heterogeneity, some secondary outcomes, including the risk of receiving blood transfusion, showed heterogeneity. Finally, important methodological differences between the studies could have affected the validity of the findings. One important example is differing restrictive transfusion thresholds.

Half of the studies used a restrictive threshold of 7g/dL, while the other half used 8g/dL. A transfusion threshold of 7 g/dL was more common in studies of ICU patients. Study populations varied widely, and different transfusion thresholds may affect different patients and conditions differently. This is apparent in the subgroup analysis where there were signals of possible benefit and harms. Subgroup analysis of patients with acute gastrointestinal bleeding demonstrated benefit with the restrictive protocol, while analysis of patients with acute myocardial infarction demonstrated possible harm. The implication of these findings is unclear, though they suggest areas for research.

While there was no difference in mortality or adverse outcomes, restrictive transfusion criteria did decrease transfusions, a potential benefit we see as important based on resource utilization and harm risks associated with transfusion.

Of note, a systematic review published in 2020 evaluated the effects of a restrictive compared to a liberal hemoglobin threshold in intensive care patients, and supports the findings of the Cochrane review.4 Eight randomized controlled trials and 3415 subjects were included and no significant difference was found in short-term mortality, length of hospital stay, length of ICU stay, or ischemic events.

As there was no evidence suggesting benefits with the routine use of a liberal transfusion threshold and since a liberal transfusion threshold leads to greater blood administration, we have assigned a color recommendation of Black (harms > benefits). In summary, routine use of a liberal hemoglobin threshold demonstrated no clear medical benefit or harm in regards to 30- day mortality and morbidity outcomes. However, the liberal threshold did lead to increase utilization of blood products, which could pose a risk of infection and transfusion reaction.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

John Z. Hillenkamp, MD; Allan B. Wolfson, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

October 15, 2020

Source

de Cassan S, Thompson MJ, Perera R, et al. Corticosteroids as standalone or add-on treatment for sore throat. Cochrane Database Syst Rev. 2020;5(5):CD008268.

Study Population: 1319 adult and pediatric outpatients with presumed infectious sore throat from 9 trials

Efficacy Endpoints

Complete resolution of pain at 24 and 48 hours, mean time to onset of pain relief, and mean time to pain resolution

Harm Endpoints

Hospitalization, peritonsillar or parapharyngeal abscess, pneumonia, severe tonsillitis, pneumonia death

Narrative

In the United States, up to 2% of ambulatory visits are related to sore throats.1 ‘Sore throat’ is a broad term with a wide range of etiologies. Infections are the most common cause, with group A beta-hemolytic streptococcus (GABHS) accounting for 10% of cases in adults and up to 30% in pediatric patients.2^, 3 A variety of interventions can be used to reduce pain, and while antibiotics are frequently used to treat infection, their overall analgesic effect is likely small and their overuse may lead to bacterial resistance and other adverse events.4^, 5 Another option is corticosteroids, which are thought to reduce oropharyngeal inflammation. Corticosteroids may have analgesic/anti-inflammatory effects.6 They have demonstrated efficacy in other inflammatory conditions such as viral croup and glandular fever.7^, 8 The publication of a recent large, rigorous trial of corticosteroids for sore throat,9 prompted the updated meta-analysis discussed here.10

The systematic review summarized here included double-blind, placebo-controlled, randomized trials (RCTs) of corticosteroids versus placebo for patients aged 3 and older with any of the following: signs of acute tonsillitis, pharyngitis, or clinical syndrome of sore throat (defined as painful throat or odynophagia).10 All corticosteroid routes and types were eligible. Hospitalized patients and those with documented glandular fever, sore throat following tonsillectomy or intubation, and peritonsillar abscess were excluded. Primary outcomes included pain resolution at 24 and 48 hours, mean time to onset of pain relief, and mean time to complete resolution of pain.

The authors of the meta-analysis identified 9 RCTs with 1319 subjects in aggregate. Study settings included the emergency department and general practices in 5 countries. In 8 studies, patients received antibiotics, and in 1 study, clinicians offered a delayed antibiotic prescription or no antibiotics. Corticosteroids included betamethasone up to 8 mg, dexamethasone up to 10 mg, or prednisone 60 mg. Seven trials utilized a single dose of corticosteroids, one used prednisone for 1 or 2 days, and one used dexamethasone for 1 or 3 days. Routes included intramuscular, oral, or both.

When used with antibiotics, corticosteroids improved the likelihood of complete pain resolution at 24 hours (absolute reduction 20.8%; number needed to treat [NNT] 5; relative risk [RR] 2.4, 95% confidence interval [CI] 1.3-4.5) and at 48 hours (absolute reduction 24.4%; NNT 5; RR 1.5, 95% CI 1.3-1.8). Corticosteroids reduced the mean time to onset of pain relief and mean time to complete resolution of pain by 6.0 hours (95% CI 3-9 hours) and 11.6 hours (95% CI 1-22 hours), respectively.

Caveats

While this meta-analysis found that corticosteroids improved resolution of pain, there are several limitations that should be considered. The first is that in the majority of studies patients also received antibiotics. While most patients with sore throat do not need or benefit from antibiotics, it is unclear whether their concomitant use may have influenced the outcomes (pain relief or adverse events). The included trials did not control for the possibility that corticosteroids and antibiotics have a synergistic effect.

Trials varied with regard to types and routes of corticosteroids, as well as the specific outcome measures. Therefore we are unable to determine the optimal route, formulation, or dosing regimen. In addition to clinical heterogeneity, there was statistical heterogeneity regarding complete resolution of pain at 24 hours, mean time to onset of pain relief, and mean time to complete resolution of pain. However, the authors performed a sensitivity analysis excluding each trial in-turn and demonstrated no loss of significance for mean time to onset of pain relief. Two trials accounted for heterogeneity regarding mean time to complete resolution of pain, and removing these trials resulted in a mean time to complete pain resolution of 21 hours. Thus, this heterogeneity likely does not impact the clinical outcomes. The estimates of the mean times to onset of pain relief and complete resolution of pain were further limited by patient recall and recording bias. Only 2 studies included pediatric patients, preventing the authors from drawing conclusions about this population. Finally, the harm endpoints included in the analysis do not include the most common adverse events associated with corticosteroids. Only 2 studies (n=690 patients) reported adverse events, with insufficient power to adequately assess this outcome.

Based on the evidence supporting the significant improvement in pain relief in patients with sore throat, we have assigned a color recommendation of Green (benefit > harm) for corticosteroids. However, we recognize that adverse events were poorly reported. Further data are needed to better evaluate adverse events, corticosteroid efficacy in children, and use of corticosteroids without antibiotics.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Kwan I, Bunn F, Roberts I. Spinal immobilisation for trauma patients. Cochrane Database Syst Rev. 2001;2001(2):CD002803.

Study Population: Trauma patients with potential spinal cord injuries

Efficacy Endpoints

Mortality, neurological disability, spinal stability

Harm Endpoints

Pressure ulcers, discomfort/pain, airway compromise

Narrative

Spinal immobilization or spinal motion restriction after blunt trauma is a routine practice and standard care in most settings due to fear of movement causing or worsening spinal cord injury. This is ingrained in guidelines for trauma management. Immobilization is typically performed with backboards, cervical collars, sandbags, straps, and vacuum mattresses. The aim of these interventions is to restrict mobility in order to prevent secondary spinal cord injury during extrication and transport. However, the benefits of this approach are unclear while harms are common. Cervical collars have been shown to have several deleterious effects. Studies have shown that cervical collars may increase ICP1 and that these increases may in fact be worse in the setting of underlying elevated ICP.2 Additionally, cervical collars may reduce mouth opening and complicate airway mechanics which may impede rapid airway management.3^, 4 Collars may also increase the risk of aspiration,5 conceal wounds or cause local pressure injury,6 resulting inpatient discomfort.

The Cochrane systematic review discussed here7 searched multiple electronic databases for randomized controlled trials studying the effects of spinal immobilization. No studies of sufficient quality were found, and the authors were unable to make firm conclusions about the effect of immobilization on mortality, neurologic injury, spinal stability, or adverse effects.

Caveats

This Cochrane systematic review restricted the inclusion criteria to randomized trials. Given that no such trials exist, the authors conclude that there is insufficient evidence for or against spinal immobilization in obtunded blunt trauma patients. However, absence of trial data points out that further research is needed. In the meantime, lower level studies are the only sources to help answer the question.

In the absence of trial data, common sense and general principles of causal inference are important. For instance, the moment at which irreversible traumatic cord damage occurs is overwhelmingly the moment of impact. This seems true anecdotally and intuitively, as any massive energy transfer capable of compromising spinal column integrity will rarely be trumped by, or simply a set up for, a comparatively minor movement during patient care. In rare cases such events have been reported, though it is difficult to know how often. Since these events are indistinguishable from cases in which impact causes ischemia, compression, or hemorrhage that leads to deterioration in the period during patient care and transport. Regardless, it is apparent that the overwhelming majority of blunt trauma patients do not have spinal instability and thus can only be harmed by immobilization.

While a complete literature review is beyond the scope of this summary it is notable that validated criteria for spine injury clearance are widely and safely applied both in the emergency department and the prehospital setting,8 and a recent trend favors spinal motion restriction without boards or collars.9^, 10 Limited studies have detected no increases in cord injury following implementation of any such protocols.9 Though some professional societies including the Society for Academic Emergency Medicine (SAEM) do recommend cervical collars in the setting of blunt trauma, the position paper cited by SAEM does not provide any evidence of efficacy of cervical collars or rationale for their presumed benefit.11

As enumerated in a letter to the editor in response to a 2006 summary of this Cochrane review, one non-peer reviewed estimate of NNT for spinal immobilization to prevent secondary neurologic injury was reported to be between 625 and 3333.12 This was calculated based on three case series’ published 32 to 64 years ago,13^, 14^, 15 and assumes all episodes of deterioration were both directly caused by movement during transport and were preventable by immobilization. This claim, however, is critically limited by the now antiquated manner in which these cases were managed on presentation—in many cases no cervical spine imaging was performed despite major trauma or a clear spinal injury threat. The workup of severe trauma in the modern emergency department universally includes high-quality CT imaging in such cases, which seems almost certain to have identified causes of neurologic compromise early in the hospital course of these patients. Conversely, harms of spinal immobilization have not been estimated or quantified in this fashion, though one review suggests morbidity and mortality likely trumps any benefits.16

The weakness of supporting evidence, known harms, and burdens of immobilization have resulted in calls to discontinue the practice. Yet it remains intuitively possible that immobilization may provide a benefit in some cases. Perhaps subgroups more likely to see that benefit can be identified and targeted (e.g. patients with neurological deficits or those with gross anatomical deformity of the spine10), relieving systems and individuals of the harms and devotion of resources associated with widespread, routine immobilization. Well-designed large trials are needed to answer these questions. Based on the existing evidence, we have assigned a color recommendation of Yellow (Unclear if benefits) to this intervention.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Olivia Serigano, MD; Matthew Riscinti, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

September 21, 2020

Source

Chabok A, Påhlman L, Hjern F, Haapaniemi S, Smedh K. Randomized clinical trial of antibiotics in acute uncomplicated diverticulitis. Br J Surg. 2012;99(4):532-9.

Daniels L, Ünlü Ç, De korte N, et al. Randomized clinical trial of observational versus antibiotic treatment for a first episode of CT-proven uncomplicated acute diverticulitis. Br J Surg. 2017;104(1):52-61.

Study Population: 1151 patients from two RCTs

Efficacy Endpoints

Requiring additional treatment or intervention during initial episode

Harm Endpoints

Antibiotic-related morbidity

Narrative

In the United States there are over 2.6 million outpatient visits and 200,000 inpatient admissions for diverticulitis annually.1 Diverticulitis can be divided into uncomplicated and complicated forms. The term complicated is used when diverticultitis is associated with abscess formation, fistula, and bowel obstruction or perforation. About 5-15% of patients develop an abscess or fistula, while bowel obstruction and frank perforation are rare.2 The mainstay of treatment for uncomplicated diverticulitis has been antibiotic therapy with bowel rest. However, recent studies have questioned the role of antibiotics.3^, 4 Systematic reviews have examined outcomes of acute uncomplicated diverticulitis treated with or without antibiotics.5^, 6^, 7^, 8

The reviews included randomized trials and observational studies. Because of the high risk of bias and confounding in observational studies, we report only results from the two randomized trials.3^, 4 Endpoints included treatment failure, recurrence of diverticulitis, complications, re-admission, and mortality. Follow up was one month in one trial and 50 months in the second.3^, 4

The difference between groups was not significant for any major endpoints including rates of recurrence (56/571 vs 54/580, p=0.77), complications (18/571 vs 10/580, p=0.12), readmission (113/571 vs 81/580, p=0.26), or mortality (3/571 vs 1/580, p=0.4).3^, 4

Treatment failure, a secondary outcome initially reported in the systematic review by Emile et al. (defined as deterioration prompting initial or expanded antibiotic treatment), was lower in the antibiotic group (Odds Ratio: 0.6, 95% CI; 0.3 to 0.97); absolute risk reduction 3.1%, NNT 32, n=1151).3^, 4^, 5

While harms were not reported in the systematic review because of inconsistent reporting, the two randomized trials briefly mention adverse events. In the study by Chabok et al. only 3 patients in the antibiotic group experienced adverse events (allergic reactions). Daniels et al. reported 22 adverse events in the antibiotic group (all “antibiotic related without specification) and 1 in the control group (Odds Ratio: 25.68, 95% CI, 3.47 to 190.14; absolute risk reduction: 4.1%; Number-needed-to-harm: 24).3^, 4

Caveats

There are several limitations of these data including having only two studies, along with important variations in outcomes and definitions. The Daniels et al. study included patients with small peri-colonic abscesses while the Chabok et al. study patients with pericolonic abscesses, which may have contributed to a higher treatment failure rate in the Daniels study (10.7% vs 3.2%), though the definition of treatment failure also differed between the studies. In Chabok et al., antibiotic usage was guided by C-reactive protein (CRP) levels, while the Daniels study did not describe a defined protocol or guidance for the use of antibiotics in the control group. Antibiotic treatments also differed between studies.3^, 4

Both studies had significant limitations and potential for bias. Both were unblinded. Enrollment rates also varied by center, leading to a risk of selection bias, however concealment and randomization may have decreased this risk.3^, 4

Based on limited data with high potential for bias we have chosen a recommendation color of Yellow (unclear benefits vs. harms, more data needed). Ongoing studies and future efforts will, we hope, be both more rigorous and better standardized to allow for more reliable pooling and comparison.

This series is coordinated by Christopher Bunt, MD, AFP Assistant Medical Editor, and Daniel Runde, MD, from the NNT Group.

Author

Shiva Poola, MD; Michael Ritchie, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

September 1, 2020

Source

Kirkilesis G, Kakkos SK, Bicknell C, Salim S, Kakavia K. Treatment of distal deep vein thrombosis. Cochrane Database Syst Rev. 2020;4:CD013422.

Study Population: 8 trials comprising 1,239 participants

Efficacy Endpoints

Recurrent VTE (recurrent distal DVT, progression of DVT to proximal veins, or pulmonary embolism), post thrombotic syndrome

Harm Endpoints

Major bleeding

Narrative

Deep venous thrombosis (DVT) has an incidence of 45 to 117 per 100,000 person-years.1 Distal DVT, defined as a venous thrombosis isolated to the calf veins, comprises one-third to one-half of all DVTs.2 However, the treatment of distal DVTs is more controversial with significant variations in practice with regard to decisions on anticoagulation.3

This systematic review and meta-analysis discussed here included five randomized controlled trials (n = 503 patients) comparing anticoagulation with no intervention or placebo for the treatment of distal DVTs, as well as three randomized controlled trials (n = 736 patients) comparing anticoagulation for six weeks versus 12 weeks or more.4 The trials included adults over 18 years of age with a distal DVT identified on ultrasound or venography. The primary outcomes were recurrence of VTE (defined as DVT recurrence in the calf veins, progression of DVT to proximal veins, and pulmonary embolism [PE]) and major bleeding (defined as a 2 g/dL decrease in hemoglobin, transfusion of ≥ 2 units of red cells, bleeding in a critical site, or bleeding contributing to death). Secondary outcomes included recurrence of DVT, PE, clinically relevant non-major bleeding (defined as overt bleeding not meeting the criteria for major bleeding but associated with medical intervention, unscheduled contact with a physician, interruption or discontinuation of study treatment, or associated with any other discomfort such as pain or impairment of activities of daily life), overall mortality, mortality related to PE or major bleeding, post-thrombotic syndrome, and resolution of symptoms.

According to the meta-analysis, the recurrence of VTE was lower in patients who were anticoagulated (5 studies; 503 participants; Relative risk [RR]: 0.34; 95% confidence interval [CI], 0.15 to 0.77; absolute risk difference [ARD]: 6%; NNT: 17; high-certainty evidence). The risk of DVT was also significantly lower in anticoagulated (5 studies; 503 participants; RR 0.25; 95% CI, 0.10 to 0.67; ARD: 6%; NNT: 17; high-certainty evidence). There was also an increased risk of clinically relevant non-major bleeding (2 studies; 329 participants; RR: 3.34; 95% CI, 1.07 to 10.46; ARD: 5%; NNH: 20; high-certainty evidence). There was no significant difference in risk of PE, post-thrombotic syndrome, major bleeding, or mortality between the groups (all low-certainty evidence).

Caveats

This systematic review and meta-analysis has several important limitations. First, the overall sample size was low, and it is possible that a significant difference in PE or major bleeding may have been identified with a larger sample size. Additionally, there were differences in the treatment periods, with some studies treating patients with anticoagulation for 3 months while others treated for 6 weeks. All studies used vitamin K antagonists (following an initial course of heparin), so it is unclear if similar findings would be present with direct oral anticoagulants (DOACs). There was also significant heterogeneity between patient populations, including whether the DVTs were provoked or unprovoked. Moreover, most studies excluded patients with cancer-associated or recurrent distal DVTs, so it is unclear how the findings of this meta-analysis would apply to this population. The primary outcome included recurrence of distal DVT, progression of DVT to proximal veins, and PE. However, recurrence of a distal DVT is a less clinically significant outcome than the latter two outcomes. Two studies were partially funded by pharmaceutical companies. Due to the open-label design in most of the trials, participants were not blinded to their treatment allocation. Furthermore, the risk of post-thrombophlebitis syndrome, a major complication of distal lower extremity DVT, was only assessed in one of the five trials. Therefore, the impact of anticoagulation on reducing the risk of post-thrombotic syndrome is inconclusive. Finally, while this study focused on anticoagulation, the CHEST guidelines recommend serial imaging for two weeks in patients that have a distal DVT without severe symptoms or risk factors for extension, while reserving anticoagulation only for those with more severe disease or a high risk of extension.5 Therefore, it is important to engage in shared decision-making with patients to individualize the decision based on an individual’s risk of recurrence versus risk of bleeding.

Based on the existing data, anticoagulation was associated with a reduced rate of recurrent VTE with no significant increase in the risk of major bleeding compared to placebo or no intervention. However, given the limitations of the specific studies and need to individualize patient-oriented risks and benefits, we have assigned a color recommendation of Yellow (unclear if benefits) to this intervention. Further study is needed with larger sample sizes, patients taking DOACs for treatment of distal DVT, and other populations including those with cancer-associated or recurrent distal DVTs.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Michael Gottlieb, MD; Brit Long, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Chavez-tapia NC, Barrientos-gutierrez T, Tellez-avila FI, Soares-weiser K, Uribe M. Antibiotic prophylaxis for cirrhotic patients with upper gastrointestinal bleeding. Cochrane Database Syst Rev. 2010;(9):CD002907.

Study Population: 1241 patients with cirrhosis hospitalized with upper gastrointestinal bleeding from 12 trials comparing prophylactic antibiotics to no antibiotic prophylaxis

Efficacy Endpoints

Mortality, infection during hospitalization (spontaneous peritoneal peritonitis, bacteremia, pneumonia, urinary tract infection)

Harm Endpoints

Not Reported

Narrative

Cirrhotic patients often develop bleeding from gastric or esophageal varices that occur secondary to portal hypertension. Gastrointestinal (GI) bleeding is fatal in approximately 20% of these episodes and bacterial infections are an important contributor to this mortality. Patients with cirrhosis are also known to have impaired immune function and also at higher risk of translocation of bacteria from the gut into the bloodstream.1 Therefore, the administration of prophylactic antibiotics during the bleeding event might help prevent such infections.

The Cochrane systematic review discussed here included 12 trials (n = 1241) involving cirrhotic patients with upper GI bleeding. Of the 12 included trials, only 1 was placebo controlled, the other 11 examined antibiotics vs. no intervention.2 These trials enrolled adult patients with cirrhosis and upper GI bleeding regardless of the severity or etiology of the cirrhosis. They excluded patients who had bacterial infections at the time of admission, positive blood cultures or who underwent surgery in the first 12-24 hours of hospitalization. Among the included trials, the length of follow up for determining mortality endpoint ranged from in-hospital to 90 days. The analysis demonstrated a clear decrease in overall rate of hospital-acquired bacterial infections, with marked reductions in nosocomial bacteremia, pneumonia, spontaneous bacterial peritonitis and urinary tract infections (Odds ratio [OR]: 0.36, 95%CI, 0.27 to 0.49; Absolute risk difference [ARD]: 23%; Number-needed-to-treat [NNT]: 4). With the exception of pneumonia, all of the infections were confirmed by cultures. The trials also noted an overall decrease in mortality (OR: 0.79, 95%CI, 0.63 to 0.98; ARD: 4.6%; NNT: 22). The choice of antibiotic regimen appeared to have no effect, although all antibiotics used in these trials were chosen because of their activity against gram negative organisms (the most common infecting agents for the targeted infection types). The most common antibiotics used in the trials were quinolones followed by cephalosporins. The subgroup analysis showed more benefits from cephalosporins than quinolones for reducing bacterial infections.

One study that was published after the Cochrane systematic review, reported a retrospective analysis of 381 patients with cirrhosis and variceal upper GI bleeding. This represented one of the most relevant studies on this topic since the completion of the most recent Cochrane systematic review, even though its retrospective nature would preclude it from being included in an updated Cochrane review on this topic. It found that antibiotic prophylaxis was associated with a lower risk of infection (OR: 0.37; 95% CI 0.31 to 0.74) but no significant change in overall mortality.3 However, subgroup analysis did find a mortality benefit that was severity-dependent: in patients with Child-Pugh class C (i.e. more severe cirrhosis) antibiotics reduced 6-week mortality by approximately 50% (from 62% in those not exposed to antibiotics to 35% in those exposed to antibiotics). In patients with Child-Pugh class B (i.e. less severe cirrhosis), the drop in mortality after using antibiotics was from 7% in nonexposed to 5% in exposed groups. The mortality in patients with Child-Pugh class A (i.e. mild cirrhosis) was negligible regardless of antibiotic administration.3 These findings generally support the conclusions of the systematic review discussed here.

Caveats

We should also note that while this review is over a decade old, a review of the literature did not reveal any new trials that would have impacted the conclusions of this review. None of the included trials reported harms or adverse effects associated with administration of antibiotics. The authors of the Cochrane systematic review themselves make a point to note that, “Adverse events, quality of life, and the economic impact of the intervention were not explored in the trials included, remaining important areas of uncertainty and requiring further data to establish an evidence‐based conclusion.” Furthermore, only 1 trial was placebo controlled, which introduces a significant risk of bias. Perhaps more importantly, as noted by the Cochrane review authors, the data for decreased mortality in the intervention group was not as compelling as it was for preventing infection. Many of the included trials were not powered to determine a mortality benefit. Furthermore, trial sequential analysis (a statistical tool used to evaluate the strength of results found during meta-analysis) found that the 12 included trials were not enough to produce a definitive conclusion regarding the survival benefit. This indicates that a large, high quality, methodologically rigorous randomized trial has the potential to trump the results of this review, as it may well generate results that are in disagreement with the results of this review (for all outcomes including infection rates). We would like to see a trial like this performed, and we believe that given the poor quality of existing trial data there is clinical equipoise adequate to perform such a trial.

The trials included in the meta-analysis suffered from various methodological limitations and therefore the produced evidence is subject to bias. None of the included trials were rated as low risk of bias. The heterogeneity for the mortality endpoint was low but it was significant for bacterial infection. Regardless of these limitations, as this review represents the best available data, and given the reported benefits of reducing hospital-acquired infections and the possible decrease in mortality, it seems appropriate to recommend this intervention in cirrhotic patients with upper GI bleeding. However, the adverse events associated with antibiotics such as allergic reactions, rash, gastrointestinal upset, clostridium difficile infections, antibiotic resistance, etc. should be balanced against these benefits on a case by case basis.

Based on the existing evidence, we have assigned a color recommendation of Green (benefits outweigh harms), as the reported benefits are significant and clinically relevant. As noted above, further data would be welcome to better characterize the degree of benefit, and assess any adverse events associated with antibiotic prophylaxis.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Daniel Runde, MD
Supervising Editor: Shahriar Zehtabchi, MD

Published/Updated

August 24, 2020

Source

RECOVERY Collaborative Group, Horby P, Lim WS, et al. Dexamethasone in Hospitalized Patients with Covid-19 - Preliminary Report [published online ahead of print, 2020 Jul 17]. N Engl J Med. 2020.

Study Population: 6425 hospitalized adults with COVID-19 in 176 hospitals in UK

See Systemic Corticosteroids for COVID-19.

Efficacy Endpoints

All-cause mortality within 28 days, discharge from hospital within 28 days, and requiring invasive mechanical ventilation or extracorporeal membrane oxygenation

Harm Endpoints

Not reported

Narrative

The novel severe acute respiratory syndrome coronavirus (SARS-CoV2) emerged in 2019 resulting in a global pandemic with millions infected and hundreds of thousands dead from coronavirus disease 2019, i.e. ‘COVID-19’.1 COVID-19 can cause lung inflammation and respiratory failure believed to be due to an exaggerated immune response.2 Corticosteroids such as dexamethasone may temper this response.

RECOVERY is a complicated randomized, controlled, open label trial exploring multiple treatment options for COVID-19.3 The results for the dexamethasone arm have been released in a preliminary report. A total of 6425 hospitalized adult patients with suspected or confirmed SARS-CoV-2 infection in the UK were randomized (2:1) to receive usual care or usual care plus dexamethasone 6 mg once daily for up to 10 days. The primary outcome was all-cause mortality. Secondary outcomes included time to discharge and invasive mechanical ventilation or extracorporeal membrane oxygenation. Outcomes were measured at discharge or on day 28 if still in the hospital.3

Administration of dexamethasone was associated with lower mortality overall (relative risk [RR]: 0.83, 95% CI, 0.8-0.9; absolute risk difference [ARD]: 2.8%; NNT: 36). In subgroup analyses this was more pronounced for those receiving invasive mechanical ventilation (RR: 0.6, 95% CI, 0.5-0.8; ARD 12.1%; NNT 8) or requiring oxygen but not ventilation (RR: 0.8, 95% CI, 0.7-0.9; ARD 2.9%; NNT 34). In non-ventilated patients, dexamethasone reduced progression to invasive mechanical ventilation (RR: 0.8, 95% CI, 0.6-0.9; ARD 2.1%; NNT 48). Dexamethasone group mortality was higher in those not on oxygen, but the difference was not statistically significant.3

Caveats

This trial was neither blinded nor placebo-controlled. These limitations introduce significant systemic bias into the findings and limit the validity of the results. There are other limitations as well. Usual care, for instance, was not defined or protocolized which may have led to differential care decisions such as more common withdrawal of measures in the control group or more aggressive care in the treatment group. Furthermore, dexamethasone harms are not detailed in this study even as prior reviews have at times suggested possible increased risks for superinfection, metabolic derangements, and other corticosteroid adverse effects.4

The readers should exercise caution interpreting the results of subgroup analyses from this trial as it is not clear if these analyses were pre-planned or appropriately powered. Moreover, if they were both preplanned and adequately powered the differences would still be considered hypothesis-generating, not definitive.

Although the preliminary results of this study are promising, more research is needed to confirm these findings. Administration of dexamethasone seems reasonable, and hopeful, based on these data. But the overall lack of methodologic rigor, as well as hints of harm in subgroup analysis, suggest these results are unlikely to be stable. Replication studies and subgroup-based investigations will be the only path to accurately answering questions on the utility of dexamethasone in Covid-19. We believe clinical equipoise remains, and have therefore assigned a color recommendation of yellow (unclear if benefits, more data needed). We hope to see large scale further randomized trials as long as the pandemic continues.

Author

Kenneth Lu, MD; Eric Tang, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

Source

Annane D, Bellissant E, Bollaert PE, et al. Corticosteroids for treating sepsis in children and adults. Cochrane Database Syst Rev. 2019;12(12):CD002243.

Study Population: 12,192 adults and children with sepsis, from 61 trials

Efficacy Endpoints

28-day, 90-day, long-term, in-hospital, and ICU mortality, shock reversal at 7 and 28 days, SOFA score at day 7, length of stay in ICU and hospital

Harm Endpoints

Gastroduodenal bleeding, superinfection, hyperglycemia, hypernatremia, muscle weakness, neuropsychiatric event, stroke, cardiac event

Narrative

Sepsis remains a major contributor to morbidity and mortality in the United States,1 and is now recognized as a dysregulated inflammatory response.2 There is laboratory evidence corticosteroids may affect this dysregulated response.3

The review summarized here by Annane et al4 found that corticosteroids may confer a short term, but not overall, mortality benefit for patients with sepsis. Specifically, they found moderate certainty evidence for reduced 28-day mortality (relative risk [RR] 0.91, 95% confidence interval [CI] 0.84 to 0.99; n=11233).

The review also found decreased intensive care unit (ICU) mortality (RR 0.89, 95% CI, 0.83-0.96; n=7267) and hospital mortality (RR 0.90, 95% CI, 0.82-0.99; n=8183) with high to moderate certainty evidence respectively. Annane et al suggested a signal of decreased 90-day mortality (RR 0.93, 95% CI, 0.87-1.00; n=5934), however it failed to reach statistical significance. Beyond 90 days, they found low certainty evidence that there is no mortality benefit (RR 0.97, 95% CI, 0.91-1.03; n=6236). The review found additional benefits of corticosteroids in sepsis including decreased intensive care unit length of stay (mean difference [MD] -1.07 days, 95% CI -1.95 to -0.19, n=7612) and hospital length of stay (MD -1.63 days, 95% CI -2.93 to -0.33, n=8795), decreased organ dysfunction or sequential organ failure assessment (SOFA) score at day 7 (MD -1.37, 95% CI - 1.84 to -0.90, n=2157), and increased number of patients with shock reversal at day 7 (RR 1.23, 95% CI, 1.13-1.34; n=6711) and day 28 (RR 1.06, 95% CI, 1.03-1.08; n=6779).

Complications of corticosteroids included an increase in hyperglycemia (RR 1.17, 95% CI, 1.13- 1.22; n=8594), hypernatremia (RR 1.66, 95% CI, 1.34-2.06; n=5069), and muscle weakness (RR 1.21, 95% CI, 1.01-1.44; n=6145). There were no statistically significant harms in other proposed complications of corticosteroids including gastroduodenal bleeding, superinfection, neuropsychiatric events, stroke or cardiac events.

Caveats

We recognize that many systematic reviews and meta-analyses have explored the use of corticosteroids in sepsis. We chose to summarize the Cochrane Systematic review by Annane et al,4 as it includes the largest (61 trials including 12192 adults and children) and most recent data (including trials up to July 25, 2019), including trials in non-English language as well as unpublished data.

For the primary outcome of 28-day mortality, the review decreased the certainty of evidence from high to moderate for concerns of heterogeneity (I² =29.47%) across trial results. One reason for this significant heterogeneity may be broad inclusion criteria including trials dating back as early as 1972. A priori subgroup analyses found 28-day mortality benefit with long course (≥3 days) of low-dose (≤400mg hydrocortisone or equivalent) corticosteroids (RR 0.91, 95% CI, 0.87-0.97; n=9902) but not with short course of high-dose corticosteroids (RR 0.96, 95% CI, 0.8-1.16; n=910). Furthermore, differences in methodological quality may contribute to heterogeneity, with subgroup analysis of studies with low risk of bias demonstrating 28-day mortality benefit (RR 0.91, 95% CI, 0.84-0.98; n=7896).

Of note, the review includes secondary outcomes of 90-day and long-term (defined as “longest available follow-up beyond three months) mortality. While 28-day was mortality was the primary outcome for the majority of the 61 included trials, it is noteworthy that the two most recent and largest studies5^, 6 used 90-day mortality as their primary outcome. If the future studies continue to include 90-day mortality as a primary outcome, future systematic reviews may provide further clarity if patient-centered long-term benefits or harms truly exist.

The review identifies additional areas for future investigation. The review suggests a possible added benefit of mineralocorticoids, with 28-day mortality benefit of hydrocortisone plus fludrocortisone (RR 0.87, 95% CI, 0.77-0.98; n=1564). Secondly, investigation of the ideal modality of corticosteroid (intermittent bolus vs continuous infusion) found no significant difference (test for subgroup differences Chi² =0.41). Additionally, tapering of corticosteroids was not associated with 28-day mortality difference (RR 1.04, 95% CI, 0.92-1.18; n=2136) whereas absence of taper was (RR 0.87, 95% CI, 0.78-0.98; n=8770). Finally, this review does not find sufficient evidence to suggest the role of corticosteroids in children, different causes of sepsis(for example, ARDS vs community associated pneumonia), or in those with or without shock.

In summary, we assign a recommendation of yellow (potential benefits may outweigh harms) for the use of corticosteroids in patients with sepsis. This review demonstrates benefits in short term mortality, without clear long-term benefit, and at the expense of harms.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

See theNNT.com's previous reviews of this topic:
Systemic Steroids for Sepsis Syndromes , September 7, 2010

Author

Robert Allen, MD; Peter Tepler, MD
Supervising Editor: Kabir Yadav, MD

Published/Updated

July 1, 2020

Source

Van der gaag WH, Roelofs PD, Enthoven WT, Van tulder MW, Koes BW. Non-steroidal anti-inflammatory drugs for acute low back pain. Cochrane Database Syst Rev. 2020;4:CD013581.

Study Population: 2,232 patients with acute low back pain from 9 trials comparing NSAIDs with placebo

Efficacy Endpoints

Pain intensity, disability, and proportion of patients experiencing global improvement

Harm Endpoints

Adverse drug effects

Narrative

Low back pain is a common reason for emergency department presentation and is a leading cause of acute disability.1 Up to 70% of individuals experience low back pain during their lifetime.2 Many patients recover from acute back pain within 6 weeks, but some patients experience multiple episodes or chronic pain.3^, 4^, 5^, 6 Guidelines recommend avoidance of bedrest, early activity, patient education, attention to patient psychosocial factors, and administration of analgesic medications.7^, 8 Clinical practice guidelines from the American College of Physicians recommend non-steroidal anti-inflammatory drugs (NSAIDs) as a treatment option due to their analgesic and anti-inflammatory properties.8

The Cochrane Review, “Non-steroidal anti-inflammatory drugs for acute low back pain”, summarized here included randomized controlled trials (RCTs) evaluating patients aged 18 years and older who were treated with NSAIDs for acute non-specific low back pain, which was defined as pain above the inferior gluteal folds but below the costal margin that was present for less than 12 weeks.9 The meta-analysis authors excluded studies evaluating chronic low back pain, those with sciatica or acute exacerbations of sciatica, and studies with patients who had underlying pathological conditions such as neoplasm, fracture, or infection. The meta-analysis included study-level data, and primary outcomes included pain intensity, disability, global improvement, adverse events, return to work status, and number of days off of work. Pain intensity and disability were evaluated as continuous outcomes. Pain intensity was measured using the visual analogue scale or numeric rating scale, and disability was evaluated using the Roland-Morris Disability Questionnaire scale in studies investigating NSAIDs versus placebo. A between group difference of more than 10% on the utilized scale was determined to be clinically relevant.

The authors identified 32 RCTs (n = 5356 patients) that met inclusion criteria, of which 9 RCTs (n = 2232 patients) evaluated NSAIDs versus placebo.9 For this Brass Tacks, we focused on studies evaluating NSAIDs versus placebo, which demonstrate the highest quality data and greatest applicability to emergency medicine. The NSAIDs evaluated included ibuprofen, Accepted Article piroxicam, dipyrone, tenoxicam, and diclofenac. Treatment periods ranged from 1 day to 4 weeks, and follow-up periods from 1 day to 2 months. Study settings included general practitioner clinics, outpatient clinics, and the emergency department.

NSAID use was associated with a mean reduction in pain intensity of -7.29 (95% confidence interval [CI]: -10.98 to -3.61) on a 100-point visual analogue scale, with a number needed to treat (NNT) of 14. NSAIDs reduced disability compared to placebo (MD: -2.02, 95% CI: -2.89 to -1.15) on the 24-point Roland-Morris Disability Questionnaire scale, with a NNT of 12. NSAIDs demonstrated an 8% absolute increase and NNT of 13 for global improvement compared to placebo. Time to return to work and the proportion of patients who had adverse events were not significantly different between NSAIDs and placebo.

Caveats

There are important limitations of these findings, most importantly the variable quality of evidence and risk of bias. All nine included trials compared NSAIDs with placebo, finding a reduction in pain intensity and short-term disability and an increase in global improvement with the former. Four studies reported on pain intensity, with moderate quality of evidence; the mean difference in pain intensity was 7.29, less than the clinically significant difference of 12 on a 100 point scale.10 Only two trials reported on short-term disability within 3 weeks; the differences were small and of unclear clinical relevance. Five trials reported on global improvement, but there was significant heterogeneity in the scales used and the types of outcomes measured. Studies also used different cutoff points for outcomes such as global improvement and modes of medication delivery. Most of the trials evaluating adverse events had small sample sizes and were not sufficiently powered to exclude a clinically significant difference.

Further limitations include variation in the duration of follow-up among studies, with only 3 trials having follow-up greater than 3 weeks. At least five trials were industry-sponsored, with other studies not fully describing sponsorship, which can increase the risk of bias. The most commonly identified biases were performance bias and attrition bias. Incomplete information about randomization and allocation concealment introduced a risk of selection bias, and most studies were not registered, increasing the risk of selective reporting. The diversity of study populations resulted in heterogeneity, and subjects were drawn from a variety of general practitioner and outpatient clinics, which may not fully reflect the same patients evaluated for this in the emergency department. Finally, not all outcome measures were reported in the 9 RCTs evaluating NSAIDs with placebo; for example, outcomes related to return to work and long-term follow-up were unavailable for many trials.

Based on this evidence, we have assigned a color recommendation of Yellow (Unclear if benefits) for use of NSAIDs in patients with acute LBP, as the magnitude of effect was small and of questionable clinical relevance. Further data are needed to more fully evaluate the role of NSAIDs for acute low back pain in the emergency department setting.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Allan Wolfson, MD; Fredrik Amell, MD

Published/Updated

Source

Kew KM, Kirtchuk L, Michell CI. Intravenous magnesium sulfate for treating adults with acute asthma in the emergency department. Cochrane Database of Systematic Reviews 2014.

Study Population: 11 studies with 1769 primarily adult patients with asthma exacerbation

Efficacy Endpoints

Hospitalization

Harm Endpoints

Adverse events (flushing, fatigue, nausea, headache, hypotension)

Narrative

Asthma is a common chronic respiratory disease with acute exacerbations recognized clinically by the signs and symptoms of dyspnea, cough, chest tightness, and wheezing. In the United States approximately 25 million individuals currently have asthma and in 2017, asthma exacerbations accounted for approximately 1.8 million emergency department (ED) visits and 3,500 deaths.1 Acute asthma exacerbations range in severity from mild to deadly. The first line medications for the treatment of asthma exacerbations in the emergency department are oxygen (as needed to maintain arterial oxygen saturation of 93-95%), short-acting beta2-agonists, and systemic corticosteroids, with ipratropium bromide being recommended for severe exacerbations. In cases of treatment failure or severe exacerbations, some studies have recommended intravenous magnesium sulfate.2 Magnesium is believed to treat asthma exacerbations through numerous mechanisms including direct smooth muscle relaxation, an anti-inflammatory effect, and by blocking calcium ion influx into bronchial smooth muscle.3^, 4 The goal of the systematic review summarized here is to provide an updated review of the safety and efficacy of intravenous magnesium sulfate in the treatment of moderate to severe asthma exacerbations in the ED.5

The systematic review discussed here included fourteen randomized controlled trials composed of 2313 primarily adult patients who presented to the ED for acute asthma exacerbations of at least moderate severity.6 Patients were excluded if they had diabetes, congestive cardiac disease, hypertension, chronic renal failure, temperature > 38 °C , pneumonia, pregnancy, or required ventilation. Eleven studies with 1769 patients contributed to the primary outcome analysis. The trials compared the efficacy of intravenous magnesium sulfate in any dose (typically 1.2 g - 2 g) to placebo. In both the magnesium sulfate and placebo groups, patients were treated with standard medications for acute asthma. The primary outcome was hospitalization. Secondary outcomes included adverse medication effects, ED treatment duration, intensive care unit admissions, vital signs, forced expiratory volume in 1 second (FEV₁) and peak expiratory flow (PEF), and symptom scale scores. These secondary outcomes were mostly of low to moderate quality of evidence and, with the exception of adverse medication effects, FEV₁, and PEF are omitted from this summary as they do not provide additional patient-centric information.5

According to the results of the Cochrane meta-analysis, intravenous magnesium sulfate therapy was associated with a reduced risk of hospitalization (Odds ratio [OR]: 0.75; 95% confidence interval [CI]: 0.60 to 0.92; absolute risk difference [ARD]: 7.1%; number needed to treat [NNT]: 14), an increased FEV1 % predicted (Mean difference [MD]: 4.41%; 95% CI: 1.75 to 7.06; 4 studies, 523 patients, highquality evidence), and an increased PEF % predicted (MD: 4.78%; 95% CI: 2.14 to 7.43; 3 studies, 1129 patients, high-quality evidence).5

Adverse events were inconsistently reported and could not be aggregated. Commonly cited adverse events were mild and included flushing, fatigue, nausea, headache, and hypotension. Most of the serious adverse events that occurred (death, intubation, arrhythmia), were believed to be due to the asthma exacerbation rather than the intravenous magnesium sulfate. In the largest trial included in this meta-analysis, intravenous magnesium sulfate was associated with a slightly increased risk of any adverse events (OR: 1.68; 95% CI: 1.07 to 2.63; ARD: 3.3%; Number Needed to Harm [NNH]: 30). The study was not powered to determine differences between serious adverse events.7

There is however good data from the obstetrics literature to support the safety of intravenous magnesium sulfate. In obstetrics, magnesium sulfate is used for the treatment of pre-eclampsia and eclampsia as well as for fetal neuroprotection before anticipated early preterm delivery. Typical dosing in obstetrics is a 4-6 gram bolus along with maintenance infusions of 1-2 g/hr, which is far higher than the dosing used in the treatment of asthma. In the most comprehensive systematic review of the maternal safety of antenatal magnesium sulfate to date, compared with placebo or no treatment, treatment with magnesium sulfate did not increase the risk of maternal death, cardiac arrest, or respiratory arrest (4 trials, 13,977 women). It did however increase the risk of ‘any side effects’ (e.g. warmth, flushing, itching, hypotension, fatigue, nausea/vomiting), which were generally mild (RR: 4.62; 95% CI: 2.42 to 8.83; ARD: 29.5%; NNH: 3; 4 trials, 13,322 women).8

Caveats

The systematic review reported the quality of evidence to be high, the risk of bias low, and no evidence of significant heterogeneity. However, there are several limitations. First, this metaanalysis was not able to determine the ideal dose of magnesium, and there was some variation in each of these among the trials. The most common dose administered in the trials was 1 to 2 grams slow intravenous infusion over 15-30 minutes. Second, the criteria for hospital admission was not standardized and varied between trials, limiting the generalizability of the findings. Third, although there is high quality data showing an improvement in spirometric values, which may help explain the improved primary outcome, the improvement is minor (less than 5%), and the clinical significance of these changes is questionable. Fourth, in some studies evaluating the safety and efficacy of magnesium therapy, patients with life threatening asthma were excluded. This is important to keep in mind because outside of clinical trials, these patients are the ones who are most often considered for magnesium therapy. However, as they are often excluded from studies, there is limited direct evidence to support magnesium therapy for these patients. Lastly, the data was not conducive to determine the specific subgroup of patients most likely to benefit from this treatment. Future large trials are needed to allow analysis of treatment benefits in subgroups classified based on disease severity.

Current guidelines do not recommend intravenous magnesium sulfate for routine mild to moderate asthma exacerbations. They instead recommend considering it only for patients with severe exacerbations or those who have failed initial treatment.2

In summary, for adults presenting to the ED with moderate to severe asthma exacerbations, intravenous magnesium sulfate therapy used as an adjunct to routine treatment (oxygen, short acting beta agonists and systemic corticosteroids) or when these treatments fail, reduces the need for hospitalization and likely has minimal adverse events. Therefore, we have assigned a color recommendation of Green (benefits > harm) to this treatment.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

See theNNT.com's previous reviews of this topic:
Intravenous Magnesium Sulfate Given During an Asthma Attack, January 10, 2010

Author

John Conway, MD; Benajmin Friedman, MD
Supervising Editor: Shahriar Zehtabchi, MD

Published/Updated