Whispers of the Outbreak: Modeling Hidden Cholera Transmission in Informal Settlements

Ioanna Kyriazis¹, Sophie Turner², Nikos Papadopoulos³

Abstract

Background:
Cholera outbreaks in informal settlements often remain under-detected due to poor surveillance, unregulated migration, and limited access to testing. Conventional surveillance may overlook the true burden and transmission routes of cholera in such high-density, underserved environments.

Objective:
To develop a predictive spatial-temporal model of cholera transmission in informal urban settlements, identifying hidden clusters and evaluating key transmission drivers.

Methods:
A retrospective analysis was conducted in two major informal settlements in Dhaka and Lagos between January 2018 and December 2023. Reported cholera cases (n = 1,728) were geo-tagged, and wastewater sampling data were integrated with demographic and environmental indicators (e.g., sanitation coverage, water source type, rainfall patterns). A Bayesian spatio-temporal model was used to infer hidden transmission pathways. Model outputs were compared against known outbreaks for validation.

Results:
Modeling revealed 3 previously undetected transmission hotspots with high posterior probabilities (>0.9). These areas had poor sanitation access (<35%) and population densities >50,000/km². Rainfall intensity was a strong predictor of cluster emergence (p < 0.001), with transmission peaks following rainfall spikes by 7–10 days. Integration of wastewater signals increased detection sensitivity by 26.3%. Areas missed by standard surveillance accounted for an estimated 41.5% of total cases.

Conclusion:
Cholera transmission in informal settlements is shaped by complex, underrecognized spatial patterns. Traditional case reporting underestimates the burden. Integration of wastewater monitoring and probabilistic modeling enhances early detection and can guide targeted intervention zones. Investing in data infrastructure and real-time geospatial tools may dramatically improve outbreak containment.