Search by:

Source

Johari F. Does this patient have Lyme disease? Academic Emergency Medicine. 2025;32(7):807-809.

Study Population: Thirteen trials assessing the diagnostic accuracy of a two-tier serological test protocol versus clinical diagnosis

Narrative

Lyme disease (LD), caused by Borrelia burgdorferi, is the most common tick-borne illness in the U.S., with approximately 40,000 cases reported annually—though the true number is likely higher due to underreporting.1 LD progresses through three stages: early localized, early disseminated, and late. In the early localized stage, LD typically presents with erythema migrans—a red, expanding rash at the tick bite site—accompanied by systemic symptoms such as malaise, headache, fever, myalgia, and arthralgia. While most cases remain in this stage, about 20% progress to early disseminated disease, characterized by multiple erythema migrans lesions, flu-like symptoms, lymphadenopathy, cranial nerve palsies (especially CN VII), lymphocytic meningitis, and cardiac manifestations like conduction abnormalities or myocarditis. Late-stage LD often presents as pauciarticular arthritis, primarily affecting large joints such as the knees. With timely and appropriate treatment, LD is curable.2

With timely and appropriate treatment, LD is curable with a course of antibiotics. Early localized LD is typically treated with a 10- to 14-day course of oral doxycycline, amoxicillin, or cefuroxime. In cases of disseminated or late-stage disease, treatment duration may be extended, and in some cases intravenous ceftriaxone may be required for severe neurological or cardiac involvement.3

The quality of evidence regarding the sensitivity and specificity of signs and symptoms of LD is generally very low due to methodological limitations in the available studies.3 Among these signs and symptoms, erythema migrans exhibits the highest sensitivity and specificity. As the characteristic “bull's-eye” rash, erythema migrans serves as a key clinical indicator of LD. Its sensitivity varies, being lower in children (40%) than in adults (67%), while maintaining high specificity in both groups (99% in children and 88% in adults).4

The diagnosis of LD is based on clinical presentation, epidemiological history, and serologic testing, with test interpretation depending on symptom onset. To minimize false positives and negatives, testing is recommended only for patients with typical symptoms and exposure to Lyme-endemic areas. Misdiagnosis can lead to unnecessary treatments, underscoring the importance of accurate clinical evaluation.1 The Association for Public Health Laboratories (APHL) has developed guidelines to enhance test interpretation and reporting, focusing on the standard and modified two-tier testing algorithms with recommendations for clear, standardized result communication.1

The systematic review and meta-analysis discussed here examined 485 global studies on LD diagnosis, narrowing the focus to 48 North American studies (1995–2013). The analysis assessed the diagnostic utility of various tests, with this report specifically addressing the meta-analysis of 13 studies evaluating the two-tier serological test protocol for LD diagnosis at different disease stages.5

The two-tier serologic testing algorithm (see Figure 1) recommends an initial enzyme immunoassay or immunofluorescence assay, followed by a Western immunoblot for positive or equivocal results.4^, 5 The meta-analysis demonstrated that this algorithm has low sensitivity (46%, 95% CI 39%–54%) in early-stage LD (Stage 1) but higher sensitivity in Stage 2 (90%, 95% CI 78%–95%) and Stage 3 (99%, 95% CI 96%–99%). The specificity of the algorithm is high across all stages (98%–99%). The algorithm's negative likelihood ratio (LR–) is sufficient to rule out the disease only in Stage 3 LD (LR– 0.006).5


Figure 1
Standard two-tiered testing for diagnosis of Lyme disease recommended by the Centers for Disease Control and Prevention.1

Caveats

The Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool used in the systematic review identified an unclear risk of bias in 84% of studies, citing issues such as inappropriate sample population enrollment, lack of blinding, and unexplained data exclusions. Additionally, 29% of studies had financial ties to commercial test suppliers, with four exhibiting a high risk of funding bias. Despite these limitations, no studies were excluded based on QUADAS assessments.5

The generalizability of test performance may be affected by regional variations in LD prevalence and Borrelia species. Results from high-prevalence areas may not accurately reflect test accuracy in low-incidence regions with different Borrelia strains. Many diagnostic tests have reduced sensitivity in early-stage LD, as antibodies may not yet be detectable. Furthermore, comparisons between FDA-cleared tests and in-house assays may introduce variability due to differences in test conditions, standardization, and result interpretation. A lack of uniformity in defining positive and negative results across studies further contributes to potential bias.1^, 6^, 7

Overreliance on serologic testing without incorporating clinical judgment could lead to misdiagnosis or missed diagnoses. Additionally, the study may not fully address emerging Borrelia species that could impact test accuracy.5

In summary, the two-tiered testing algorithm remains the standard approach for LD diagnosis,8 with sensitivity improving in later disease stages and high specificity maintained across all stages. To minimize false positives and negatives, testing should be reserved for patients with typical symptoms and a history of exposure to Lyme-endemic areas.

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

Author

Fatima Johari, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

August 14, 2025

What are Likelihood Ratios?

LR, pretest probability and posttest (or posterior) probability are daunting terms that describe simple concepts that we all intuitively understand.

Let's start with pretest probability: that's just a fancy term for my initial impression, before we perform whatever test it is that we're going to use.

For example, a patient with prior stents comes in sweating and clutching his chest in agony, I have a pretty high suspicion that he's having an MI – let's say, 60%. That is my pretest probability.

He immediately gets an ECG (known here as the "test") showing an obvious STEMI.

Now, I know there are some STEMI mimics, so I'm not quite 100%, but based on my experience I'm 99.5% sure that he's having an MI right now. This is my posttest probability - the new impression I have that the patient has the disease after we did our test.

And likelihood ration? That's just the name for the statistical tool that converted the pretest probability to the posttest probability - it's just a mathematical description of the strength of that test.

Using an online calculator, that means the LR+ that got me from 60% to 99.5% is 145, which is about as high an LR you can get (and the actual LR for an emergency physician who thinks an ECG shows an obvious STEMI).

(Thank you to Seth Trueger, MD for this explanation!)