Efficacy of typhoid conjugate vaccine: final analysis of a 4-year, phase 3, randomised controlled trial in Malawian children.
Patel PD, Liang Y, Meiring JE, Chasweka N, Patel P, Misiri T, Mwakiseghile F, Wachepa R, Banda HC, Shumba F, Kawalazira G, Dube Q, Nampota-Nkomba N, Nyirenda OM, Girmay T, Datta S, Jamka LP, Tracy JK, Laurens MB, Heyderman RS, Neuzil KM, Gordon MA, TyVAC team
- DOI
- 10.1016/S0140-6736(23)02031-7
- Record issued
- 2026-08-10
- Engine
- 7.29.0
- Exported
- 2026-09-20
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/4889539c-19f9-4195-b617-5f2c13ab0f5f is authoritative.
How this rating was calculated
- IntegrityIntegrity concern−0.5★
- ReportingStatistical analysis not met−0.5★
- 01Statistical reporting inadequate
Statistical methods and reporting are generally strong (tests named, exact p-values, CIs, software), but Table 3 contains a demonstrable arithmetic inconsistency in the third-year Vi-TT incidence rate.
“Third year | | Vi-TT | 14 043 | 14 031 | 3 | 93·1 (54·1–160·3) | 77·0% (16·4–95·8)”
Table 3Find in source - 02Internal contradictions in the reported numbers
Table 3's third-year Vi-TT incidence rate (93·1 per 100 000 person-years) is inconsistent with its own numerator and denominator (3 cases / 14 031 person-years = 21·4 per 100 000) and duplicates the MenA row's rate and CI.
“Third year | | Vi-TT | 14 043 | 14 031 | 3 | 93·1 (54·1–160·3) | 77·0% (16·4–95·8) | | MenA | 13 976 | 13 963 | 13 | 93·1 (54·1–160·3) | Ref |”
Table 3Find in source
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This is a well-designed and generally well-reported Phase 3 clinical trial with a strong scientific premise and robust ethical approvals. However, a critical arithmetic inconsistency in a key results table and the absence of reported baseline weight for a pediatric cohort slightly diminish its rigor. The lack of explicit reference to a reporting guideline also represents a minor reporting gap.
This review synthesized two independent rigor evaluations and a copyedit report. The reviewers largely agreed, with the main divergences concerning the completeness of biological variable reporting and the severity of the reporting guideline omission. The statistical analysis dimension's 'fail' status is driven by a clear arithmetic error identified by Reviewer 2, the integrity component, and the copyedit pass.
Numerical inconsistencies
1 finding · worst mediumValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Internal contradictions in the reported numbersAssessed
Recomputed 1 test: 1 consistent, 0 inconsistent; 1 via agent-written checks.
- CONSISTENTreported p = .770 · recomputed p = .764Reviewer 2Two-tailed p for the annual decline slope from its 95% CI
“The trend in efficacy by yearly increments, analysed using a random effects meta-regression taking into account the precision for each annual estimate, suggests there was a 1·3% decline in vaccine efficacy each year (slope –0·013, 95% CI –0·098 to 0·072, p=0·77, ).”
Taken as given: the 95% CI is a symmetric normal-based confidence interval for the slope; the reported p is two-tailedMethod: Back-computed a two-tailed p from estimate and 95% CI under a normal approximation using pCI.How we recomputed it: pCI(-0.013, -0.098, 0.072, 0)
- mediuminternal contradictionTable 3's third-year Vi-TT incidence rate (93·1 per 100 000 person-years) is inconsistent with its own numerator and denominator (3 cases / 14 031 person-years = 21·4 per 100 000) and duplicates the MenA row's rate and CI.
“Third year | | Vi-TT | 14 043 | 14 031 | 3 | 93·1 (54·1–160·3) | 77·0% (16·4–95·8) | | MenA | 13 976 | 13 963 | 13 | 93·1 (54·1–160·3) | Ref |”
Table 3Find in source
Overstated conclusions
None foundConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
Checked — nothing surfaced.
6 major claims checked against the paper's own evidence: all adequately supported.
- supportedReviewers 1, 2A single dose of Vi-TT is durably efficacious for at least 4 years among children aged 9 months to 12 years.The primary result of 78.3% efficacy (95% CI 66.3-86.1) over a median follow-up of 4.3 years directly supports this claim.Evidence: Table 2 and text: vaccine efficacy 78.3% (95% CI 66.3-86.1) in ITT analysis.
“A single dose of Vi-TT is durably efficacious for at least 4 years among children aged 9 months to 12 years”
AbstractFind in source - supportedReviewers 1, 2The vaccine shows efficacy in all age groups, including children younger than 2 years.Subgroup analysis shows efficacy of 70.6% in children <2 years, 79.6% in 2-4 years, and 79.3% in 5-12 years, with no significant heterogeneity (p=0.85).Evidence: Table 2 and text: age-stratified efficacy estimates with CIs.
“Efficacies by age group were 70·6% (6·4–93·0) for children aged 9 months to 2 years; 79·6% (45·8–93·9) for children aged 2–4 years; and 79·3% (63·5–89·0) for children aged 5–12 years.”
Results ¶7Find in source - supportedReviewers 1, 2These results support current WHO recommendations for typhoid conjugate vaccine use in endemic areas.The durable efficacy data provide evidence for the WHO-recommended strategy of mass campaigns and routine introduction.Evidence: Discussion states that results support WHO recommendations, and the paper provides relevant efficacy data.
“These results support current WHO recommendations in typhoid-endemic areas for mass campaigns among children aged 9 months to 15 years, followed by routine introduction in the first 2 years of life.”
Discussion ¶1Find in source - supportedReviewer 1The loss of vaccine efficacy over time was estimated to be only 1.3% per annum.The trend analysis reports a slope of -0.013 (95% CI -0.098 to 0.072, p=0.77), which is consistent with minimal waning, though the CI includes the possibility of no decline.Evidence: Results, text and appendix: 'a 1.3% decline in vaccine efficacy each year (slope –0.013, 95% CI –0.098 to 0.072, p=0.77)'.
suggests there was a 1.3% decline in vaccine efficacy each year (slope –0.013, 95% CI –0.098 to 0.072, p=0.77)
Results ¶6reviewer’s wording - supportedReviewer 1The estimated number needed to vaccinate is now 163 children.The number needed to vaccinate is calculated from the absolute risk reduction and reported in Table 2.Evidence: Table 2: 'Number needed to vaccinate: 163 (129–222)'.
Number needed to vaccinate 163 (95% CI 129–222)
Table 2reviewer’s wording - supportedReviewer 2Vaccine efficacy declined by only 1·3% per year over 4 years.The reported slope from the random-effects meta-regression is –0·013, with a wide CI including zero; the claim is presented as an estimate and is consistent with the point estimate.Evidence: Slope –0·013, 95% CI –0·098 to 0·072, p=0·77.
“We estimated that vaccine efficacy reduced over time by only 1·3% per year over 4 years.”
Research in context, Added valueFind in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointPrimary endpoint is blood culture-confirmed typhoid fever, a hard clinical outcome.
“The primary outcome was first occurrence of blood culture-confirmed typhoid fever.”
- ADEQUATEEffect sizeVaccine efficacy of 78.3% (95% CI 66.3–86.1) against blood culture-confirmed typhoid fever, with number needed to vaccinate of 163. This is a clinically meaningful reduction in disease incidence.
“Overall vaccine efficacy in preventing a first episode of blood culture-confirmed typhoid fever for the entire study period was 78·3% (95% CI 66·3–86·1) in the intention-to-treat analysis.”
Data authenticity concerns
None foundAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
Checked — nothing surfaced.
Reporting gaps
1 finding · worst highRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Statistical reporting inadequateAssessed
The introduction and 'Research in context' sections cite extensive prior work, including Cochrane reviews and earlier efficacy trials, and explicitly note the gap in long-term durability data. The rationale for the study is logically derived from the cited evidence, and the paper discusses how the limitations of prior studies (e.g., short follow-up, lack of data in children under 2 years) are addressed in this trial.
“A Cochrane systematic review estimated that the efficacy of Vi-PS fell from 69% (95% CI 63–74) after 1 year to 55% (95% CI 30–70) after 3 years.”
“We present the final efficacy results of single-dose Vi-TT over an extended median follow-up time of 4·3 years, and by age group, including children younger than 2 years of age.”
“Data on the durability of protection beyond 2 years are sparse.”
“Data on the durability of protection beyond 2 years are sparse.”
“This is the first randomised, controlled, double-blind trial to evaluate the longer-term efficacy of single-dose TCV in a typhoid fever-endemic setting from 9 months of age.”
“A Cochrane systematic review estimated that the efficacy of Vi-PS fell from 69% (95% CI 63–74) after 1 year to 55% (95% CI 30–70) after 3 years.”
Randomization uses block randomization with block sizes 6-12, generated by an unmasked statistician using the blockrand package in R. The unit of randomization is individual children. Blinding is double-blind, with participants, parents, and study team masked; only nurses administering vaccines were unmasked, which is acceptable. A power analysis was performed assuming 75% efficacy, 90% power, and a minimum of 30 cases. Inclusion/exclusion criteria are specified. The analysis populations (ITT and per-protocol) are defined, and handling of missing data is addressed through censoring in survival analysis.
“Block randomisation was used with block sizes of 6–12. The randomisation allocation sequence was generated by an unmasked statistician using the blockrand package (version 1.3) in R (version 3.4.1).”
“Participants, their parents or guardians, and the study team involved in subsequent disease surveillance screening, eligibility assessment, and follow-up were masked to vaccine allocation.”
“Details of sample size and power calculations have been reported. Briefly, assuming 75% vaccine efficacy, the minimum number of total cases needed to test the null hypothesis that the vaccine has no protective efficacy (ie, vaccine efficacy ≤0), with 90% power, was 30.”
“Block randomisation was used with block sizes of 6–12. The randomisation allocation sequence was generated by an unmasked statistician using the blockrand package (version 1.3) in R (version 3.4.1).”
“Participants, their parents or guardians, and the study team were masked to vaccine allocation. Nurses administering vaccines were unmasked.”
“assuming 75% vaccine efficacy, the minimum number of total cases needed to test the null hypothesis that the vaccine has no protective efficacy (ie, vaccine efficacy ≤0), with 90% power, was 30.”
Sex is reported in the baseline characteristics table (Table 1), with both sexes enrolled. Age is reported as mean, median, and range. Health status is described as 'free of acute illness' at enrolment. Demographics include ethnicity (Black African) and study site. Since both sexes are enrolled, a justification for single-sex is not applicable.
“Age at enrolment (years) | | Mean (SD) | 6·1 (3·3) | 6·2 (3·3)”
“Inclusion criteria required that children were aged 9 months to 12 years, resided in the study areas, had no previous history of typhoid vaccination, and were free of acute illness.”
Local ethics approval from the Malawi National Health and Sciences Research Committee and international approvals from the University of Liverpool Ethics Review Board and the University of Maryland Baltimore Institutional Review Board are stated. Written informed consent from parents/guardians and assent from children aged 8+ are described. Compliance with Good Clinical Practice guidelines is mentioned.
“Local ethics approval was obtained from the Malawi National Health and Sciences Research Committee, and international ethics approval from the University of Liverpool Ethics Review Board and the University of Maryland Baltimore Institutional Review Board.”
“Written informed consent was obtained from the parent or guardian of all children before any study procedures, and written assent was also required for all children aged 8 years and older.”
“This study was conducted under Good Clinical Practice guidelines, with internal and external monitoring for quality.”
“Local ethics approval was obtained from the Malawi National Health and Sciences Research Committee, and international ethics approval from the University of Liverpool Ethics Review Board and the University of Maryland Baltimore Institutional Review Board.”
“Written informed consent was obtained from the parent or guardian of all children before any study procedures, and written assent was also required for all children aged 8 years and older.”
“This study was conducted under Good Clinical Practice guidelines, with internal and external monitoring for quality.”
The investigational products (Vi-TT and MenA) are identified with manufacturer and location. Statistical software (Stata/SE version 17, R, blockrand package) and the REDCap database are named. No other biological resources (antibodies, cell lines, etc.) are used, so those criteria are not applicable.
“Vi-TT (Bharat Biotech International, Hyderabad, India) or a meningococcal capsular group A conjugate control vaccine (MenA, Serum Institute of India, Pune, India).”
“All analyses were performed using Stata/SE version 17.”
“the blockrand package (version 1.3) in R (version 3.4.1)”
“Vi-TT (Bharat Biotech International, Hyderabad, India) or a meningococcal capsular group A conjugate control vaccine (MenA, Serum Institute of India, Pune, India).”
“All analyses were performed using Stata/SE version 17.”
“using the blockrand package (version 1.3) in R (version 3.4.1)”
We verified that the reported third-year Vi-TT incidence rate of 93·1 per 100 000 person-years cannot co-occur with 3 cases and 14 031 person-years (3/14 031 = 21·4 per 100 000); the printed value and CI duplicate the MenA row. All other checked percentages, incidence rates, VE, and NNV values are arithmetically consistent. Tests are named (log-rank, Wilcoxon–Breslow–Gehan, Mantel-Haenszel, random-effects meta-regression), p-values are reported (e.g., p=0·77, p=0·85), effect estimates carry 95% CIs, and Stata/SE 17 is identified.
“The test of homogeneity in incidence rate ratio and vaccine efficacy across subgroups was conducted using the Mantel-Haenszel method.”
“Vaccine efficacy was not significantly different by sex (82%, 95% CI 64·3–91·8 among males and 74·5%, 53·6–86·9 among females; p=0·44)”
“Third year | | Vi-TT | 14 043 | 14 031 | 3 | 93·1 (54·1–160·3) | 77·0% (16·4–95·8)”
“slope –0·013, 95% CI –0·098 to 0·072, p=0·77”
“All analyses were performed using Stata/SE version 17.”
The paper states that the full data set will be available on the Vivli platform, which is a managed-access repository for clinical trial data. The ClinicalTrials.gov identifier is provided. No custom code was developed, so code sharing is not applicable.
“The full data set will be available on the Vivli platform ( https://search.vivli.org/studyDetails/fromSearch/0bb6a051-8149-4161-bb99-d95cdc700714 ).”
“This trial is registered at ClinicalTrials.gov (http://ClinicalTrials.gov) , NCT03299426 .”
“The full data set will be available on the Vivli platform ( https://search.vivli.org/studyDetails/fromSearch/0bb6a051-8149-4161-bb99-d95cdc700714 ).”
Methods are detailed enough for replication. The trial is registered. All outcomes are reported. Limitations are discussed. Conclusions are proportional. Funding and conflicts of interest are disclosed. However, the paper does not explicitly mention adherence to a specific reporting guideline like CONSORT, though the trial profile and reporting are consistent with it.
“This trial is registered at ClinicalTrials.gov (http://ClinicalTrials.gov) , NCT03299426 .”
“This trial had some limitations. The number of children presenting for passive surveillance and blood culture collection decreased during the first 2 years.”
“YL, SD, LPJ, MBL, and KMN receive funding from the TyVAC grant (grant number OPP1151153).”
“This trial is registered at ClinicalTrials.gov (http://ClinicalTrials.gov) , NCT03299426 .”
“This trial had some limitations. The number of children presenting for passive surveillance and blood culture collection decreased during the first 2 years.”
Registered (1 ID: ClinicalTrials.gov). No reporting guideline cited.
Broken references and links
None foundReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Checked — nothing surfaced.
Checked 29 references by DOI: 26 verified — 3 no DOI (shown, not verified).
- NO DOITyphoid fever—level 4 causeNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOITyphoid vaccines: WHO position paperNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOITyphoid vaccines: WHO position paper—March 2018No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
2 data/code links checked; 2 live.
- datahttps://search.vivli.org/studyDetails/fromSearch/0bb6a051-8149-4161-bb99-d95cdc700714LIVEHTTP 200Resolves, but the content could not be matched to the paper.
- datahttp://ClinicalTrials.govLIVEHTTP 200Resolves, but the content could not be matched to the paper.
Copyediting
1 finding · worst lowWording, consistency and formatting errors that need correcting before submission.
- Wording or formatting errors that need correctingAssessed
2 copyedit issues flagged (1 major): mostly consistency, other.
- MAJORconsistencyTable 3, 'Third year' row“93·1 (54·1–160·3)”→ 21·4 (95% CI computed from 3 events / 14,031 person-years)Value duplicates MenA row; cannot be derived from the reported 3 cases.
- MINORotherBeginning of the paper text“pmc 1 Efficacy of typhoid conjugate vaccine: ...”→ Remove the 'pmc 1' prefix, which appears to be an artifact from the source.This is likely a formatting remnant and does not affect scientific content.
This published work is robust in its design and execution, but an informed reader should be aware of the arithmetic inconsistency in Table 3, which warrants a correction or erratum. The absence of baseline weight data and the lack of explicit CONSORT guideline adherence are minor reporting gaps that do not undermine the core findings but could be improved in future publications.
- 1.HIGHstatisticsIssue a correction or erratum for Table 3, specifically for the 'Third year' Vi-TT incidence rate and its 95% CI.The reported incidence rate of 93·1 per 100 000 person-years for Vi-TT in the third year is arithmetically inconsistent with the stated 3 cases over 14,031 person-years (which calculates to 21·4 per 100 000) and duplicates the MenA row, representing a demonstrable error in a key results table.
- 2.HIGHreportingAdd an explicit statement in the Methods section referencing adherence to the CONSORT 2010 statement.Explicitly citing the reporting guideline enhances transparency and confirms that the trial's reporting follows established best practices for clinical trials.
- 3.HIGHrigorReport baseline weight (or height/BMI) in Table 1 to fully satisfy the age/weight/health biological-variables criterion for this paediatric cohort.Weight is an important demographic and potential covariate in pediatric studies, and its inclusion would provide a more complete picture of the study population.
- 4.MEDIUMdata codeIn the Data sharing section, specify the access conditions and review timeframe for the Vivli dataset (e.g., proposal review, de-identification status) alongside the URL.Providing details on data access conditions improves clarity and facilitates data reuse for interested researchers.
- 5.MEDIUMrigorState whether the per-protocol exclusions (e.g., duplicate-ID error) were pre-specified in the statistical analysis plan.Clarifying the pre-specification of per-protocol exclusions strengthens the description of analysis populations and outlier handling.
- 6.LOWcopyeditRemove the 'pmc 1' prefix at the beginning of the paper text.This appears to be a formatting artifact and does not belong in the published text.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.