Risk Factors for Cholera Case Fatality: A Systematic Review and Meta-Analysis of Demographic, Clinical, and Health-System Determinants, 1961–2026

Comprehensive Overview of the Research Project Purpose and Rationale This research project represents the first systematic review and meta-analysis to quantitatively synthesize the risk factors associated with cholera case fatality across the entire duration of the seventh pandemic from 1961 to the present day. The study emerges from a critical recognition that cholera, despite being a readily treatable disease costing less than one US dollar to manage with oral rehydration solution, continues to claim thousands of lives annually. In 2023 alone, reported cholera deaths rose by 71 percent globally compared with the previous year, with Africa experiencing a case fatality ratio of 1.9 percent, nearly double the World Health Organization's Global Task Force on Cholera Control elimination target of less than 1 percent. Several countries reported case fatality ratios exceeding this threshold, with Chad reaching a staggering 6.8 percent. These figures do not reflect biological inevitability but rather profound health system failures, delayed care seeking, and inequitable access to treatment. The impetus for this investigation comes directly from a 2024 scoping review that examined 77 studies spanning six decades of cholera mortality research. That review explicitly identified critical knowledge gaps and called for a quantitative synthesis to understand which demographic groups, clinical presentations, and health system contexts confer the greatest mortality risk. Despite the treatability of cholera and its designation as a neglected tropical disease targeted for elimination, no previous study has attempted to pool data across settings to produce robust, generalizable estimates of case fatality ratios by risk factor. This project directly addresses that gap. The primary objective is to identify and quantitatively pool the demographic, clinical, and health system risk factors associated with cholera case fatality during the seventh pandemic. Demographic factors include age group and sex, which have been the most frequently reported risk factors yet have shown inconsistent results across different outbreak settings. Clinical factors encompass dehydration severity, comorbidities including malnutrition, pregnancy status, and HIV infection, as well as time from symptom onset to care seeking. Health system factors include whether deaths occur in facilities or in the community, access to antibiotics and oral rehydration solution, healthcare worker training, and facility type. Secondary objectives extend beyond the primary aim to provide a comprehensive understanding of cholera mortality patterns. These include estimating pooled case fatality ratios stratified by age group and sex, quantifying the proportion of community deaths versus facility deaths, assessing health system contributions to excess mortality, identifying comorbidities associated with elevated case fatality where data permit, evaluating temporal trends in case fatality ratios before and after updated World Health Organization case management protocols, and quantifying heterogeneity across geographic regions and outbreak contexts. Methodological Approach The study will follow Preferred Reporting Items for Systematic Review and Meta Analysis Protocols 2015 guidelines for protocol development and will be reported in accordance with PRISMA 2020 and Meta analyses of Observational Studies in Epidemiology guidelines. The protocol will be registered with PROSPERO before commencing database searches, and any deviations from the published protocol will be documented with rationale in the final manuscript. The eligibility criteria are structured around a Population Index factor Factors Outcome Study design framework appropriate for observational studies where no intervention is being evaluated. The population includes patients of all ages with confirmed or suspected Vibrio cholerae O1 or O139 infection across all settings including urban, rural, refugee camps, health facilities, and communities throughout the entire seventh pandemic period. The index factors encompass demographic characteristics including age group and sex, clinical features including dehydration severity, comorbidities, malnutrition, pregnancy status, HIV status, and time to care, and health system factors including facility versus community death, antibiotic access, oral rehydration solution availability, and healthcare worker training. No factors are excluded a priori. The primary outcome is the case fatality ratio defined as the number of cholera deaths divided by the number of cholera cases expressed as a percentage. Secondary outcomes include time to death and the proportion of deaths occurring in the community versus health facilities. The study design includes all primary empirical studies reporting at least two fatal cholera cases with at least one stratified predictor. Eligible designs include outbreak reports, cohort studies, case control studies, cross sectional surveys, and surveillance studies. Animal studies, modelling studies, and outbreak studies from before 1961 will be excluded. Additional exclusion criteria include studies referring solely to non O1 or non O139 Vibrio cholerae, duplicate outbreak reports where the most complete report will be retained, studies reporting case fatality without stratification by predictor which will be included for the overall estimate only and coded separately, and case series with fewer than two deaths. The search strategy encompasses nine databases and repositories including PubMed MEDLINE as the primary biomedical index using MeSH terms and free text, EMBASE for European coverage using Emtree and free text, Web of Science for citation tracking, Scopus for broad biomedical coverage, LILACS for Latin American and Caribbean literature using DeCS Spanish and Portuguese vocabulary, African Journals Online for Sub Saharan African grey and peer reviewed literature, the Cochrane Library for reviews and trials, OpenGrey and SIGLE for unpublished reports, and institutional repositories including the WHO Institutional Repository for Information Sharing, ReliefWeb, and the Global Task Force on Cholera Control. The core PubMed MEDLINE search string combines cholera and Vibrio cholerae terms with case fatality rate, mortality, CFR, case fatality, death, and fatal terms, and with risk factors, predictors, determinants, sex, age, comorbidity, malnutrition, pregnancy, and health services accessibility terms. Supplementary search methods include backward snowballing of reference lists, forward citation tracking of the 2025 scoping review, hand searching of Lancet Infectious Diseases, Tropical Medicine and International Health, and PLOS Neglected Tropical Diseases from 2015 to 2025, and systematic grey literature browsing of the Global Task Force on Cholera Control, WHO Institutional Repository for Information Sharing, ReliefWeb, Médecins Sans Frontières Science Portal, and Epicentre publications. The study selection process will use Rayyan for title and abstract screening and Covidence for full text review. Two reviewers will independently screen at all stages with a target Cohen's kappa of at least 0.80 before proceeding to full text review. Conflicts will be resolved by discussion or by a third reviewer. A PRISMA 2020 compliant flow diagram will be produced at each stage including reasons for all full text exclusions. Data extraction will use a standardised form piloted on five studies before full scale use with pilot success defined as at least 80 percent agreement between reviewers on all extracted fields. Two reviewers will extract independently with discrepancies reconciled by discussion. Extraction will be performed in a pre specified spreadsheet with one row per study and one tab per domain. Extracted variables include study identifiers such as author, year, country, region, setting, study design, and outbreak versus endemic status. Case definition variables include laboratory confirmed versus suspected, serogroup, and biotype. Population variables include total number, age distribution, sex, and special populations. Case fatality ratio outcomes include overall, by age, by sex, with 95 percent confidence intervals, deaths, and cases. Place of death variables include facility versus community, deaths on arrival, and distance to care. Clinical factors include dehydration severity, time to care, comorbidities, antibiotics, and oral rehydration solution. Health system factors include facility type, staffing, and surge capacity. Risk of bias items include case definition clarity, laboratory confirmation, age and sex in deaths, numerator and denominator stated, and selection bias. Risk of bias will be assessed using design appropriate tools. Outbreak reports and cross sectional surveillance studies will be evaluated using the AXIS critical appraisal tool. Cohort and case control studies will be assessed using the Newcastle Ottawa Scale. Institutional and programmatic reports will be evaluated using GRADE. The minimum quality threshold for meta analysis inclusion requires a clear cholera case definition, explicit case fatality ratio numerator and denominator, and at least one predictor reported in stratified fashion. Studies failing any criterion will be included in narrative synthesis only. Sensitivity analyses for risk of bias will restrict to laboratory confirmed cases only, restrict to low risk studies only, and restrict to post 2000 studies. Statistical Analysis and Meta Analysis The statistical approach will employ random effects meta analysis using the DerSimonian Laird method given the anticipated high heterogeneity across settings, eras, and case definitions. Because case fatality ratio is a proportion, the Freeman Tukey double arcsine transformation will be applied before pooling with back transformation for presentation. Results will be reported as pooled case fatality ratio percentages with 95 percent confidence intervals. Heterogeneity will be assessed using the I squared statistic with values below 25 percent indicating low heterogeneity, 25 to 75 percent indicating moderate heterogeneity, and above 75 percent indicating high heterogeneity. Cochran's Q test with a threshold of p less than 0.10 will also be used, and tau squared will measure between study variance. If I squared exceeds 75 percent, pooled estimates will be presented with caution and explored via subgroup analyses and meta regression. Pre specified subgroup analyses will examine geographic region including Sub Saharan Africa, South Asia, Middle East and North Africa, and the Americas. Temporal era will be categorised as pre 2000, 2000 to 2010, and 2011 to present. Setting will be analysed as cholera treatment centre or hospital versus community. Case definition rigor will be compared as confirmed versus suspected cases. Age groups will include less than 5 years, 5 to 14 years, 15 to 49 years, and 50 years and older. Sex will be analysed as male versus female. Outbreak magnitude will be categorised as greater than 10,000 cases versus 10,000 or fewer cases. Conflict or humanitarian setting will be analysed as yes versus no. Meta regression will model case fatality ratio as a function of year of outbreak, the proportion laboratory confirmed, region, and setting if at least 10 studies are available per stratum. Continuous moderators will be mean centred. Publication bias will be assessed via funnel plot asymmetry using visual inspection and Egger's test if at least 10 studies are available. Trim and fill adjustment will be applied if asymmetry is detected. The systematic grey literature search is designed to minimise publication bias. All analyses will be conducted in R version 4.3 using the meta, metafor, and dmetar packages with Stata 18 for secondary validation. Forest plots will be produced in R and exported at 600 dots per inch. GRADE will be used to rate the certainty of evidence for each key finding as high, moderate, low, or very low. Domains assessed include risk of bias, inconsistency, indirectness, imprecision, and publication bias. A GRADE evidence profile will be produced as a summary table. Expected Outcomes Based on the comprehensive scoping review that preceded this project, we anticipate identifying approximately 80 to 120 studies for qualitative synthesis and 50 to 80 for quantitative meta analysis. The included studies are expected to span 30 to 50 countries across all six WHO regions with study designs including outbreak reports, surveillance studies, cohort analyses, and case control investigations. We anticipate significant heterogeneity with I squared exceeding 75 percent for most pooled estimates, driven by regional differences, temporal evolution of case definitions and treatment protocols, and variation in facility versus community based surveillance. Meta regression will quantify the independent contributions of year, laboratory confirmation rate, and WHO region. Funnel plot asymmetry may indicate underreporting of high case fatality ratio outbreaks or community based studies with limited publication capacity. Trim and fill will adjust estimates if asymmetry is detected. Sensitivity analyses restricting to laboratory confirmed cases may reduce pooled case fatality ratio by testing case definition bias. Restricting to low risk studies with AXIS scores of 8 or higher will test robustness against methodological limitations. Restricting to post 2000 studies will test protocol era effects. We anticipate that outbreak reports will score moderately on AXIS with a median around 6 out of 11, reflecting incomplete reporting of denominators and selection processes. Cohort and case control studies are expected to score higher on the Newcastle Ottawa Scale with a median around 7 out of 9. The findings from this meta analysis will provide several key contributions to the field. If the analysis confirms recent findings from Zambia showing male excess mortality with a hazard ratio of 1.89 and dramatically elevated risk in adults aged 50 years and older with a hazard ratio of 6.02, these groups should be prioritised for rapid rehydration and close monitoring in resource limited settings. The mechanisms may be biological with higher comorbidity burden in males and reduced physiological reserve in older adults, or behavioural with delayed care seeking. Quantifying the community death proportion anticipated to be 40 to 60 percent based on the scoping review will inform decisions on decentralised treatment strategies and community health worker deployment. The Global Task Force on Cholera Control 2030 roadmap targets a case fatality ratio below 1 percent globally. With several countries currently exceeding this threshold, our findings will have immediate policy relevance. If health system factors such as care delay, facility access, and oral rehydration solution availability explain more variance than biological factors, the policy response must shift from clinical training alone to infrastructure investment and decentralised care. The temporal analysis comparing case fatality ratios before and after updated WHO case management protocols will provide ecological evidence on whether guideline changes have translated into mortality reductions. Strengths and Limitations This review has several methodological strengths. It is the first quantitative synthesis of cholera mortality determinants. It spans six decades with no language restrictions. It includes grey literature to minimise publication bias. It uses pre specified subgroups and sensitivity analyses. It will provide GRADE certainty ratings for each key finding. The inclusion of institutional and programmatic reports ensures representation of outbreak contexts where peer reviewed publication is limited. However, several limitations are anticipated. Observational data predominance carries inherent risk of residual confounding. Sparse comorbidity data may preclude pooled estimates for HIV, malnutrition, and pregnancy. High anticipated heterogeneity may leave some variance unexplained despite extensive subgroup analyses. Potential publication bias may persist despite grey literature inclusion. Evolving case definitions and treatment protocols over six decades may introduce temporal confounding that meta regression can only partially address. Priority Areas for Future Research This project will identify several priority areas for future research. Primary data collection on comorbidities in outbreak settings using standardised instruments is needed. Intervention studies targeting modifiable health system factors such as care delay and community oral rehydration solution availability should be prioritised. Prospective standardisation of case fatality ratio reporting across outbreak surveillance systems would greatly enhance future syntheses. Conclusion This systematic review and meta analysis will provide the first robust, generalisable estimates of cholera case fatality by risk factor. The findings will inform clinical practice, outbreak response, and policy toward the WHO and Global Task Force on Cholera Control goal of eliminating cholera deaths by 2030. As an open access publication, these findings will reach the outbreak responders, clinicians, and policymakers who need them most. The research represents a significant step toward understanding why cholera remains a lethal disease despite being readily treatable, and how resources can be most effectively deployed to save lives in the world's most vulnerable populations.