Protection from killed whole-cell cholera vaccines: a systematic review and meta-analysis

Abstract

Background: Killed whole-cell oral cholera vaccines (kOCVs) are a standard prevention and control measure in cholera-endemic areas and during outbreaks and humanitarian emergencies. New evidence has emerged and the ways in which the vaccines are used have changed. We aimed to provide an updated synthesis of evidence on protection conferred by kOCV.

Methods: In this systematic review and meta-analysis, we used the same search procedure as a previous systematic review to identify randomised clinical trials (RCTs) and observational studies that reported estimates of protection conferred by kOCVs against medically attended, confirmed cholera. Eligible studies in English, French, Spanish, or Chinese published up until March 8, 2024, including those identified in the previous review, were included. Data on efficacy and effectiveness were extracted, as were the number of doses, duration of follow-up, and age group. Efficacy and effectiveness estimates were summarised separately using random-effect models to estimate protection by time since vaccination; meta-regression models were used to estimate protection, by dose, as a function of time since vaccination. This updated study is registered along with the original review with PROSPERO (CRD42016048232).

Findings: We identified 8205 records published online up until March 8, 2024, including 6224 articles from the previous review and 1981 articles from our new search (after Jan 1, 2016). Of these, 53 were eligible for full-text review. Five RCTs and ten observational studies from 23 publications were included. Average two-dose efficacy 12 months after vaccination was 55% (95% CI 46-62), declining to 44% (25-59) 48 months after vaccination. Average two-dose effectiveness was 69% (58-78) 12 months after vaccination, declining to 47% (9-70) 48 months after vaccination. Only one RCT assessed one-dose efficacy and found sustained protection for 24 months (57% [42-69]) among those 5 years and older with no significant protection in younger children. Average one-dose effectiveness 12 months after vaccination was 60% (51-68) and after 24 months was 47% (34-58). Using age group-specific meta-analysis, we found that average two-dose efficacy in children younger than 5 years was half that of older individuals.

Interpretation: Two doses of kOCV provide protection against medically attended cholera for at least 4 years after vaccination. One dose of kOCV provides protection for at least 2 years after vaccination, but wanes faster than that of two doses. Children younger than 5 years are less protected by kOCVs than those aged 5 years and older, regardless of the number of doses received.

Funding: Bill & Melinda Gates Foundation.

Figure 1

PRISMA flow chart of the record screening process The screening process is shown for the combined dataset of records identified from the 2016 and 2023–24 searches. OCV=oral cholera vaccine.

Figure 2

Stratified and meta-regression estimates of the efficacy and effectiveness of two doses of kOCV as a function of time since vaccination The upper panels show stratified estimates of efficacy (A) and effectiveness (B) by time since vaccination. The trial done in Matlab, Bangladesh used three doses of kOCV but was also included. Estimates are grouped into the five follow-up duration categories by the midpoint of the time window during which the estimate was measured. Bars and squares show 95% CIs and point estimates, respectively, of efficacy or effectiveness for each study, coloured by vaccine type. Black diamonds are the estimated pooled efficacy or effectiveness and 95% CI by follow-up period, with numerical values shown beneath the x-axis. If there is only one estimate in the follow-up period, the estimate from the study is presented on the x-axis. Estimates of I² for the pooled estimates in panel A are 0% (0–12 months), 0% (12–24 months), 25% (24–36 months), and 81% (36–48 months). Estimates of I² for the pooled estimates in panel B are 36% (0–12 months), 0% (12–24 months), and 52% (24–36 months). The bottom panels illustrate meta-regression results for mean two-dose efficacy (C) and effectiveness (D) as a function of time since vaccination, with the shaded envelope representing the 95% CIs (darker region) and 95% prediction intervals (lighter region). The horizontal grey lines represent the data from the literature that were used to fit the meta-regression models; the length of the line indicates the duration of follow-up (months since vaccination) and the line's position on the y-axis marks the magnitude of the point estimate (%). The dashed horizontal line at y=0 denotes no protective effect (0%) of kOCV. kOCV=killed whole-cell oral cholera vaccine.

Figure 3

Stratified and meta-regression estimates of the efficacy and effectiveness of one dose of kOCV as a function of time since vaccination (A) Stratified estimates of efficacy or effectiveness by 6-month time periods after vaccination. Bars and squares show 95% CIs and point estimates, respectively, of efficacy (dashed lines) or effectiveness (solid lines), coloured by vaccine type. Black diamonds are the estimated pooled effectiveness and 95% CI by follow-up period, with numerical values beneath the x-axis in black (first row). If there is only one effectiveness estimate in the follow-up period, the estimate from that study is presented. Efficacy estimates for those 5 years and older. Estimates of I² for the pooled estimates in panel A are 36% (0–6 months), 0% (12–18 months), and 0% (24–30 months). (B) Meta-regression results for mean one-dose effectiveness as a function of time since vaccination, with the shaded envelope representing the 95% CIs (darker region) and 95% prediction intervals (lighter region). The horizontal grey lines represent the data from the literature that were used to fit the meta-regression models; the length of the line indicates the duration of follow-up (months since vaccination), and the position of the line on the y-axis marks the magnitude of the point estimate (%). The dashed horizontal line at y=0 denotes no protective effect (0%) of kOCV. kOCV=killed whole-cell oral cholera vaccine.