Abstract:Background Accurate and reliable diagnostics, including rapid diagnostic tests (RDTs), are critical components of cholera control programs, though their performance has varied greatly across studies. While poorly understood, this variability may be due to the reference assay choice, patient-level and/or sampling characteristics, which hinder test result interpretation and performance evaluation. Methods We enrolled all suspected cholera cases seeking care at two healthcare facilities in Sitakunda, Bangladesh over 19 months. All stool samples were tested with the Cholkit RDT, and a subset by PCR and culture. Test performance was estimated using a latent-class Bayesian framework accounting for imperfect test performance, incomplete PCR and culture testing, and time-varying changes in cholera incidence. Patient-level (including age, antibiotic use) and sampling (season, testing delays) factor effects were estimated, and simulations were used to assess the bias in RDT performance estimates when using traditional reference assays. Findings We enrolled 3,744 suspected cases, 692 of whom were RDT-positive. Among the RDTpositives, 573 were PCR-positive and 450 culture-positive. For RDT, PCR and culture, we estimated a sensitivity of 93.5% (95% Credible Intervals, CrI: 91.3-95.4), 90.3% (88.4-92.1), and 73.7% (70.8-76.5), and a specificity of 97.3% (96.7-97.8), 97.2% (96.6-97.8), and 100% (culture specificity assumed perfect), respectively. We found that younger age (≤ 5), antibiotic use, and testing delays decreased culture sensitivity, but RDT performance remained relatively constant. The RDT positive predictive value ranged from <15% in children <5 years to >80% in adults, varying greatly across seasons. Simulations demonstrated underestimation of RDT sensitivity and specificity in low and high cholera prevalence settings, respectively, when evaluated against PCR or culture. Interpretation Our results shed light on the potential mechanisms leading to heterogeneous cholera RDT performance estimates in previous studies, including the use of culture as a reference assay. Across various patient and sampling characteristics, Cholkit RDT had high performance in this cholera-endemic setting, supporting its use for cholera surveillance and control. Accounting for epidemiologic context is crucial both for individual-level clinical test interpretation, and for the future evaluation of diagnostics like RDTs.

Rethinking cholera diagnostic test performance, interpretation and evaluation: a field-based latent-class analysis in Bangladesh