Decentralised rapid diagnostic tests enable cholera diagnosis by non-laboratory health workers during outbreaks in Cameroon

Cholera has re-emerged as a major global public health emergency, with reported cases and deaths tripling between 2022 and 2025 especially in Africa, due to climate shocks, population displacement, fragile water and sanitation systems, and delays in laboratory confirmation. Early detection is critical to interrupt transmission and reduce mortality, underscoring the need for decentralised point-of-care diagnostics. We conducted a prospective field evaluation of three cholera rapid diagnostic tests (RDTs): SD Bioline Cholera Ag O1/O139, Crystal VC O1, and Cholkit, during active cholera outbreaks in Cameroon. Tests were performed on fresh stool samples by laboratory technicians and by non-laboratory health workers at primary health-care facilities. Diagnostic performance was assessed using PCR as the reference standard. In a subset of samples, RDTs were also performed after alkaline peptone water (APW) enrichment to assess its effect on diagnostic accuracy. Inter-operator agreement and ease of use were evaluated. Among 492 suspected cholera cases enrolled, 377 samples had PCR results available for analysis. When performed at the point of care, RDT sensitivity ranged from 88% to 95%, comparable to laboratory-based testing, while specificity ranged from 72% to 85%. APW enrichment was associated with a consistent reduction in sensitivity across all three RDTs and reduced specificity for SD Bioline. Inter-operator agreement was high (Cohen’s κ 0·81–0·89). More than 80% of end users reported that RDTs were easy to use under outbreak conditions. Cholera RDTs demonstrated high sensitivity, strong inter-operator reliability, and operational feasibility when deployed directly on fresh stool samples by non-laboratory staff. Integrating RDTs into decentralised surveillance systems, without APW enrichment, can accelerate outbreak detection and support timely response in high-burden, resource-limited settings.