RESUMEN
Background: Diagnosis of cutaneous leishmaniasis (CL) usually relies on invasive samples, but it is unclear whether more patient-friendly tools are good alternatives for diverse lesions when used with polymerase chain reaction (PCR). Methods: Patients with suspected CL were enrolled consecutively in a prospective diagnostic accuracy study. We compared dental broach, tape disc, and microbiopsy samples with PCR as index tests, using PCR with skin slit samples as reference test. Subsequently, we constructed a composite reference test including microscopy, the 3 index tests and skin slit PCR, and we compared these same tests with the composite reference test. We assessed diagnostic accuracy parameters with 95% confidence intervals for all comparisons. Results: Among 344 included patients, 282 (82.0%) had CL diagnosed, and 62 (18.0%) CL absence, by skin slit PCR. The sensitivity and specificity by PCR were 89.0% (95% confidence interval, 84.8%-92.1%) and 58.1% (45.7%-69.5%), respectively, for dental broach, 96.1% (93.2%-97.8%) and 27.4% (17.9%-39.6%) for tape disc, and 74.8% (66.3%-81.7%) and 72.7% (51.8%-86.8%) for microbiopsy. Several reference test-negative patients were consistently positive with the index tests. Using the composite reference test, dental broach, and skin slit had similar diagnostic performance. Discussion: Dental broach seems a less invasive but similarly accurate alternative to skin slit for diagnosing CL when using PCR. Tape discs lack specificity and seem unsuitable for CL diagnosis without cutoff. Reference tests for CL are problematic, since using a single reference test is likely to miss true cases, while composite reference tests are often biased and impractical as they require multiple tests.
RESUMEN
BACKGROUND: Cutaneous leishmaniasis (CL) is common in Ethiopia, mainly affecting impoverished populations in rural areas with poor access to health care. CL is routinely diagnosed using skin slit smear microscopy, which requires skilled staff and appropriately equipped laboratories. We evaluated the CL Detect Rapid Test (InBios, Washington, USA), which is supplied with a dental broach sampling device, as a diagnostic alternative which could be used in field settings. METHODOLOGY/PRINCIPAL FINDINGS: We evaluated the diagnostic accuracy of the CL Detect Rapid Test on skin slit and dental broach samples from suspected CL patients at the Leishmaniasis Research and Treatment Center in Gondar, Ethiopia. A combined reference test of microscopy and PCR on the skin slit sample was used, which was considered positive if one of the two tests was positive. We recruited 165 patients consecutively, of which 128 (77.6%) were confirmed as CL. All microscopy-positive results (n = 71) were also PCR-positive, and 57 patients were only positive for PCR. Sensitivity of the CL Detect Rapid Test on the skin slit was 31.3% (95% confidence interval (CI) 23.9-39.7), which was significantly higher (p = 0.010) than for the dental broach (22.7%, 95% CI 16.3-30.6). Sensitivity for both methods was significantly lower than for the routinely used microscopy, which had a sensitivity of 55.5% (IQR 46.8-63.8) compared to PCR as a reference. CONCLUSIONS/SIGNIFICANCE: The diagnostic accuracy of the CL Detect Rapid Test was low for skin slit and dental broach samples. Therefore, we do not recommend its use neither in hospital nor field settings. TRIAL REGISTRATION: This study is registered at ClinicalTrials.gov as NCT03837431.
