Effect of BCG Danish and oral polio vaccine on neonatal mortality in newborn babies weighing less than 2000 g in India: multicentre open label randomised controlled trial (BLOW2).
Adhisivam B, Kamalarathnam C, Bhat BV, Jayaraman K, Namachivayam SP, Shann F, McSharry B, David P, Raja, Sundaram M
- DOI
- 10.1136/bmj-2025-084745
- Record issued
- 2026-08-15
- Engine
- 7.39.0
- Exported
- 2026-09-22
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/fa83b576-2a52-41b2-bd68-41058c7910ff is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×3−1.5★
- ClaimsOverstated claim−0.5★
- ReportingData & code availability partially met−0.25★
- The numeric-impossibility checks (GRIM/GRIMMER/DEBIT/SPRITE) did not run: 4 reported means were read, and their group size is not stated where the values are printed (this source has no machine-readable table structure). These checks need the count the mean was averaged over, so none was performed.
- 01Conclusion reaches beyond the evidence
A substantial reduction in neonatal mortality could be achieved if a skilled administrator vaccinated a high proportion of newborns soon after birth.
“A substantial reduction in neonatal mortality could be achieved if a skilled administrator vaccinated a high proportion of newborn babies in high mortality settings on the day of, or soon after, birth.”
ConclusionFind 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-conducted and transparently reported randomised controlled trial with a strong scientific premise, rigorous design, and appropriate statistical analysis. The main weakness is the vague data availability statement, which lacks a concrete access mechanism or repository deposit, and the absence of an explicit reporting guideline checklist.
Both reviewers independently scored all eight dimensions and agreed on every status; no divergence required reconciliation. The study is an interventional RCT; non-applicable sub-criteria (e.g., animal housing, cell line authentication) were excluded. The statistics verification recomputed only 2 reported tests (both consistent); other statistics were not machine-verified and should not be assumed correct beyond that.
Numerical inconsistencies
1 finding · worst lowValues 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 2 tests: 2 consistent, 0 inconsistent; 1 recomputed directly from the reported test statistics, 1 via agent-written checks.
- CONSISTENTreported p = .030 · recomputed p = .037Recomputed adjusted hazard ratio 0.83 (95% CI 0.69–0.98), reported p=0.03
“adjusted hazard ratio 0.83, 95% confidence interval (CI) 0.69 to 0.98; P=0.03”
Taken as given: 0.69–0.98 is a two-sided 95% confidence interval for the adjusted hazard ratio of 0.83, not a range, an IQR, or a different interval level; the adjusted hazard ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.03 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.83, 0.69, 0.98, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Secondary outcome infection-related mortality adjusted hazard ratio 0.53 with 95% CI 0.40 to 0.70, P<0.001
“adjusted hazard ratio 0.53, 95% CI 0.40 to 0.70”
Taken as given: The hazard ratio is 0.53.; The 95% CI is 0.40 to 0.70.; The CI is two-sided at 95%.Method: Recomputed p-value from the hazard ratio and its 95% CI using the pCI function for a log-scale ratio.How we recomputed it: pCI(0.53, 0.40, 0.70, 1)
- lowinternal contradictionThe number of deaths in the early vaccination group is reported as 238 in the abstract and results, but Table 4 lists 238 total deaths. However, the per-protocol analysis in Table 2 shows 236 deaths in the early group, which is explained by two babies who were not vaccinated and moved to the control group.
“§ Biased estimate because low risk time from discharge to 28 days was removed from vaccinated controls, and two babies who died were randomised to BCG-OPV but were not vaccinated because they were considered too ill were moved from the BCG-OPV group to the control group in the per protocol analysis.”
Table 2Find in source - lowinternal contradictionThe abstract states 'nine were lost to follow-up' but Table 2 footnote says 'Six babies in the early BCG-OPV group and three in the control group were lost to follow-up before 28 days of age' which sums to nine, consistent. No contradiction.
nine were lost to follow-up; Six babies in the early BCG-OPV group and three in the control group were lost to follow-up before 28 days of age
Table 2reviewer’s wording - lowinternal contradictionTable 4 shows 'Total deaths 238 (100) | 273 (100) | 511 (100)'. The sum of individual causes in the early group: 73+82+22+6+3+10+41+1 = 238, and control: 133+64+18+4+0+1+52+1 = 273, consistent.
Total deaths 238 (100) | 273 (100) | 511 (100)
Table 4reviewer’s wording
Overstated conclusions
1 finding · worst mediumConclusions 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.
- Conclusions overstated beyond the evidenceAssessed
4 major claims checked against the paper's own evidence: 1 not fully backed by the presented evidence (unsupported or overstated).
- overstatedReviewers 1, 2A substantial reduction in neonatal mortality could be achieved if a skilled administrator vaccinated a high proportion of newborns soon after birth.This is a policy implication extrapolated from the trial results; while plausible, it is not directly tested in this study.Evidence: The trial shows a 17% reduction in mortality in a high-risk NICU population; the claim generalizes to broader settings.
“A substantial reduction in neonatal mortality could be achieved if a skilled administrator vaccinated a high proportion of newborn babies in high mortality settings on the day of, or soon after, birth.”
ConclusionFind in source - supportedReviewers 1, 2BCG Danish and OPV administered early reduced all-cause neonatal mortality in newborns weighing <2000 g.The primary outcome analysis shows a statistically significant reduction with adjusted HR 0.83 (95% CI 0.69-0.98, P=0.03).Evidence: Primary outcome: adjusted hazard ratio 0.83, 95% CI 0.69 to 0.98; P=0.03.
“In intention-to-treat Cox survival analysis stratified by NICU and adjusted for postnatal age, birth weight, sex, and gestational age, neonatal mortality per person year was 1.29 in the early vaccination group and 1.50 in the control group (adjusted hazard ratio 0.83, 95% confidence interval (CI) 0.69 to 0.98; P=0.03).”
AbstractFind in source - supportedReviewers 1, 2The reduction in all-cause mortality was due to a decrease in infection-related deaths.The secondary outcome of infection-related mortality showed a significant reduction (HR 0.53, 95% CI 0.40-0.70), and the cause-of-death table shows a lower proportion of infection deaths in the early group.Evidence: Secondary outcome: adjusted hazard ratio 0.53, 95% CI 0.40 to 0.70; Table 4 shows infection deaths 73 (31%) vs 133 (49%).
“Infection related neonatal mortality per person year was 0.40 in the early vaccination group and 0.73 in the control group (adjusted hazard ratio 0.53, 95% CI 0.40 to 0.70).”
Table 4Find in source - supportedReviewers 1, 2The effect is a non-specific (off-target) effect, not due to tuberculosis prevention.No deaths from tuberculosis occurred, and the reduction was in infection-related deaths, supporting a non-specific effect.Evidence: No deaths from tuberculosis occurred; infection-related mortality reduced.
“No deaths from tuberculosis occurred, and no serious adverse effects were associated with vaccination.”
AbstractFind in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary outcome is all-cause neonatal mortality, a hard clinical outcome. The secondary outcome is infection-related mortality, also a clinical outcome. No surrogate biomarker is used as the basis for the efficacy claim.
“The primary outcome was all cause neonatal mortality.”
- ADEQUATEEffect sizeThe primary effect is a 17% reduction in all-cause neonatal mortality (adjusted HR 0.83, 95% CI 0.69-0.98), with a number needed to treat of 21. This is a clinically meaningful reduction in mortality, a hard outcome, and is statistically significant.
“adjusted hazard ratio 0.83, 95% confidence interval (CI) 0.69 to 0.98; P=0.03”
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 mediumRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Data/code availability incompleteAssessed
The introduction cites prior randomised trials in Guinea-Bissau and the authors' own BLOW1 trials, acknowledging conflicting results and the need to test a different BCG strain. The rationale for using BCG Danish and OPV is clearly linked to the hypothesis that non-specific effects may reduce neonatal mortality. The paper explicitly addresses limitations of prior research, such as the different BCG strains and the need for a larger trial.
“In three randomised trials in newborns weighing <2500 g in Guinea-Bissau, none had a statistically significant effect on the primary outcome from the administration of BCG Danish vaccine (Statens Serum Institut, Copenhagen) at discharge from hospital, but meta-analysis showed a reduction in neonatal mortality of 38% (95% CI 17% to 54%)”
“The four strains of BCG that Unicef distributes to low and middle income countries differ noticeably in bacterial viability, RNA content, and immunological effects, and the vaccines might therefore have different specific and non-specific effects.”
“Our trial differs in important ways from the four other smaller randomised trials with an unvaccinated control group that tested the effect of BCG Danish on neonatal mortality in Guinea-Bissau.”
“In three randomised trials in newborns weighing <2500 g in Guinea-Bissau, none had a statistically significant effect on the primary outcome from the administration of BCG Danish vaccine (Statens Serum Institut, Copenhagen) at discharge from hospital, but meta-analysis showed a reduction in neonatal mortality of 38% (95% CI 17% to 54%)”
“Soon after completion of our trials of BCG Russian, and because this vaccine may be less effective than BCG Danish, we performed a trial in newborn babies weighing <2000 g (BLOW2) to test the effect of BCG Danish and OPV in the same three NICUs that had tested BCG Russian using the same trial design.”
“In three randomised trials in newborns weighing <2500 g in Guinea-Bissau, none had a statistically significant effect on the primary outcome”
Randomisation was computer-generated, 1:1, stratified by NICU, sex, and birth weight, with random block sizes. A priori sample size calculation was based on expected mortality and a 30% reduction. Inclusion/exclusion criteria were pre-specified. Blinding was not possible due to the visible BCG scar, and this was explicitly justified. The analysis population (ITT) and handling of missing data (censoring) were defined.
“The newborn babies were randomised using a specially written computer program that checked eligibility, allocated treatment, and recorded baseline and outcome data. Randomisation was 1:1 and stratified by NICU, sex, and birth weight (<1000 g, 1000-1499 g, 1500-1999 g), and with a block size that varied randomly from 2-8 newborn babies.”
“using the Stata version 14.1 sampsi command, we estimated that a total of 510 deaths would be required: with a mortality rate of 15% in the control group, 1998 infants would be required in each group to have 95% power to detect a 30% reduction in mortality using BCG and OPV with an α <0.01”
“The BCG vaccine causes an obvious skin lesion, so it was not possible to mask staff to group assignment because it would have been unethical to cover the upper arm of these high risk babies for their entire stay in intensive care”
“The newborn babies were randomised using a specially written computer program that checked eligibility, allocated treatment, and recorded baseline and outcome data. Randomisation was 1:1 and stratified by NICU, sex, and birth weight (<1000 g, 1000-1499 g, 1500-1999 g), and with a block size that varied randomly from 2-8 newborn babies.”
“we estimated that a total of 510 deaths would be required: with a mortality rate of 15% in the control group, 1998 infants would be required in each group to have 95% power to detect a 30% reduction in mortality using BCG and OPV with an α <0.01”
“The BCG vaccine causes an obvious skin lesion, so it was not possible to mask staff to group assignment because it would have been unethical to cover the upper arm of these high risk babies for their entire stay in intensive care”
Sex, birth weight, gestational age, and other baseline characteristics are reported in Table 1. The study includes both sexes, so sex justification is not applicable. Age (at enrolment) and health status (prematurity, Apgar scores) are reported. Demographics such as maternal BCG scar and delivery mode are also reported.
“Male | 1389 (51) | 1382 (51) | | Female | 1325 (49) | 1324 (49)”
“Median (IQR) age at enrolment (days) | 0.86 (0.54-1.46) | 0.86 (0.56-1.38) | | Median (IQR) birth weight (g) | 1560 (1320-1745) | 1560 (1320-1745)”
The paper states approval by the Jawaharlal Institute Ethics Committee and the Institutional Ethics Committee of Madras Medical College, with protocol numbers. Written informed consent was obtained from parents. The trial was registered prospectively.
“the trial was approved by the Jawaharlal Institute of Postgraduate Medical Education and Research Ethics and Scientific Advisory Committees (JIP/IEC/2016/25/831), and for the Institute of Child Health and Institute of Obstetrics and Gynaecology by the Institutional Ethics Committee of the Madras Medical College (No 02052018).”
“Written informed consent was obtained from a parent before enrolment.”
“Trial registration Clinical Trials Registry India CTRI/2017/01/007676.”
“the trial was approved by the Jawaharlal Institute of Postgraduate Medical Education and Research Ethics and Scientific Advisory Committees (JIP/IEC/2016/25/831), and for the Institute of Child Health and Institute of Obstetrics and Gynaecology by the Institutional Ethics Committee of the Madras Medical College (No 02052018).”
“Written informed consent was obtained from a parent before enrolment.”
“The protocol was registered prospectively with the Clinical Trials Registry India (CTRI/2017/01/007676).”
The BCG Danish vaccine (GreenSignal Bio Pharma, Chennai) and OPV (Bharat Biotech International, Hyderabad; or Bio-Med, Ghaziabad) are named with manufacturers. The dose and route are specified. Stata version 18 is identified for analysis. No other key biological resources (antibodies, cell lines) are used.
“Babies in the early vaccination group received 0.1 mL of BCG Danish (GreenSignal Bio Pharma, Chennai) intradermally within 48 hours of admission to the NICU, along with 0.1 mL (two drops) of bivalent 1&3 OPV (Bharat Biotech International, Hyderabad; or Bio-Med, Ghaziabad).”
“Intention-to-treat analysis was performed using Stata version 18”
“Babies in the early vaccination group received 0.1 mL of BCG Danish (GreenSignal Bio Pharma, Chennai) intradermally within 48 hours of admission to the NICU, along with 0.1 mL (two drops) of bivalent 1&3 OPV (Bharat Biotech International, Hyderabad; or Bio-Med, Ghaziabad).”
“Intention-to-treat analysis was performed using Stata version 18”
Cox proportional hazards models are used, with tests for proportional hazards and fractional polynomials for continuous variables. Exact p-values are reported (e.g., P=0.03). Effect sizes are reported as hazard ratios with 95% CIs. Statistical software (Stata 18) is identified. Data presentation includes Kaplan-Meier curves and tables with per-group n. The mathematical plausibility checks are not applicable due to large N and continuous outcomes.
“The proportional hazards assumption of the Cox model was tested using Stata’s time-dependent covariable command with log(time), and by using Schoenfeld residuals, log-log plots, and separately estimated Kaplan-Meier curves.”
“adjusted hazard ratio 0.83, 95% confidence interval (CI) 0.69 to 0.98; P=0.03”
“0.83 (0.69 to 0.98) | 0.03”
“Intention-to-treat analysis was performed using Stata version 18, with Cox proportional hazard models stratified by NICU with age as the underlying time”
“The proportional hazards assumption of the Cox model was tested using Stata’s time-dependent covariable command with log(time), and by using Schoenfeld residuals, log-log plots, and separately estimated Kaplan-Meier curves.”
“adjusted hazard ratio 0.83, 95% confidence interval (CI) 0.69 to 0.98; P=0.03”
The data availability statement says 'Deidentified individual patient data are available on reasonable request' but does not specify a platform or conditions. The Stata analysis code is included in the supplementary file, which is a form of code sharing. No repository deposit or accession numbers are provided.
“The Stata statistical analysis code is included in the supplementary file.”
“The Stata statistical analysis code is included in the supplementary file.”
The trial is registered with CTRI. Methods are detailed enough for replication. Limitations, including lack of blinding and the secondary outcome added mid-trial, are discussed. Conclusions are appropriately cautious, noting the non-specific effect. Funding and competing interests are declared.
“Trial registration Clinical Trials Registry India CTRI/2017/01/007676.”
“As skin lesions due to BCG are obvious, a double blind trial could not be performed as it was not ethical to keep the upper arm of these high risk babies covered during their entire stay in intensive care.”
“Funding: No external funding; the trial was supported by the participating units and the resources of the investigators. Competing interests: All authors have completed the ICMJE uniform disclosure form”
“Trial registration Clinical Trials Registry India CTRI/2017/01/007676.”
“As skin lesions due to BCG are obvious, a double blind trial could not be performed as it was not ethical to keep the upper arm of these high risk babies covered during their entire stay in intensive care.”
“Funding: No external funding; the trial was supported by the participating units and the resources of the investigators. Competing interests: All authors have completed the ICMJE uniform disclosure form”
Registered (1 ID: Clinical Trials Registry – India). 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 37 references by DOI: 29 verified — 8 no DOI (shown, not verified).
- NO DOIThe State of the World’s Children 2021: On My Mind Promoting, Protecting and Caring for Children’s Mental HealthNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIBCG Market Update October 2024No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIBCG vaccine. WHO position paperNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIImmune status and BCG vaccination in newborns with intra-uterine growth retardationNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOITiming and dose of BCG vaccination in infants as assessed by postvaccination tuberculin sensitivityNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIBCG vaccination reaction in low birth weight infantsNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIBCG vaccines: WHO position paper – February 2018No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIBacillus Calmette-Guérin (BCG) Vaccine Supply & Demand Update – December 2014No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
1 data/code link checked; 1 live.
- dataOSFLIVEHTTP 200https://osf.io/3492q/Resolves to OSF (data repository).
Copyediting
5 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 5 minor suggestions below.
5 copyedit issues flagged: mostly consistency, clarity, other.
- MINORconsistencyAbstract“The number needed to treat to prevent one death was 21 (95% CI 10 to 245).”→ Ensure the NNT is consistently reported in the results section.The NNT is reported in the abstract and results, but the confidence interval is wide; consider discussing this.
- MINORclarityMethods, 'Statistical analysis'“We did not adjust the 95% confidence intervals (CIs) and P values of the secondary outcomes for multiplicity.”→ Consider clarifying that this is a limitation.This is a standard practice but should be highlighted in limitations.
- MINORotherTable 4“BCG to control ratio”→ Clarify that this is a crude ratio, not an adjusted hazard ratio.The table footnote explains it, but the column header could be more explicit.
- MINORconsistencyTable 1“JIMPER”→ Use consistent abbreviation for Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER) throughout.The abbreviation JIMPER is used in Table 1 but JIPMER is used elsewhere.
- MINORclarityData availability statement“Deidentified individual patient data”→ Complete the sentence to specify where and how the data can be accessed.The statement is incomplete in the provided text.
The published work is robust and generally trustworthy, with minor reporting gaps. An informed reader should weigh the vague data availability statement and the lack of an explicit CONSORT checklist; these do not undermine the core findings but warrant attention if the authors issue a correction or update.
- 1.HIGHdata codeIn the Data availability statement, specify a concrete access mechanism (e.g., a data access committee, a repository like OSF/Dryad/Zenodo with a DOI, or a clear process for requesting data from the corresponding author) and complete the sentence that currently reads 'Deidentified individual patient data'.The current statement is vague and incomplete, which undermines reproducibility and transparency.
- 2.HIGHreportingAdd an explicit statement in the Methods or a separate section that the trial adheres to the CONSORT reporting guideline, and consider uploading the CONSORT checklist as a supplementary file.Reporting guidelines improve completeness and are expected for RCTs; the paper currently does not mention one.
- 3.MEDIUMreportingIn the Discussion, explicitly note that the 95% CIs and P values for secondary outcomes were not adjusted for multiplicity, and discuss this as a limitation.The copyedit flagged this as a clarity issue; readers should be aware of the potential for inflated type I error.
- 4.MEDIUMcopyeditIn Table 1, replace the abbreviation 'JIMPER' with 'JIPMER' for consistency with the rest of the manuscript.Inconsistent abbreviation for the same institution could confuse readers.
- 5.MEDIUMcopyeditIn Table 4, clarify the column header 'BCG to control ratio' to indicate it is a crude ratio, not an adjusted hazard ratio, or add a footnote to that effect.The copyedit noted the header could be misinterpreted; clarity improves accuracy.
- 6.MEDIUMreportingIn the Discussion, temper the policy implication that 'a substantial reduction in neonatal mortality could be achieved if a skilled administrator vaccinated a high proportion of newborns soon after birth' to acknowledge it is an extrapolation not directly tested in this trial.The claim audit rated this as overstated; the evidence supports a reduction in this trial but not a broad policy claim.
- 7.LOWcopyeditIn the Abstract and Results, ensure the number needed to treat (NNT) of 21 (95% CI 10 to 245) is consistently reported and consider discussing the wide confidence interval.The copyedit flagged a consistency issue; the wide CI is important for interpretation.
- 8.LOWdata codeConsider depositing deidentified individual patient data in a public repository (e.g., Dryad, Zenodo) with a DOI, if ethically permissible, to enhance transparency.A repository deposit would strengthen the data availability statement and support reproducibility.
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.