Time-series characterisation of cholera outbreaks and optimisation of oral cholera vaccine resource allocation in Africa, 2010–2023

Abstract

Background: Cholera remains a major cause of epidemic disease in Africa, and the Global Task Force on Cholera Control (GTFCC) roadmap targets a 90% reduction in cholera mortality by 2030 through improved water and sanitation, surveillance and targeted oral cholera vaccine (OCV) use. Optimal allocation of a limited global OCV stock across countries with different transmission dynamics is a central operational question, yet quantitative, surveillance-based guidance is scarce. Methods: Using the Johns Hopkins Global Cholera Monitoring Database — a harmonised, location-masked, country-level weekly time series of suspected and confirmed cases and deaths for 34 African countries, 2010–2023 (1·004·144 reported suspected cases) — we characterised national transmission patterns by k-means clustering of incidence, seasonality and persistence features, with hierarchical clustering for cross-validation. We then calibrated a susceptible–infected–recovered–environment (SIRB) transmission model with seasonal forcing and importation to each pattern, and simulated preventive, reactive and mixed OCV strategies (10–50% coverage; vaccine effectiveness 75%). We computed cost-effectiveness (USD per DALY averted; CFR 2%; $2 per dose) and optimised cross-pattern dose allocation by greedy marginal benefit. Results: Three reproducible transmission patterns were identified: persistent-endemic (3 countries; 43·4% of cases), rainy-season-driven (18 countries; 55·4%) and sporadic-imported (10 countries; 1·2%). Preventive annual campaigns (25% coverage) averted 89·1% of cases in persistent-endemic and 77·3% in rainy-season-driven settings, exceeding reactive campaigns (85·5% and 60·6%) at equal coverage; in sporadic-imported settings, even 50% reactive coverage averted only 33·0% of cases. Across full-Africa strategies, ICERs ranged from $9·27/DALY (preventive 10%) to $14·67/DALY (targeted preventive), all far below a $1·600/DALY willingness-to-pay threshold and robust to sensitivity analyses (ICER range $4·7–$51·4/DALY). Greedy allocation concentrated doses in rainy-season-driven and persistent-endemic patterns, sparing sporadic-imported settings, and averted 49·7% more cases than equal per-pattern allocation at the same budget. Conclusion: Transmission-pattern-aware OCV programming — preventive campaigns timed to seasonal peaks in endemic and seasonally driven settings, with reactive response reserved for sporadic importation hotspots — could substantially improve the efficiency of the global cholera stockpile and accelerate progress towards the GTFCC 2030 goals.