Abstract
<title>Abstract</title> <p> Background Cholera remains a persistent global public health threat, heavily driven by climate change and infrastructural vulnerabilities. We analyzed 203,201 suspected cholera cases and 277 deaths reported in Amran Governorate, Yemen, from 2016–2025 to characterize long-term transmission dynamics, environmental drivers, and spatial vulnerabilities. Methods We conducted a retrospective surveillance study using the electronic Integrated Disease Surveillance and Response (eIDSR) database, including 203,201 suspected cholera cases and 277 deaths reported between September 2016 and May 2025. Bayesian estimation of the time-varying effective reproduction number (Rt) and joinpoint regression were used to characterize epidemic dynamics. Environmental associations were assessed using distributed lag nonlinear models (DLNMs), seasonal autoregressive integrated moving average models with exogenous variables (SARIMAX), and negative binomial regression. Spatial heterogeneity and WASH vulnerability were evaluated using geographic information system–based analyses, Global Moran's I, Local Indicators of Spatial Association (LISA), and a composite vulnerability index. Multivariable logistic regression identified predictors of mortality and characteristics associated with the recent resurgence. Results The epidemic evolved from explosive transmission during Wave 2 (2017–2018) to persistent endemic circulation, followed by a substantial urban-centered resurgence during 2023–2025. Weekly rainfall independently predicted cholera incidence after a 1-week lag (incidence rate ratio = 1.011, 95% confidence interval: 1.007–1.016; <italic>P</italic> < 0.001), while the resurgence was preceded by an extreme drought anomaly. Distributed lag analyses demonstrated the greatest transmission risk following intense rainfall events with short lag periods. Spatial analyses identified a WASH severity paradox, whereby highly deprived rural districts reported lower cholera incidence than urban districts despite greater infrastructural vulnerability, suggesting substantial surveillance underascertainment. Mortality was independently associated with severe dehydration (adjusted odds ratio = 104.97, 95% confidence interval: 33.40–329.96) and age > 60 years (adjusted odds ratio = 6.82, 95% confidence interval: 4.51–10.31). Compared with Wave 2, the recent resurgence disproportionately affected urban residents, older adults, and patients presenting with more severe dehydration. Conclusions The persistence of cholera in Yemen is increasingly shaped by climatic variability, urban vulnerability, and surveillance inequities. Strengthening climate-resilient WASH infrastructure and targeted surveillance in underserved districts is essential. </p>