Improving estimation of vaccine effectiveness during outbreaks in low-resource settings: A case study of oral cholera vaccination during the 2022-2023 cholera outbreak in Malawi

Résumé

Abstract Background Use of oral cholera vaccine (OCV) is globally recommended as a public health response to cholera outbreaks, alongside water, sanitation and hygiene (WASH) interventions. Estimating vaccine effectiveness during emergencies in low- and middle-income countries is challenging because vaccination campaigns are often implemented over short time frames, while individual-level data are frequently incomplete due to constraints in infrastructure, resources and data systems. There is a need for pragmatic approaches that can generate timely, policy-relevant evidence using routinely collected data. Methods We analysed routine surveillance data from a large 2022–2023 cholera outbreak in Blantyre District, Malawi. The EpiEstim framework was used to generate estimates of the time-varying reproduction number (Rₜ) from line-listed case data. We modelled changes in 𝑅 𝑡 as a function of cumulative OCV coverage using a log-linear framework and propagated uncertainty through posterior sampling. Lagged WASH exposure variables were incorporated in the model to generate adjusted vaccine effectiveness estimates and to explore potential interaction effects. Sensitivity analyses assessed robustness to alternative lag structures. Findings The Blantyre outbreak was characterised by an initial period of low-level transmission followed by a sharp increase in cases from late November 2022, after which transmission declined steadily through April 2023. This decline coincided with the implementation of a reactive OCV campaign. The majority of the cases were among middle-aged men living in urban Blantyre. The unadjusted vaccine-associated reduction in transmission was estimated at 53.52% (95% credible interval (CrI):42.5–64.1%). After adjusting for a 7-day rolling average WASH activity, total vaccine effectiveness increased to 62.1% (95% CrI: 49.3–74.9%). Sensitivity analyses using alternative lag structures for WASH exposure produced comparable adjusted estimates. Interpretation Implementation of OCV contributed to a substantial reduction in cholera transmission during the outbreak. This study demonstrates a feasible approach for estimating vaccine-attributable impact whilst accounting for public health and social measures, such as WASH interventions. The methods described will be useful in outbreaks where classical observational designs are not possible, providing actionable evidence to policy makers for outbreak response in resource-limited settings. Funding MRC Discovery Medicine North (DiMeN) Doctoral Training Partnership (UKRI), National Institute for Health and Care Research (NIHR) Global Health Research Group on Gastrointestinal Infections and Wellcome through the core grant to the Malawi-Liverpool-Wellcome Research Programme. Research in context Evidence before this study Before undertaking this analysis, our team and collaborators generated detailed epidemiological and genomic evidence describing the 2022–2023 cholera outbreak in Blantyre, Malawi. Our previous fine-scale spatial work demonstrated marked heterogeneity in cholera burden across urban Blantyre. We found that transmission concentrated in densely populated informal settlements and areas with poor water and sanitation infrastructure. Our genomic analysis of outbreak isolates showed that the epidemic was driven by a recently introduced seventh-pandemic El Tor Vibrio cholerae sub-lineage AFR15 linked to regional and international transmission events, with epidemic expansion occurring during late 2022. These studies characterised where transmission occurred and the evolutionary origin of the outbreak strain. The population-level effect of reactive oral cholera vaccination (OCV) or concurrent and improved water, sanitation and hygiene (WASH) interventions on transmission dynamics during the outbreak was not quantified. In linked work, we conducted a systematic review of post-licensure vaccine impact and effectiveness studies from sub-Saharan Africa (CRD42023436851) to assess the principal study designs and the extent to which they adjusted for concurrent public health and social measures (PHSMs), such as WASH interventions. We searched PubMed, EMBASE, MEDLINE, CINAHL, and Google Scholar for vaccine impact or effectiveness studies conducted in children under five years and screened reference lists of included studies. Across all eligible studies, none measured or adjusted for concurrent PHSMs. Added value of this study Using surveillance, vaccination, and WASH data collected from Blantyre district by the Malawi Ministry of Health, we estimated the vaccine-associated reduction in transmission during a large reactive OCV campaign. By linking time-varying R ₜ estimates to cumulative vaccine coverage within a posterior sampling framework, we provide field-relevant effectiveness estimates from a setting where traditional individual-level designs were not feasible. We further evaluated whether concurrent WASH activity materially altered vaccine effect estimates. Implications of all the available evidence Taken together with prior spatial and genomic analyses of the same outbreak, our findings provide a more complete picture of cholera control in Malawi. The earlier work clarified transmission geography and pathogen introduction. The present analysis quantifies the population-level impact of reactive vaccination under real-world programme conditions. This integrated evidence base supports ministries of health in interpreting vaccine performance during rapidly evolving outbreaks and in strengthening the routine data systems required for timely evaluation of interventions.

Citer ce document

Accès au document

Texte intégral en lecture en ligne, réservé aux abonnés SPHAERO et aux membres de l'institution. Se connecter

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0