Systemic Corticosteroids for Acute Exacerbations of Chronic Obstructive Pulmonary Disease (COPD)

Reduce treatment failure and length of stay; increase medication side effects

Benefits in NNT

1 in 9 were helped (treatment failure prevented)
12.2% lower risk of treatment failure

Harms in NNT

1 in 6 were harmed (adverse effect associated with corticosteroids)
19.6% increase in all adverse effects associated with corticosteroid treatment
View As:

Efficacy Endpoints

Treatment failure within 30 days, relapse within 1-4 months, 30-day mortality, and mean length of hospital stay

Harm Endpoints

Side effects of corticosteroid


Chronic obstructive pulmonary disease (COPD), currently the fourth leading cause of death worldwide, is characterized by respiratory symptoms and airflow limitation often caused by long-term exposure to noxious particles or gases, most commonly tobacco.1, 2 Symptoms include dyspnea, cough, and sputum production, and episodes of acute worsening are termed exacerbations. The most common cause of exacerbations is infection, and treatments may include antibiotics, bronchodilators, and systemic corticosteroids. The latter are believed to decrease lower airway obstruction by reducing airway edema.

The Cochrane review3 summarized here analyzed 16 randomized trials of 1787 subjects (mean age 68, 80% male) comparing corticosteroid to placebo for COPD exacerbation. Patients with COPD of any severity were recruited from outpatient, inpatient, critical care, or emergency department settings for “acute functional deterioration” consistent with exacerbation. Corticosteroids were oral prednisone, or intravenous prednisolone, methylprednisolone, or hydrocortisone. The length of treatment varied from 3 to 19 days; only one study extended beyond 14 days.

The three primary outcomes analyzed were treatment failure (≤30 days), relapse (>30 days), and 30-day mortality. Secondary outcomes included length of hospital stay, adverse effects, and hyperglycemia.

Treatment failure was defined as an increase in pharmacologic treatment, hospital admission, or return emergency department visit. The analysis involved nine studies and showed a significant decrease with corticosteroids (odds ratio [OR]: 0.48: 95% confidence interval [CI], 0.35 to 0.67; absolute risk difference [ARD]: 12.2%; number-needed-to-treat [NNT]: 9; quality of evidence: high). Analysis of three studies comparing oral to parenteral corticosteroids showed no significant difference.

Relapse between 1-4 months was reported in five studies showing no significant difference between corticosteroid and placebo groups. Thirty-day mortality was assessed in 11 studies and also showed no difference.

A secondary endpoint, hospital stay, was reported in two studies and showed a mean decrease of 1.22 days (95% CI -2.26 to -0.18) among those in the general inpatient setting treated with corticosteroids.

Finally, utilizing 8 studies, the Cochrane review also reported on adverse effects with corticosteroid use. These included a composite of hyperglycemia, GI bleeding, dyspepsia, weight gain, depression, anxiety, psychiatric disorder, insomnia, delirium, secondary infection, and ventilatory associated pneumonia. The effects occurred twice as frequently with corticosteroids as compared to placebo (OR: 2.33, 95% CI, 1.59-3.43; ARD: 19.6%; NNH: 6; quality of evidence: high).


The quality of the trials included in the Cochrane review was judged to be good overall. For studies assessing primary endpoints and adverse effects, heterogeneity was minimal, while in two studies reporting hospital length of stay there was substantial heterogeneity. It should be noted that this analysis included studies of ED patients, hospital inpatients, and ICU patients, clinically a very heterogenous group. Though all patients suffered from the same disease process, the effect of corticosteroids may well vary over the spectrum of disease severity. This is an area where further studies would be beneficial.

Among the 16 studies included, there was significant variation in corticosteroids used, doses prescribed, and duration of treatment. However, in the most recent 2020 Gold report1 5-7 days of corticosteroid treatment is considered adequate. Moreover, while the meta-analysis showed no survival benefit of corticosteroids, there appeared to be a clear benefit in preventing treatment failure.

One in six patients who received corticosteroids experienced adverse effects, however severity of the events were not described. Subgroup analysis revealed that one in seven were noted to be hyperglycemic though this resolved with cessation of the drug. Short-term hyperglycemia is not a meaningful patient-centered effect. It is reasonable to expect that the difference in the reported composite outcome would not persist if hyperglycemia were not included.

In conclusion, the Cochrane review discussed here quantifies reported harms and benefits associated with corticosteroids in acute COPD exacerbation. Short-course corticosteroids reduced short-term (2-30 day) treatment failure, and patient-centered adverse events appeared to be uncommon. In critical care settings, and with use of intravenous corticosteroids, the adverse event profile may be more significant, though these data are limited. We have thus assigned a color recommendation of Green (benefits > harms) to the use of corticosteroids in the treatment of acute exacerbations of COPD.

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

See's previous reviews of this topic:
Systemic Steroids for Acute COPD Exacerbations, August 19, 2010


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


May 1, 2020