BCG vaccine — Vaccine Data Navigator

← All vaccines

disease risks

post-acute and long-term sequelae

vaccine efficacy and effectiveness

vaccine safety and adverse events

sources

Tuberculosis (all forms) Mycobacterium tuberculosis

Disease risks

Acute risks of the infection itself: complications, hospitalisation, ICU, death.

Outcome (O)	Population (P)	Estimate	Design (S)	Source
Death in children diagnosed with tuberculosis severity: death case definition: Population-representative paediatric (<15 years) TB case series with known outcomes, studies after 1980	children under 15 years vaccination status: not reported country: multiple period: after 1980 (treatment era) Infant <1y1-4y5-11y12-17y	0.9 (0.5–1.6) CFR (%) per diagnosed and mostly treated child TB case	meta-analysis	Jenkins et al. 2017 Lancet Infect Dis quote In studies in the recent era, when most children had tuberculosis treatment, the pooled case fatality ratio was 0·9% (95% CI 0·5-1·6). abstract, FINDINGS Treated-era figure. The review stresses that most child TB deaths occur in children who are never diagnosed or treated, which this CFR does not capture (see bcg-bur-018).
Death from untreated smear-positive pulmonary tuberculosis severity: death case definition: Untreated smear-positive pulmonary TB in HIV-negative people, pre-chemotherapy era studies	all ages in pre-chemotherapy cohorts immune status: HIV-negative vaccination status: unvaccinated country: multiple period: pre-chemotherapy era All ages	70 10-year CFR (%), weighted mean per untreated smear-positive case	systematic review	Tiemersma et al. 2011 PLoS One quote Untreated smear-positive tuberculosis among HIV negative individuals has a 10-year case fatality variously reported between 53% and 86%, with a weighted mean of 70%. abstract, PRINCIPAL FINDINGS The 53-86% in the same sentence is the range across studies, not a confidence interval, so no CI is recorded. Natural-history figure that underlies modelled untreated-TB mortality.
Death from untreated culture-positive smear-negative pulmonary tuberculosis severity: death case definition: Untreated culture-positive smear-negative pulmonary TB, HIV-negative, pre-chemotherapy era	all ages in pre-chemotherapy cohorts immune status: HIV-negative vaccination status: unvaccinated country: multiple period: pre-chemotherapy era All ages	20 10-year CFR (%), indirect estimate per untreated smear-negative case	systematic review	Tiemersma et al. 2011 PLoS One quote Ten-year case fatality of culture-positive smear-negative tuberculosis was nowhere reported directly but can be indirectly estimated to be approximately 20%. abstract, PRINCIPAL FINDINGS An indirect estimate, as the source says; no CI.
Death in children with tuberculosis, pre-treatment era severity: death case definition: Population-representative paediatric TB series published before 1946	children under 15 years vaccination status: not reported country: multiple period: pre-treatment era (before 1946) Infant <1y1-4y5-11y12-17y	21.9 (18.1–26.4) CFR (%) per child TB case	meta-analysis	Jenkins et al. 2017 Lancet Infect Dis quote Among children with tuberculosis included in studies in the pre-treatment era, the pooled case fatality ratio was 21·9% (95% CI 18·1-26·4) overall. abstract, FINDINGS Pre-chemotherapy era, so this approximates the untreated natural history in children. Age-specific figures follow.
Death in children aged 0-4 years with tuberculosis, pre-treatment era severity: death case definition: Population-representative paediatric TB series published before 1946	children aged 0-4 years vaccination status: not reported country: multiple period: pre-treatment era (before 1946) Infant <1y1-4y	43.6 (36.8–50.6) CFR (%) per child TB case	meta-analysis	Jenkins et al. 2017 Lancet Infect Dis quote The pooled case fatality ratio was significantly higher in children aged 0-4 years (43·6%, 95% CI 36·8-50·6) than in those aged 5-14 years (14·9%, 11·5-19·1). abstract, FINDINGS The same sentence gives 14.9% (95% CI 11.5-19.1) for children aged 5-14 years: untreated TB is far more lethal in the under-fives.
Incident tuberculosis disease after close exposure severity: any case definition: 2-year cumulative TB incidence in exposed children with a positive TB infection test who did not receive preventive therapy	children below 5 years of age prior infection: positive test for tuberculosis infection after household or close exposure vaccination status: mixed country: multiple (34 countries) setting: household contact period: cohorts published 1998-2018 n: 130 512 exposed children followed Infant <1y1-4y	19.0 (8.4–37.4) 2-year cumulative incidence of TB (%) per infected, untreated exposed child under 5	meta-analysis	Martinez et al. 2020 Lancet quote Children not receiving preventive therapy with a positive result for tuberculosis infection had significantly higher 2-year cumulative tuberculosis incidence than children with a negative result for tuberculosis infection, and this incidence was greatest among children below 5 years of age (19·0% [95% CI 8·4-37·4]). abstract, FINDINGS Individual-participant meta-analysis of 46 cohorts. BCG status was mixed; the same analysis estimated preventive-therapy effectiveness of 63% overall and 91% in infected children.
Death from tuberculosis, all ages severity: death case definition: WHO-estimated TB deaths including people with HIV	all ages vaccination status: mixed country: global period: 2024 All ages	1.23 TB deaths per year, millions per population per year	expert body	who.int quote Youth movement against TB in Hua Dan primary school, Chang Muong Commune, Muong La, Son La (A remote commune). © Credits Tuberculosis 24 March 2026 Reading time: Key facts A total of 1.23 million people died from tuberculosis (TB) in 2024 (including 150 000 among people with HIV). Key facts WHO fact sheet dated 24 March 2026 reporting the 2024 estimate; the 2024 Global TB Report (previous records) covers 2023. The same fact sheet gives 10.7 million incident cases in 2024 including 1.2 million children.
Progression from TB infection to TB disease severity: any case definition: Lifetime risk of TB disease after infection with M. tuberculosis, as stated by WHO	all ages vaccination status: mixed country: global period: current All ages	5–10 lifetime risk of TB disease among infected people (%) per infected person	expert body	who.int quote About 5–10% of people infected with TB will eventually get symptoms and develop TB disease. Overview A range stated by WHO without a confidence interval. WHO adds in the next sentence that babies and children are at higher risk of developing disease if infected; the paediatric household-contact figure (bcg-bur-016) quantifies that.
Incident tuberculosis, all forms severity: any case definition: TB cases ascertained over 15 years in a controlled trial of tuberculin-negative British adolescents	14-15 years at entry, followed 15 years vaccination status: unvaccinated country: Great Britain setting: community period: 1950-1967 n: 54 239 trial participants 12-17y18-49y	1.28 annual incidence per 1,000 (unvaccinated tuberculin-negative controls) per 1,000 person-years	RCT	Fourth 1972 Bull World Health Organ quote The annual incidence (per 1 000) of TB was 0.28 among the 13 598 BCG-vaccinated participants and 1.28 among the tuberculin-negative unvaccinated participants admitted concurrently (a 78% reduction); for vole bacillus vaccine (5 817 participants) the figures were 0.29 and 1.50 (an 81% reduction). abstract The 1.28 is the unvaccinated-control incidence; the same sentence gives 0.28 per 1,000 in BCG recipients (a 78% reduction, filed as bcg-eff-027). A high-income, pre-decline setting in which TB was still common.
Incident bacillary pulmonary tuberculosis severity: bacillary pulmonary disease case definition: Sputum-positive pulmonary TB found by repeated surveys and passive case finding in Chingleput district	all ages registered in the trial population vaccination status: mixed country: India setting: community, high transmission period: 1968-1983 n: 366,625 registered All ages	2.6 incidence per 1,000 person-years (whole trial population) per 1,000 person-years	RCT	1999 Indian J Med Res quote This represents a small fraction of the total incidence of 2.6 per 1000 person-years most of which came from those who were initially tuberculin positive. abstract Most incident cases arose in people already tuberculin-positive at intake. Very high background infection rate; this is the setting in which BCG showed no protection in adults.
Incident tuberculosis, all forms severity: any case definition: TB events in trial records, registries and death certificates, 1948-1998	participants vaccinated as children/young people in 1935-1938, followed to 1998 vaccination status: unvaccinated country: United States (American Indians and Alaska Natives) setting: community period: 1948-1998 n: 1309 placebo recipients 5-11y12-17y18-49y50-64y65+	138 incidence per 100,000 person-years (placebo group) per 100,000 person-years	RCT	Aronson et al. 2004 JAMA quote RESULTS: The overall incidence of tuberculosis was 66 and 138 cases per 100 000 person-years in the BCG vaccine and placebo groups, respectively, for an estimate of vaccine efficacy of 52% (95% confidence interval, 27%-69%). abstract, RESULTS 138 is the placebo-group incidence; the vaccine group had 66 per 100,000 person-years in the same sentence (bcg-eff-023). A high-risk indigenous population in a high-income country.
Incident tuberculosis disease, all forms severity: any case definition: WHO-estimated incident TB cases (new and relapse), all ages	all ages vaccination status: mixed country: global period: 2023 All ages	10.8 (10.1–11.7) incident TB cases, millions per year per population per year	surveillance	who.int quote An estimated global total of 10.8 million people (95% uncertainty interval [UI]: 10.1–11.7 million) fell ill with TB in 2023, equivalent to 134 incident cases (95% UI: 125–145) per 100 000 population ( Table 1.1.1 ). Section 1.1 TB incidence, para 2 WHO modelled estimate with a 95% uncertainty interval, not a confidence interval. The same sentence gives the rate per 100,000 (next record). WHO notes the 2024 edition burden estimates are superseded by each new annual report.
Incident tuberculosis disease, all forms severity: any case definition: WHO-estimated incident TB cases (new and relapse), all ages	all ages vaccination status: mixed country: global period: 2023 All ages	134 (125–145) incidence per 100,000 population per year per 100,000 population per year	surveillance	who.int quote An estimated global total of 10.8 million people (95% uncertainty interval [UI]: 10.1–11.7 million) fell ill with TB in 2023, equivalent to 134 incident cases (95% UI: 125–145) per 100 000 population ( Table 1.1.1 ). Section 1.1 TB incidence, para 2 95% uncertainty interval from WHO modelling. Most cases were in the WHO South-East Asia, African and Western Pacific regions.
Incident tuberculosis disease in children severity: any case definition: WHO-estimated incident TB cases among children aged 0-14 years	children aged 0-14 years vaccination status: mixed country: global period: 2023 Infant <1y1-4y5-11y12-17y	1.25 (1.16–1.35) incident TB cases in children, millions per year per child population per year	surveillance	who.int quote The highest burden is in adult men (aged ≥15 years), with an estimated 6.0 million cases (95% UI: 5.5–6.4 million) in 2023, equivalent to 55% of the estimated total; by comparison, there were an estimated 3.6 million cases (95% UI: 3.3–3.9 million) among adult women (aged ≥15 years), equivalent to 33% of the estimated total, and 1.25 million (95% UI: 1.16–1.35 million) cases among children (aged 0–14 years), equivalent to 12% of the estimated total. Section 1.1 TB incidence, age and sex para Children are 12% of the global estimated total in the same sentence. The same page reports that notifications covered only 55% of estimated incidence in children, so most paediatric TB is undiagnosed or unreported.
Death from tuberculosis among HIV-negative people severity: death case definition: WHO-estimated TB deaths among HIV-negative people	all ages vaccination status: mixed country: global period: 2023 All ages	1.09 (0.98–1.2) TB deaths among HIV-negative people, millions per year per population per year	surveillance	who.int quote Globally in 2023, there were an estimated 1.09 million deaths among HIV-negative people (95% uncertainty interval [UI]: 0.98–1.2 million) and an estimated 161 000 deaths among people with HIV (95% UI: 132 000–193 000). Section 1.2 TB mortality, para 3 Deaths among people with HIV are counted separately (next record) because ICD classifies them as HIV/AIDS deaths.
Death from tuberculosis among people with HIV severity: death case definition: WHO-estimated TB deaths among people living with HIV	all ages immune status: HIV-positive vaccination status: mixed country: global period: 2023 All ages	161 000 (132 000–193 000) TB deaths among people with HIV per year per population per year	surveillance	who.int quote Globally in 2023, there were an estimated 1.09 million deaths among HIV-negative people (95% uncertainty interval [UI]: 0.98–1.2 million) and an estimated 161 000 deaths among people with HIV (95% UI: 132 000–193 000). Section 1.2 TB mortality, para 3 The combined HIV-negative plus HIV-positive total in the same sentence is 1.25 million TB deaths in 2023.
Death from tuberculosis (case fatality ratio, all incident cases) severity: death case definition: WHO-estimated proportion of people who develop TB who die from it, HIV-negative and HIV-positive combined	all ages vaccination status: mixed country: global period: 2023 All ages	11.5 CFR (%) per incident TB case	surveillance	who.int quote Globally, the CFR was 11.5% in 2023, down from 12.4% in 2022 and 13.6% in 2021 (and below the pre-pandemic level of 13% in 2019). Section 1.2 TB mortality, CFR para A modelled population-level CFR that mixes treated and untreated disease; compare with the untreated pre-chemotherapy CFR of about 70% (bcg-bur-012) and the treated paediatric CFR of 0.9% (bcg-bur-011). The same sentence gives 13% for 2019 and higher values in the pandemic years.
Death from tuberculosis in children severity: death case definition: WHO-estimated TB deaths in children and young adolescents under 15 years, HIV-negative and HIV-positive combined	children and young adolescents under 15 years vaccination status: mixed country: global period: 2023 Infant <1y1-4y5-11y12-17y	191 000 (170 000–214 000) TB deaths in children under 15 per year per child population per year	surveillance	who.int quote In 2023, an estimated total of 191 000 children and young adolescents (<15 years) died from TB (95% UI: 170 000–214 000): 166 000 among those HIV-negative (95% UI: 149 000–184 000) and 25 000 among those with HIV (95% UI: 21 000–30 000). Section 1.2 TB mortality, age/sex para Children under 15 were 15% of HIV-negative TB deaths in 2023 (same page). Age split below.
Death from tuberculosis in children under 5 severity: death case definition: WHO-estimated TB deaths among HIV-negative children aged under 5 years	children aged under 5 years immune status: HIV-negative vaccination status: mixed country: global period: 2023 Infant <1y1-4y	122 000 (110 000–135 000) TB deaths in HIV-negative children under 5 per year per child population per year	surveillance	who.int quote Among HIV-negative children and young adolescents, the estimated number of TB deaths in 2023 was 122 000 among children aged <5 years (95% UI: 110 000–135 000); and 44 000 in older children and young adolescents aged 5–14 years (95% UI: 39 000–48 000). Section 1.2 TB mortality, age/sex para The same sentence gives 44 000 (95% UI 39 000-48 000) deaths in HIV-negative children aged 5-14 years, so most paediatric TB deaths are in the under-fives, the group BCG is given to.
Death from tuberculosis in children severity: death case definition: Modelled TB deaths in children younger than 15 years, 217 countries, case-fatality-based approach	children younger than 15 years vaccination status: mixed country: global period: 2015 Infant <1y1-4y5-11y12-17y	239 000 (194 000–298 000) TB deaths in children per year per child population per year	modelling	Dodd et al. 2017 Lancet Glob Health quote FINDINGS: We estimated that 239 000 (95% uncertainty interval [UI] 194 000-298 000) children younger than 15 years died from tuberculosis worldwide in 2015; 80% (191 000, 95% UI 132 000-257 000) of these deaths were in children younger than 5 years. abstract, FINDINGS The same sentence reports that 80% (191 000) of these deaths were in children under 5 years.
Death from tuberculosis in children not receiving treatment severity: death case definition: Modelled share of paediatric TB deaths occurring in children not on TB treatment	children younger than 15 years vaccination status: mixed country: global period: 2015 Infant <1y1-4y5-11y12-17y	96 proportion of child TB deaths in untreated children (%) per child TB death	modelling	Dodd et al. 2017 Lancet Glob Health quote More than 96% (230 000, 185 000-289 000) of all tuberculosis deaths occurred in children not receiving tuberculosis treatment. abstract, FINDINGS Shows that paediatric TB mortality is almost entirely a problem of undiagnosed, untreated disease; a treated-child CFR (bcg-bur-011) is below 1%.
Incident tuberculosis disease in children severity: any case definition: Modelled number of children developing TB disease in the 22 high-burden countries	children under 15 years vaccination status: mixed country: 22 high-burden countries setting: LMIC period: 2010 Infant <1y1-4y5-11y12-17y	650,977 (424,871–983,118) incident child TB cases per year (median) per child population per year	modelling	Dodd et al. 2014 Lancet Glob Health quote In 2010, the median number of Mycobacterium tuberculosis infections in children was 7,591,759 (5,800,053-9,969,780), and 650,977 children (424,871-983,118) developed disease. abstract, FINDINGS Interval is an IQR of model runs, not a 95% CI. The model estimated that only 35% of paediatric cases were detected by notification systems.

Post-acute and long-term sequelae

What can follow the acute illness: chronic disease, disability, cancers, congenital syndromes, post-infectious conditions.

Outcome (O)	Population (P)	Comparator	Estimate	Design (S)	Source
All-cause mortality after tuberculosis treatment severity: death case definition: Deaths after treatment in cohorts with bacteriological or clinical TB confirmation and a general-population comparison group	all ages in included cohorts vaccination status: not reported country: multiple period: cohorts published 1997-2018 n: 40 781 individuals, 6922 deaths All ages	vs general population or matched controls	2.91 (2.21–3.84) standardised mortality ratio vs general population follow-up: at least 12 months after treatment completion	meta-analysis	Romanowski et al. 2019 Lancet Infect Dis quote The pooled SMR for all-cause mortality among people with tuberculosis, compared with the control group, was 2·91 (95% CI 2·21-3·84; I2=99%, pheterogeneity<0·0001). abstract, FINDINGS I2 99%. People who develop TB differ from the general population (poverty, HIV, smoking, alcohol, comorbidity), so part of the excess is confounding rather than a consequence of TB. Most post-treatment deaths were cardiovascular.
All-cause mortality after confirmed TB cure or treatment completion severity: death case definition: As above, restricted to confirmed treatment completion or cure	all ages in included cohorts vaccination status: not reported country: multiple period: cohorts published 1997-2018 All ages	vs general population or matched controls	3.76 (3.04–4.66) standardised mortality ratio vs general population follow-up: at least 12 months after treatment completion	meta-analysis	Romanowski et al. 2019 Lancet Infect Dis quote When restricted to people with confirmed treatment completion or cure, the pooled SMR was 3·76 (95% CI 3·04-4·66; I2=95%). abstract, FINDINGS Same confounding caveat as bcg-pa-006; I2 95%.
Reduced lung function after treated drug-susceptible tuberculosis severity: lung impairment case definition: FEV1 as percent of predicted on spirometry in TB survivors	adults and children in 54 included studies vaccination status: not reported country: multiple period: studies to November 2020 n: 14 621 people across the review All ages	vs predicted normal values (no uninfected control group)	76.6 (71.6–81.6) mean FEV1, % of predicted	meta-analysis	Ivanova et al. 2023 Eur Respir Rev quote For subjects with former drug-susceptible tuberculosis, the combined estimated mean was 76.6% (95% CI 71.6-81.6) of predicted for forced expiratory volume in 1 s (FEV1) and 81.8% (95% CI 77.4-86.2) for forced vital capacity (FVC). abstract, RESULTS Not a comparison with matched uninfected controls, so smoking, biomass exposure and pre-existing disease are not separated from the effect of TB. The same abstract gives 65.9% for multidrug-resistant TB survivors.
Severe lung impairment after tuberculosis severity: severe lung impairment case definition: Severe impairment on spirometry in TB survivors, as defined by each study	adults and children in included studies vaccination status: not reported country: multiple period: studies to November 2020 All ages	vs none (proportion of survivors)	10-15 proportion of survivors with severe lung impairment (%)	systematic review	Ivanova et al. 2023 Eur Respir Rev quote In the majority of studies, at least 10-15% of tuberculosis survivors had severe lung impairment. abstract, RESULTS A range reported as a floor ("at least") across most studies, not a pooled estimate; no CI.
People alive after surviving tuberculosis severity: any case definition: Modelled number of people who developed TB in 1980-2019 and were alive in 2020	all ages vaccination status: mixed country: global period: 2020 All ages	vs none (count)	155 (138–171) TB survivors alive, millions	modelling	Dodd et al. 2021 Lancet Infect Dis quote We estimate that 155 million tuberculosis survivors (95% UI 138 million-171 million) were alive in 2020, the largest proportion (47% [37-57]) of whom were in the WHO South-East Asia region. abstract, FINDINGS Context for the post-TB sequelae records above: the population at risk of post-TB lung disease and excess mortality is more than ten times the annual incidence.

Vaccine efficacy and effectiveness

Every estimate compiled, annotated by population (P), intervention (I), comparator (C), outcome (O) and study design (S).

Outcome (O)	Population (P)	Intervention / comparator	Estimate	Design (S)	Source
Tuberculosis, all forms severity: any case definition: TB cases as defined in each trial	all ages across 14 prospective trials country: multiple All ages	BCG, various strains platform: live attenuated M. bovis doses: 1 vs nonrecipients (unvaccinated or placebo)	0.49 (0.34–0.70) RR of TB (vaccinated vs unvaccinated); protective effect 51%	meta-analysis 14 prospective trials	Colditz et al. 1994 JAMA quote In the trials, the RR of TB was 0.49 (95% confidence interval [CI], 0.34 to 0.70) for vaccine recipients compared with nonrecipients (protective effect of 51%). abstract, DATA SYNTHESIS Random-effects summary across highly heterogeneous trials; latitude and study validity explained 66% of the heterogeneity. The abstract states age at vaccination did not improve prediction of efficacy.
Tuberculosis, all forms severity: any case definition: TB cases as defined in each case-control study	all ages across 12 case-control studies country: multiple All ages	BCG, various strains platform: live attenuated M. bovis doses: 1 vs unvaccinated controls	0.50 (0.39–0.64) OR of TB (vaccinated vs unvaccinated); protective effect 50%	meta-analysis 12 case-control studies	Colditz et al. 1994 JAMA quote In the case-control studies, the OR for TB was 0.50 (95% CI, 0.39 to 0.64), or a 50% protective effect. abstract, DATA SYNTHESIS Observational estimate, consistent with the trial estimate above.
Death from tuberculosis severity: death case definition: Tuberculous deaths in trials	all ages across 7 trials country: multiple All ages	BCG, various strains platform: live attenuated M. bovis doses: 1 vs nonrecipients	0.29 (0.16–0.53) RR of TB death; protective effect 71%	meta-analysis 7 trials	Colditz et al. 1994 JAMA quote Seven trials reporting tuberculous deaths showed a protective effect from BCG vaccine of 71% (RR, 0.29; 95% CI, 0.16 to 0.53), and five studies reporting on meningitis showed a protective effect from BCG vaccine of 64% (OR, 0.36; 95% CI, 0.18 to 0.70). abstract, DATA SYNTHESIS Higher protection against death than against all TB; the authors caution this may partly reflect less misclassification of severe outcomes rather than greater efficacy. The same sentence gives OR 0.36 (0.18-0.70) for meningitis from 5 studies.
Tuberculosis, all forms, after infant vaccination severity: any case definition: TB cases in four randomized trials of BCG given in infancy	vaccinated as newborns or infants country: multiple NeonateInfant <1y1-4y5-11y	BCG, various strains platform: live attenuated M. bovis doses: 1 schedule: at birth or in infancy vs unvaccinated infants	0.74 (0.62–0.83) protective effect (1-RR), proportion	meta-analysis 4 randomized controlled trials	Colditz et al. 1995 Pediatrics quote Overall, the protective effect of vaccination against cases of tuberculosis was 0.74 (95% confidence interval [95% CI], 0.62 to 0.83) when estimated from four randomized controlled trials, and 0.52 (95% CI, 0.38 to 0.64) when estimated from nine case-control studies. abstract, DATA SYNTHESIS Protective effect is reported as a proportion (0.74 = 74%). The same sentence gives 0.52 (0.38-0.64) from nine case-control studies.
Death from tuberculosis after infant vaccination severity: death case definition: TB deaths in trials of infant BCG	vaccinated as newborns or infants country: multiple NeonateInfant <1y1-4y5-11y	BCG, various strains platform: live attenuated M. bovis doses: 1 schedule: at birth or in infancy vs unvaccinated infants	0.65 (0.12–0.86) protective effect (1-RR), proportion	meta-analysis 5 trials	Colditz et al. 1995 Pediatrics quote Five trials reporting deaths from tuberculosis showed a BCG protective effect of 0.65 (95% CI, 0.12 to 0.86), five studies reporting on meningitis showed a protective effect of 0.64 (95% CI, 0.30 to 0.82), and three studies of disseminated tuberculosis showed a protective effect of 0.78 (95% CI, 0.58 to 0.88). abstract, DATA SYNTHESIS Wide interval. Same sentence: meningitis 0.64 (0.30-0.82) from five studies and disseminated TB 0.78 (0.58-0.88) from three.
Laboratory-confirmed tuberculosis after infant vaccination severity: laboratory-confirmed disease case definition: Laboratory-confirmed TB in three case-control studies	vaccinated as newborns or infants country: multiple NeonateInfant <1y1-4y5-11y	BCG, various strains platform: live attenuated M. bovis doses: 1 schedule: at birth or in infancy vs unvaccinated controls	0.83 (0.58–0.93) protective effect (1-OR), proportion	meta-analysis 3 case-control studies	Colditz et al. 1995 Pediatrics quote These studies documented a protective effect of 0.83 (95% CI, 0.58 to 0.93). abstract, DATA SYNTHESIS Highest estimate in the review; the authors attribute it to less outcome misclassification when cases are laboratory confirmed.
Pulmonary tuberculosis severity: pulmonary disease case definition: Pulmonary TB as defined in each randomized trial	children (school age), stringently tuberculin tested before vaccination prior infection: tuberculin-negative after stringent testing (prior M. tuberculosis infection or environmental sensitisation excluded) country: multiple 5-11y12-17y	BCG, various strains platform: live attenuated M. bovis doses: 1 vs unvaccinated or placebo	0.26 (.18–.37) rate ratio (vaccinated vs unvaccinated)	meta-analysis randomized trials (18 reporting pulmonary TB)	Mangtani et al. 2014 Clin Infect Dis quote Protection appeared greatest in children stringently tuberculin tested, to try to exclude prior infection with Mycobacterium tuberculosis or sensitization to environmental mycobacteria (rate ratio [RR], 0.26; 95% confidence interval [CI], .18-.37), or infants (RR, 0.41; 95% CI, .29-.58). abstract, RESULTS Univariable meta-regression subgroup. The same sentence gives RR 0.41 (.29-.58) for infants (next record).
Pulmonary tuberculosis severity: pulmonary disease case definition: Pulmonary TB as defined in each randomized trial	infants country: multiple NeonateInfant <1y	BCG, various strains platform: live attenuated M. bovis doses: 1 schedule: in infancy vs unvaccinated or placebo	0.41 (.29–.58) rate ratio (vaccinated vs unvaccinated)	meta-analysis randomized trials of infant vaccination	Mangtani et al. 2014 Clin Infect Dis quote Protection appeared greatest in children stringently tuberculin tested, to try to exclude prior infection with Mycobacterium tuberculosis or sensitization to environmental mycobacteria (rate ratio [RR], 0.26; 95% confidence interval [CI], .18-.37), or infants (RR, 0.41; 95% CI, .29-.58). abstract, RESULTS Infants have no prior mycobacterial exposure, which the review links to higher efficacy.
Pulmonary tuberculosis severity: pulmonary disease case definition: Pulmonary TB as defined in each randomized trial	children not stringently tuberculin tested prior infection: not stringently screened for prior tuberculin sensitivity country: multiple 5-11y12-17y	BCG, various strains platform: live attenuated M. bovis doses: 1 vs unvaccinated or placebo	0.59 (.35–1.01) rate ratio (vaccinated vs unvaccinated)	meta-analysis randomized trials	Mangtani et al. 2014 Clin Infect Dis quote Protection was weaker in children not stringently tested (RR, 0.59; 95% CI, .35-1.01) and older individuals stringently or not stringently tested (RR, 0.88; 95% CI, .59-1.31 and RR, 0.81; 95% CI, .55-1.22, respectively). abstract, RESULTS Interval touches 1. The same sentence gives older individuals: RR 0.88 (.59-1.31) stringently tested and 0.81 (.55-1.22) not stringently tested (next record).
Pulmonary tuberculosis severity: pulmonary disease case definition: Pulmonary TB as defined in each randomized trial	older individuals (adolescents and adults), stringently tuberculin tested prior infection: tuberculin-negative after stringent testing country: multiple 12-17y18-49y50-64y	BCG, various strains platform: live attenuated M. bovis doses: 1 vs unvaccinated or placebo	0.88 (.59–1.31) rate ratio (vaccinated vs unvaccinated)	meta-analysis randomized trials	Mangtani et al. 2014 Clin Infect Dis quote Protection was weaker in children not stringently tested (RR, 0.59; 95% CI, .35-1.01) and older individuals stringently or not stringently tested (RR, 0.88; 95% CI, .59-1.31 and RR, 0.81; 95% CI, .55-1.22, respectively). abstract, RESULTS No demonstrated protection against pulmonary TB when BCG is given to older people; the not-stringently-tested older group in the same sentence is RR 0.81 (.55-1.22). This is the null that drives the "variable adult efficacy" statement in policy documents.
Pulmonary tuberculosis severity: pulmonary disease case definition: Pulmonary TB in the UK MRC trial of tuberculin-negative adolescents	adolescents (14-15 years at entry) prior infection: tuberculin-negative at entry country: United Kingdom 12-17y18-49y	BCG (Danish strain, Copenhagen) platform: live attenuated M. bovis doses: 1 vs unvaccinated tuberculin-negative controls	0.22 (0.16–0.31) rate ratio (vaccinated vs unvaccinated)	systematic review	Abubakar et al. 2013 Health Technol Assess quote Protection against pulmonary tuberculosis in adults is variable, ranging from substantial protection in the UK MRC trial {rate ratio 0.22 [95% confidence interval (CI) 0.16 to 0.31]}, to absence of clinically important benefit, as in the large Chingleput trial [rate ratio 1.05 (95% CI 0.88 to 1.25)] and greater in latitudes further away from the equator. abstract, RESULTS Efficacy The high-latitude, stringently tested extreme of the efficacy range; the same sentence gives the Chingleput null (next record).
Pulmonary tuberculosis severity: pulmonary disease case definition: Bacillary pulmonary TB in the Chingleput (south India) trial	all ages (adults and children) country: India setting: community, high transmission, near the equator All ages	BCG (French and Danish strains, two doses levels) platform: live attenuated M. bovis doses: 1 vs placebo	1.05 (0.88–1.25) rate ratio (vaccinated vs placebo)	systematic review	Abubakar et al. 2013 Health Technol Assess quote Protection against pulmonary tuberculosis in adults is variable, ranging from substantial protection in the UK MRC trial {rate ratio 0.22 [95% confidence interval (CI) 0.16 to 0.31]}, to absence of clinically important benefit, as in the large Chingleput trial [rate ratio 1.05 (95% CI 0.88 to 1.25)] and greater in latitudes further away from the equator. abstract, RESULTS Efficacy No benefit. A null result in a tropical, high-exposure population with widespread non-specific tuberculin sensitivity; the review attributes low efficacy to latitude and failure to exclude prior sensitisation.
M. tuberculosis infection (IGRA positivity) after exposure severity: infection case definition: Positive interferon-gamma release assay in children under 16 with known recent exposure to pulmonary TB	children under 16 years prior infection: recent household or congregate exposure to pulmonary TB country: multiple setting: household and congregate contact n: 3855 participants Infant <1y1-4y5-11y12-17y	BCG platform: live attenuated M. bovis doses: 1 vs unvaccinated exposed children	0.81 (0.71–0.92) RR of infection (vaccinated vs unvaccinated); protective efficacy 19%	meta-analysis 14 studies	Roy et al. 2014 BMJ quote The estimated overall risk ratio was 0.81 (95% confidence interval 0.71 to 0.92), indicating a protective efficacy of 19% against infection among vaccinated children after exposure compared with unvaccinated children. abstract, RESULTS Modest protection against acquiring infection, as opposed to disease.
Active tuberculosis after exposure severity: active disease case definition: Active TB at screening in the six studies with progression data	children under 16 years prior infection: recent exposure to pulmonary TB country: multiple setting: household and congregate contact n: 1745 participants Infant <1y1-4y5-11y12-17y	BCG platform: live attenuated M. bovis doses: 1 vs unvaccinated exposed children	0.29 (0.15–0.58) RR of active TB (vaccinated vs unvaccinated); protection 71%	meta-analysis 6 studies	Roy et al. 2014 BMJ quote Restriction of the analysis to the six studies (n=1745) with information on progression to active tuberculosis at the time of screening showed protection against infection of 27% (risk ratio 0.73, 0.61 to 0.87) compared with 71% (0.29, 0.15 to 0.58) against active tuberculosis. abstract, RESULTS Same sentence: 27% protection against infection (RR 0.73, 0.61 to 0.87) in these six studies. Among those infected, protection against progression to disease was 58% (0.42, 0.23 to 0.77).
M. tuberculosis infection (QuantiFERON conversion) in household contacts severity: infection case definition: QuantiFERON conversion at 0.7 IU/mL or more among initially test-negative household contacts	household contacts over 2 years of age (children 0-2 excluded) immune status: HIV-negative prior infection: household exposure, baseline IGRA/TST negative country: multiple setting: household contact n: 1491 household contacts 1-4y5-11y12-17y18-49y	BCG (primary vaccination, neonatal policy countries) platform: live attenuated M. bovis doses: 1 vs unvaccinated household contacts	0.65 (0.51–0.82) adjusted HR of QuantiFERON conversion (vaccinated vs unvaccinated)	meta-analysis individual participant data from household contact studies	Pelzer et al. 2025 Lancet Microbe quote Among 1491 household contacts, but not among 5644 adolescents, the BCG vaccine protected against QuantiFERON conversion at the primary cutoff of 0·7 IU/mL or more with the adjusted hazard ratio (0·65, 95% CI 0·51-0·82) being consistent with that for protection against disease (0·68, 0·18-2·59). abstract, FINDINGS Same sentence: protection against disease in this group aHR 0.68 (0.18-2.59), consistent but imprecise. No protection against TST conversion was seen in any study type, and no QuantiFERON protection in 5644 adolescents.
NON-SPECIFIC EFFECT: all-cause neonatal mortality (not tuberculosis) severity: death case definition: Deaths from any cause within the neonatal period in three RCTs of early vs delayed BCG in low-weight neonates	neonates weighing under 2500 g risk group: low birth weight country: Guinea-Bissau setting: LMIC, high neonatal mortality period: 2002-2013 n: 3 trials (6583 infants per the 2018 companion paper) Neonate	BCG-Denmark (Statens Serum Institut) platform: live attenuated M. bovis doses: 1 schedule: at birth/first contact vs delayed vs local policy (delayed BCG) for low-weight neonates	0.62 (.46–.83) mortality rate ratio (early BCG vs delayed/no BCG), neonatal period follow-up: neonatal period (28 days)	meta-analysis	Biering-Sørensen et al. 2017 Clin Infect Dis quote A meta-analysis of 3 BCG trials showed that early BCG-Denmark reduced mortality by 38% (MRR, 0.62; 95% CI, .46-.83) within the neonatal period and 16% (0.84; .71-1.00) by age 12 months. abstract, RESULTS Non-specific effect: TB does not cause neonatal deaths at this timescale, so this is heterologous protection. Same sentence: 16% reduction by 12 months (MRR 0.84; .71-1.00). Open-label trials by one research group; replication elsewhere is limited.
NON-SPECIFIC EFFECT: all-cause mortality within one year (not tuberculosis) severity: death case definition: All-cause mortality in RCTs of BCG published 2011-2022, follow-up restricted to one year	all ages in included RCTs (mostly neonates and adults in COVID-era trials) country: multiple All ages	BCG, various strains platform: live attenuated M. bovis doses: 1 vs no BCG or placebo	0.79 (0.64–0.99) HR for all-cause mortality (BCG vs control), 1-year follow-up follow-up: 1 year	meta-analysis RCTs (16 trials, 34,197 participants in the review)	Trunk et al. 2023 Vaccines (Basel) quote There was evidence for a protective effect on all-cause mortality of 21% if follow-up was restricted to one year (HR 0.79, 95% CI 0.64-0.99). abstract, RESULTS Same sentence context: no evidence of effect with longer follow-up (HR 0.88, 0.75-1.03). Heterogeneous populations (neonates in LMICs and adults in COVID-19 trials) pooled together.
NON-SPECIFIC EFFECT: all-cause mortality in children under 5 (not tuberculosis) severity: death case definition: All-cause mortality in clinical trials of BCG, WHO-commissioned review	children under 5 years country: multiple NeonateInfant <1y1-4y	BCG platform: live attenuated M. bovis doses: 1 vs no BCG	0.70 (0.49–1.01) average RR of all-cause mortality (BCG vs no BCG), clinical trials	systematic review 5 clinical trials	Higgins et al. 2016 BMJ quote Receipt of BCG vaccine was associated with a reduction in all cause mortality: the average relative risks were 0.70 (95% confidence interval 0.49 to 1.01) from five clinical trials and 0.47 (0.32 to 0.69) from nine observational studies at high risk of bias. abstract, RESULTS Interval includes 1. The same sentence gives RR 0.47 (0.32 to 0.69) from nine observational studies judged at high risk of bias; the trial estimate is the one filed here.
Bacillary pulmonary tuberculosis in children severity: bacillary pulmonary disease case definition: Sputum-positive pulmonary TB over 15 years of follow-up, Chingleput trial	children (tuberculin-negative at intake) prior infection: tuberculin-negative with normal chest X-ray at intake country: India setting: community, high transmission period: 1968-1983 n: 109,873 tuberculin-negative persons with normal chest X-ray 1-4y5-11y12-17y	BCG (French and Danish strains, high and low dose) platform: live attenuated M. bovis doses: 1 vs placebo	27 (-8–50) VE (%) follow-up: 15 years	RCT	1999 Indian J Med Res quote BCG offered no overall protection in adults and a low level of overall protection (27%; 95% C.I. -8 to 50%) in children. abstract Interval includes zero. The same abstract states BCG offered no overall protection in adults in this population; the 15-year result confirmed the complete lack of efficacy seen at 7.5 years.
Tuberculosis, all forms, over 60 years severity: any case definition: TB events in health service records, registries and death certificates 1948-1998	vaccinated as children/young people in 1935-1938, followed to 1998 country: United States (American Indians and Alaska Natives) period: 1948-1998 n: 1483 BCG and 1309 placebo participants 5-11y12-17y18-49y50-64y65+	BCG (Phipps strain, intradermal) platform: live attenuated M. bovis doses: 1 time since vaccination: 10 to 60 years vs placebo	52 (27–69) VE (%) follow-up: 50-60 years after vaccination	RCT	Aronson et al. 2004 JAMA quote RESULTS: The overall incidence of tuberculosis was 66 and 138 cases per 100 000 person-years in the BCG vaccine and placebo groups, respectively, for an estimate of vaccine efficacy of 52% (95% confidence interval, 27%-69%). abstract, RESULTS Efficacy persisted for 50-60 years with only slight, non-significant waning (greater in men). Retrospective record review of a 1930s placebo-controlled trial.
Tuberculosis, all forms severity: any case definition: TB cases ascertained over 20 years in the MRC trial	14-15 years at entry prior infection: tuberculin-negative at entry country: Great Britain period: 1950-1972 n: 54 239 participants 12-17y18-49y	BCG (Danish strain) platform: live attenuated M. bovis doses: 1 time since vaccination: 0-5 years vs unvaccinated tuberculin-negative controls	84 VE (%), first five years follow-up: first 5 years	RCT	Hart et al. 1977 Br Med J quote The protective efficacy of each of the two vaccines, among those initially tuberculin negative, was 84% during the first five years, and gradually decreased, averaging 77% for each vaccine over the whole period. abstract The same sentence gives an average of 77% over the whole 20-year period. Only 27 of 610 cases arose in years 15-20, so efficacy in that final period could not be reliably assessed.
Tuberculosis, all forms severity: any case definition: TB cases over 15 years (70% pulmonary, 15% pleural effusion)	14-15 years at entry prior infection: tuberculin-negative at entry country: Great Britain period: 1950-1967 n: 13 598 BCG-vaccinated participants 12-17y18-49y	BCG (Danish strain) platform: live attenuated M. bovis doses: 1 vs unvaccinated tuberculin-negative controls admitted concurrently	78 VE (%), 15 years (0.28 vs 1.28 per 1,000 per year) follow-up: 15 years	RCT	Fourth 1972 Bull World Health Organ quote The annual incidence (per 1 000) of TB was 0.28 among the 13 598 BCG-vaccinated participants and 1.28 among the tuberculin-negative unvaccinated participants admitted concurrently (a 78% reduction); for vole bacillus vaccine (5 817 participants) the figures were 0.29 and 1.50 (an 81% reduction). abstract Same abstract: 10 cases of tuberculous meningitis or miliary TB in unvaccinated participants and none in vaccinated; protection similar in both sexes.
Tuberculosis, all forms, 10-15 years after vaccination severity: any case definition: TB cases arising 10-15 years after entry	14-15 years at entry (24-30 years at this interval) prior infection: tuberculin-negative at entry country: Great Britain period: 1960-1967 18-49y	BCG (Danish strain) platform: live attenuated M. bovis doses: 1 time since vaccination: 10-15 years vs unvaccinated tuberculin-negative controls	59 VE (%), years 10-15 follow-up: 10-15 years	RCT	Fourth 1972 Bull World Health Organ quote It decreased gradually during the period but was still substantial (59% for BCG vaccine and 73% for vole bacillus vaccine) 10-15 years after vaccination; no definite conclusion can yet be reached on the level of protection after 15 years. abstract Waning from 84% in the first 5 years to 59% at 10-15 years in the same trial.
Tuberculosis, all forms, 15-20 years after neonatal vaccination severity: any case definition: TB cases identified by record linkage in the control arm of the BCG-REVAC trial	school children aged 7-14 at enrolment, followed to 15-20 years after neonatal vaccination country: Brazil 5-11y12-17y	BCG (neonatal, Brazilian programme) platform: live attenuated M. bovis doses: 1 schedule: at birth time since vaccination: 15-20 years vs children without a BCG scar (unvaccinated) in the control arm	39 (9–58) VE (%) follow-up: 15-20 years after vaccination	RCT	Barreto et al. 2005 Int J Tuberc Lung Dis quote In the control arm of a trial of efficacy of revaccination of schoolchildren in Brazil we found substantial protection (39%; 95%CI 9-58) of neonatal BCG against all forms of tuberculosis (TB) 15-20 years after vaccination, much longer than previously believed. abstract Observational comparison within the control arm of a cluster trial (scar-based exposure), so confounding by scar failure is possible. Evidence that neonatal BCG protection lasts beyond 10 years.
Tuberculosis, all forms severity: any case definition: TB cases identified by record linkage to the Tuberculosis Control Programme	school children aged 7-14 years prior infection: one BCG vaccination in infancy (BCG scar) country: Brazil (Salvador and Manaus) period: 5-year follow-up n: 176,846 revaccination and 171,293 control children assigned 5-11y12-17y	BCG (second dose at school age) platform: live attenuated M. bovis doses: 2 schedule: neonatal dose plus revaccination at 7-14 years vs no revaccination (neonatal BCG only)	9 (-16–29) VE (%) of revaccination follow-up: 5 years	cluster RCT	Rodrigues et al. 2005 Lancet quote The efficacy of BCG revaccination was 9% (-16 to 29%). abstract, FINDINGS No substantial added protection; the authors recommend against school-age revaccination. Crude rate ratio 0.97 (95% CI 0.76-1.28) in the same abstract.
Tuberculosis, all forms severity: any case definition: TB cases by record linkage, extended 9-year follow-up	school children aged 7-14 years at revaccination prior infection: neonatal BCG vaccination country: Brazil (Salvador and Manaus) period: 9-year follow-up n: over 200,000 children 5-11y12-17y	BCG (second dose at school age) platform: live attenuated M. bovis doses: 2 vs no revaccination (neonatal BCG only)	12 (-2–24) VE (%) of revaccination follow-up: 9 years	cluster RCT	Barreto et al. 2011 Vaccine quote The overall vaccine efficacy was 12% (-2 to 24%) as compared to 9% (-16 to 29%) for the 5-year follow up. abstract Marginal at best. City-specific estimates follow.
Tuberculosis, all forms severity: any case definition: TB cases by record linkage, extended 9-year follow-up	school children aged 7-14 years at revaccination prior infection: neonatal BCG vaccination country: Brazil, Salvador (further from the equator) period: 9-year follow-up 5-11y12-17y	BCG (second dose at school age) platform: live attenuated M. bovis doses: 2 vs no revaccination	19 (3–33) VE (%) of revaccination follow-up: 9 years	cluster RCT	Barreto et al. 2011 Vaccine quote Vaccine efficacy was higher in Salvador (19%, 3 to 33%) than in Manaus (1%, -27 to 27%) with the highest vaccine efficacy in children from Salvador aged <11 years at revaccination (33%, 3 to 54%). abstract Same sentence: Manaus (nearer the equator) 1% (-27 to 27%) and Salvador children under 11 at revaccination 33% (3 to 54%). Consistent with lower efficacy where environmental mycobacteria are more prevalent.
NON-SPECIFIC EFFECT: all-cause neonatal mortality (not tuberculosis) severity: death case definition: Deaths from any cause in the neonatal period, single RCT 2008-2013	neonates weighing under 2500 g risk group: low birth weight country: Guinea-Bissau setting: LMIC, high neonatal mortality period: 2008-2013 n: 2083 early BCG vs 2089 control Neonate	BCG-Denmark (Statens Serum Institut) platform: live attenuated M. bovis doses: 1 schedule: at birth/first contact vs local policy for low-weight neonates, no early BCG	0.70 (.47–1.04) mortality rate ratio (early BCG vs control), neonatal period follow-up: neonatal period	RCT	Biering-Sørensen et al. 2017 Clin Infect Dis quote RESULTS: Early administration of BCG-Denmark was associated with a nonsignificant reduction in neonatal mortality rate (MRR, 0.70; 95% confidence interval [CI], .47-1.04) and a 34% reduction (0.66; .44-1.00) when censoring for oral poliovirus vaccine campaigns. abstract, RESULTS Prespecified primary hypothesis of a 45% reduction was not met; interval includes 1. Infectious-disease mortality MRR 0.57 (.35-.93) in the same abstract. Unblinded.
NON-SPECIFIC EFFECT: all-cause infant mortality (not tuberculosis) severity: death case definition: Deaths from any cause by 12 months in an RCT of BCG at birth vs delayed BCG in low-birth-weight children	low-birth-weight infants, birth to 12 months risk group: low birth weight country: Guinea-Bissau setting: LMIC period: 2004-2008 n: 2320 children NeonateInfant <1y	BCG (at birth) platform: live attenuated M. bovis doses: 1 schedule: at birth vs delayed vs delayed BCG (usual policy)	.83 (.63–1.08) mortality rate ratio (early vs delayed BCG), 12 months follow-up: 12 months	RCT	Aaby et al. 2011 J Infect Dis quote In the primary analysis, infant mortality was reduced insignificantly by 17% (mortality rate ratio [MRR] = .83 [.63-1.08]). abstract, RESULTS Primary analysis, not statistically significant; trial was powered for a much higher pretrial mortality than observed. Secondary analyses showed MRR .55 (.34-.89) after 4 weeks.

Vaccine safety and adverse events

Harms measured after vaccination against an unvaccinated or baseline comparator. Kept separate from post-acute sequelae, which follow the disease, not the vaccine.

Outcome (O)	Population (P)	Intervention / comparator	Estimate	Design (S)	Source
Fatal disseminated BCG disease after vaccination severity: death case definition: Fatal disseminated BCG infection reported after vaccination, as summarised by ACET/ACIP	all ages (primarily immunocompromised persons) immune status: deaths occurred primarily among immunocompromised persons country: multiple period: reports summarised to 1996 All ages	BCG, various strains platform: live attenuated M. bovis doses: 1 vs none (passive reports per doses distributed)	0.06-1.56 fatal disseminated BCG cases per million doses administered per million doses administered	expert body	cdc.gov quote Fatal disseminated BCG disease has occurred at a rate of 0.06-1.56 cases per million doses of vaccine administered ( Table_1 ); these deaths occurred primarily among immunocompromised persons. Adverse reactions section Passive-surveillance range, no controlled comparator. The same document gives BCG osteitis ranging from 0.01 per million vaccinees in Japan to 32.5 and 43.4 per million in Sweden and Finland, varying with strain and production method.
Disseminated BCG disease after vaccination in HIV-infected infants severity: disseminated vaccine-strain disease case definition: Microbiologically confirmed disseminated BCG disease in confirmed HIV-infected infants, multicentre prospective surveillance	infants aged 1 year or under immune status: HIV-infected country: South Africa setting: high HIV and TB burden period: 2004-2006 NeonateInfant <1y	BCG (Danish strain, routine birth dose) platform: live attenuated M. bovis doses: 1 schedule: at birth vs none (population-based incidence; denominator estimated from maternal HIV prevalence and transmission rates)	992 (567–1495) incidence per 100,000 BCG-vaccinated HIV-infected infants per 100,000 BCG-vaccinated HIV-infected infants	surveillance	Hesseling et al. 2009 Bull World Health Organ quote The pooled incidence over the study period was 992 (95% CI: 567-1495) per 100 000. abstract, FINDINGS Uncontrolled population-based rate, likely an under-estimate per the authors. This risk is why WHO (2018 position paper) advises deferring BCG in infants known to be HIV-infected until antiretroviral therapy is established. Not a controlled comparison.

Tuberculous meningitis and miliary tuberculosis in children Mycobacterium tuberculosis

Disease risks

Acute risks of the infection itself: complications, hospitalisation, ICU, death.

Outcome (O)	Population (P)	Estimate	Design (S)	Source
Death during treatment for childhood tuberculous meningitis severity: death case definition: Childhood TBM diagnosed by predefined criteria in 19 treatment-outcome studies	children (paediatric TBM series) vaccination status: not reported country: multiple setting: hospital period: studies published up to 2012 n: 1636 children Infant <1y1-4y5-11y12-17y	19.3 (14.0–26.1) CFR during treatment (%) per treated childhood TBM case	meta-analysis	Chiang et al. 2014 Lancet Infect Dis quote Risk of death was 19·3% (95% CI 14·0-26·1) and probability of survival without neurological sequelae was 36·7% (27·9-46·4). abstract, FINDINGS Treated cases only; untreated TBM is regarded as uniformly fatal. 47% of children were diagnosed at the most advanced stage, which worsened outcomes.
Death from tuberculous meningitis with hydrocephalus in children severity: death case definition: Probable and definite TBM with hydrocephalus treated at a South African centre, assessed at 6 months	median 3 years 3 months (range 3 months to 13 years) vaccination status: not reported country: South Africa setting: hospital period: prospective cohort, publication 2016 n: 44 children Infant <1y1-4y5-11y12-17y	16 mortality at 6 months (%) per treated TBM-with-hydrocephalus case	cohort	Rohlwink et al. 2016 Dev Med Child Neurol quote The mortality rate was 16%, three patients (6.8%) were in a persistent vegetative state, two were severely disabled (4.5%), and 11 (25%) suffered mild-moderate disability. abstract, RESULTS Small single-centre cohort of the more severe hydrocephalus subgroup; disability outcomes from the same cohort are filed as post-acute records.
Tuberculous meningitis in children severity: meningitis case definition: Modelled incident childhood tuberculous meningitis, treated and untreated	children younger than 15 years vaccination status: mixed country: global period: 2019 Infant <1y1-4y5-11y12-17y	24 000 (22 300–25 700) incident childhood TB meningitis cases per year per child population per year	modelling	du et al. 2025 Lancet Glob Health quote FINDINGS: An estimated 24 000 (95% credible interval 22 300-25 700) children younger than 15 years developed tuberculous meningitis in 2019. abstract, FINDINGS Bayesian model; interval is a 95% credible interval. Only 13 000 of these were estimated to have been diagnosed and treated; most untreated children were under 5.
Death from tuberculous meningitis in children severity: death case definition: Modelled deaths among the 24 000 children with tuberculous meningitis, assuming untreated disease is always fatal	children younger than 15 years vaccination status: mixed country: global period: 2019 Infant <1y1-4y5-11y12-17y	16 100 (14 900–17 300) childhood TB meningitis deaths per year per child population per year	modelling	du et al. 2025 Lancet Glob Health quote Among the 24 000 children with tuberculous meningitis, 16 100 (14 900-17 300) were estimated to have died in 2019, of whom 1101 (6·8%) had HIV. abstract, FINDINGS Roughly two in three children with TB meningitis die in this model, largely because untreated disease is assumed fatal. 83.1% of deaths were in children under 5 and 68.3% in untreated children (same abstract).
Death from tuberculous meningitis in children under 5 severity: death case definition: Share of modelled childhood TB meningitis deaths occurring in children younger than 5 years	children younger than 5 years vaccination status: mixed country: global period: 2019 Infant <1y1-4y	83.1 proportion of childhood TBM deaths in under-fives (%) per childhood TB meningitis death	modelling	du et al. 2025 Lancet Glob Health quote 13 380 (83·1%) of 16 100 deaths were estimated to be in children younger than 5 years and 11 000 (68·3%) were estimated to be in children who did not receive tuberculous-meningitis treatment. abstract, FINDINGS The under-fives are the BCG target group; the same sentence gives 68.3% of deaths in children who received no TBM treatment.
Death from miliary tuberculosis in children severity: death case definition: Childhood miliary TB diagnosed clinically with classic chest radiograph pattern at a South African hospital	median 10.5 months vaccination status: vaccinated country: South Africa setting: hospital period: 1985-1989 n: 94 children Infant <1y1-4y5-11y	14 CFR (%) per hospitalised miliary TB case	case series	Hussey et al. 1991 Pediatr Infect Dis J quote Meningitis occurred in 19% of patients and this was the only significant risk factor identified for mortality, the overall case fatality rate being 14%. abstract 88% of the children had a documented BCG vaccination history, so these are largely breakthrough cases in a very high-burden setting. Meningitis complicated 19% of cases and was the only risk factor for death.
Miliary tuberculosis by age severity: disseminated disease case definition: Share of childhood miliary TB cases in infants under 1 year	children, median 10.5 months vaccination status: vaccinated country: South Africa setting: hospital period: 1985-1989 n: 94 children Infant <1y1-4y5-11y	52 proportion of cases aged under 1 year (%) per childhood miliary TB case	case series	Hussey et al. 1991 Pediatr Infect Dis J quote The median age at presentation was 10.5 months, 52% of cases occurring in those younger than 1 year. abstract Shows that miliary TB concentrates in the first year of life, before and around the time BCG-induced protection is established.

Post-acute and long-term sequelae

What can follow the acute illness: chronic disease, disability, cancers, congenital syndromes, post-infectious conditions.

Outcome (O)	Population (P)	Comparator	Estimate	Design (S)	Source
Neurological sequelae after childhood tuberculous meningitis severity: neurological disability case definition: Any neurological sequelae at end of treatment among survivors in 19 studies	children (paediatric TBM series) vaccination status: not reported country: multiple setting: hospital period: studies published up to 2012 n: 1636 children Infant <1y1-4y5-11y12-17y	vs none (proportion among TBM survivors)	53.9 (42.6–64.9) proportion of survivors with neurological sequelae (%) follow-up: end of treatment	meta-analysis	Chiang et al. 2014 Lancet Infect Dis quote Among survivors, risk of neurological sequelae was 53·9% (95% CI 42·6-64·9). abstract, FINDINGS Denominator is survivors, not all cases; do not combine with the 19.3% death risk by simple addition. Sequelae definitions varied across studies.
Survival without neurological sequelae after childhood tuberculous meningitis severity: composite of death or disability case definition: Alive and without neurological sequelae at end of treatment	children (paediatric TBM series) vaccination status: not reported country: multiple setting: hospital period: studies published up to 2012 n: 1636 children Infant <1y1-4y5-11y12-17y	vs none (proportion among all treated TBM cases)	36.7 (27.9–46.4) proportion surviving without sequelae (%) follow-up: end of treatment	meta-analysis	Chiang et al. 2014 Lancet Infect Dis quote Risk of death was 19·3% (95% CI 14·0-26·1) and probability of survival without neurological sequelae was 36·7% (27·9-46·4). abstract, FINDINGS Only about one in three treated children leaves TB meningitis alive and intact.
Disability after tuberculous meningitis with hydrocephalus severity: mild-moderate disability case definition: Paediatric Cerebral Performance Category at 6 months	median 3 years 3 months vaccination status: not reported country: South Africa setting: hospital period: prospective cohort, publication 2016 n: 44 children Infant <1y1-4y5-11y12-17y	vs none (proportion of cohort)	25 proportion with mild-moderate disability at 6 months (%) follow-up: 6 months	cohort	Rohlwink et al. 2016 Dev Med Child Neurol quote The mortality rate was 16%, three patients (6.8%) were in a persistent vegetative state, two were severely disabled (4.5%), and 11 (25%) suffered mild-moderate disability. abstract, RESULTS Same sentence: 6.8% in a persistent vegetative state and 4.5% severely disabled; all cases showed neurodevelopmental deficits relative to controls. Small cohort restricted to hydrocephalus.
Moderate-to-severe cognitive impairment after childhood tuberculous meningitis severity: moderate-to-severe cognitive impairment case definition: Cognitive impairment assessed at the conclusion of anti-tuberculous treatment	children diagnosed with TBM vaccination status: not reported country: South Africa setting: hospital period: 1985-2015 n: 327 children Infant <1y1-4y5-11y12-17y	vs none (proportion of cohort)	21.7 proportion with poor developmental outcome (%) follow-up: end of anti-tuberculous treatment	cohort	Saal et al. 2023 Childs Nerv Syst quote RESULTS: A total of 327 TBM patients were included, 71 (21.7%) suffered a poor outcome. abstract, RESULTS Retrospective single-centre cohort over 30 years; outcome measured at end of treatment, so later recovery or deterioration is not captured.
Neurological sequelae in childhood tuberculous meningitis survivors severity: neurological disability case definition: Modelled number of surviving children with neurological sequelae	children younger than 15 years vaccination status: mixed country: global period: 2019 Infant <1y1-4y5-11y12-17y	vs none (count among 7900 modelled survivors)	5550 (5110–5980) survivors with neurological sequelae per year	modelling	du et al. 2025 Lancet Glob Health quote Of the 7900 (5800-10 000) children who did not die, 5550 (5110-5980) were estimated to have neurological sequelae. abstract, FINDINGS Model assumed the same age-specific sequelae risk ratios as for death; credible intervals, not CIs.

Vaccine efficacy and effectiveness

Every estimate compiled, annotated by population (P), intervention (I), comparator (C), outcome (O) and study design (S).

Outcome (O)	Population (P)	Intervention / comparator	Estimate	Design (S)	Source
Disseminated tuberculosis after infant vaccination severity: disseminated disease case definition: Disseminated (miliary) TB in studies of infant BCG	vaccinated as newborns or infants country: multiple NeonateInfant <1y1-4y5-11y	BCG, various strains platform: live attenuated M. bovis doses: 1 schedule: at birth or in infancy vs unvaccinated infants	0.78 (0.58–0.88) protective effect (1-RR or 1-OR), proportion	meta-analysis 3 studies	Colditz et al. 1995 Pediatrics quote Five trials reporting deaths from tuberculosis showed a BCG protective effect of 0.65 (95% CI, 0.12 to 0.86), five studies reporting on meningitis showed a protective effect of 0.64 (95% CI, 0.30 to 0.82), and three studies of disseminated tuberculosis showed a protective effect of 0.78 (95% CI, 0.58 to 0.88). abstract, DATA SYNTHESIS Meningitis in the same sentence: 0.64 (0.30-0.82) from five studies.
Meningeal and miliary tuberculosis severity: meningitis or disseminated disease case definition: Meningeal or miliary TB as defined in each randomized trial	infants country: multiple NeonateInfant <1y1-4y	BCG, various strains platform: live attenuated M. bovis doses: 1 schedule: in infancy vs unvaccinated or placebo	0.1 (.01–.77) rate ratio (vaccinated vs unvaccinated)	meta-analysis randomized trials reporting miliary or meningeal TB (6)	Mangtani et al. 2014 Clin Infect Dis quote Protection against meningeal and miliary tuberculosis was also high in infants (RR, 0.1; 95% CI, .01-.77) and children stringently tuberculin tested (RR, 0.08; 95% CI, .03-.25). abstract, RESULTS Few events, wide interval. Same sentence: RR 0.08 (.03-.25) in children stringently tuberculin tested (next record).
Meningeal and miliary tuberculosis severity: meningitis or disseminated disease case definition: Meningeal or miliary TB as defined in each randomized trial	children stringently tuberculin tested before vaccination prior infection: tuberculin-negative after stringent testing country: multiple 5-11y12-17y	BCG, various strains platform: live attenuated M. bovis doses: 1 vs unvaccinated or placebo	0.08 (.03–.25) rate ratio (vaccinated vs unvaccinated)	meta-analysis randomized trials reporting miliary or meningeal TB	Mangtani et al. 2014 Clin Infect Dis quote Protection against meningeal and miliary tuberculosis was also high in infants (RR, 0.1; 95% CI, .01-.77) and children stringently tuberculin tested (RR, 0.08; 95% CI, .03-.25). abstract, RESULTS Protection against the severe disseminated forms is consistently high across reviews (compare Colditz 0.78 and the MMWR summary of 86%).

Gaps

The WHO BCG position paper (WER 2018, 93:73-96) could not be snapshotted: the IRIS bitstream and apps.who.int PDF links return a 6-character DSpace stub and the cdn.who.int summary PDF is 404. Its headline efficacy and HIV-infant recommendations are therefore not quoted; the publication landing page (who-int-0d52b23290) was cached but carries only the abstract.
No quantitative record for the risk of TB meningitis or miliary TB per primary infection by age (the pre-chemotherapy natural history in Marais 2004, PMID 15141729, states that infection under 2 years frequently progresses to serious disease but gives no figures in the abstract; the full text was not snapshotted).
No TB meningitis incidence per 100,000 children for a high-income, BCG-free setting (e.g. US or ECDC) was found in the searches run; the burden records are global modelled counts and hospital series.
No controlled estimate of post-TB-meningitis hearing loss or epilepsy as separate outcomes: the sequelae records pool all neurological sequelae.
No efficacy record for BCG in HIV-exposed uninfected infants or in HIV-infected children (searches returned immunogenicity and safety studies only).
No efficacy estimate by BCG strain was filed: both Mangtani 2014 and Abubakar 2013 report no association between strain and protection, but give no strain-specific numbers in the abstract.
No safety record from a controlled design (SCCS/cohort) for BCG lymphadenitis or osteitis; the two safety records are surveillance rates and are flagged.
Household-contact IPD estimates of BCG effectiveness by age at exposure (Martinez et al.) were searched for but the target paper did not surface; only the 2020 Lancet exposure-risk IPD (filed as burden) and the 2025 Lancet Microbe infection IPD were found.

Flags for review

WHO Global TB Report 2024 figures (bcg-bur-001 to 008) carry 95% uncertainty intervals from modelling, stored in ci_low/ci_high; they are not confidence intervals. WHO states each annual report supersedes previous editions and comparisons across editions are not appropriate.
The treated-child CFR (0.9%, bcg-bur-011), the pre-treatment-era child CFR (21.9%, bcg-bur-014) and the WHO population CFR (11.5%, bcg-bur-006) have different denominators (diagnosed-and-mostly-treated children, pre-1946 series, all modelled incident cases). Present them side by side; never derive one from another.
The TB meningitis sequelae figure 53.9% (bcg-pa-001) is per survivor and the death figure 19.3% (bcg-bur-026) is per treated case; the two must not be added. The survival-without-sequelae figure (36.7%, bcg-pa-002) is the correct per-case composite.
Efficacy records bcg-eff-029 to 033 are NON-SPECIFIC (heterologous) effects on all-cause mortality, not protection against tuberculosis. They are filed under efficacy only because the outcome is all-cause mortality from randomised trials, as the brief instructs. Three of the five come from one research group in Guinea-Bissau (open-label trials); the Higgins/WHO review trial estimate and the Trunk meta-analysis of longer follow-up both include 1.
Mangtani 2014 reports CI bounds without a leading zero (e.g. ".18-.37"); the spec copies them as written so the number check passes. Renderers should display them as 0.18-0.37.
Colditz 1995 (bcg-eff-004 to 007) reports protective effect as a proportion (0.74 = 74%), whereas Colditz 1994 (bcg-eff-001 to 003) reports RR/OR. The measure field states which; do not mix the two scales when charting.
Safety records bcg-saf-001 and bcg-saf-002 are uncontrolled surveillance rates (population-based incidence and passive reports), included because they underpin the HIV contraindication; they do not meet the brief's controlled-source bar and should be reviewed before publication.
The two 60-year and 20-year duration records (bcg-eff-017, bcg-eff-021) come from retrospective record review and a scar-based observational comparison within a trial arm respectively; they conflict with the Abubakar 2013 conclusion that protection is well demonstrated only to 10 years. Both views are filed.
Chingleput (bcg-eff-015, bcg-eff-016) is a complete null in adults and near-null in children in a tropical high-transmission population, while the UK MRC trial (bcg-eff-014, 018-020) shows about 80% protection. This is the central heterogeneity of BCG and should be shown rather than averaged.
The WHO fact sheet (who-int-64be06f908, cached 2026) reports 2024 estimates (1.23 million deaths, 10.7 million cases) that post-date the 2024 Global TB Report (2023 data) used for the other WHO records; both years are labelled in period=.
bcg-bur-009 quotes the WHO fact sheet; the compiler's sentence splitter prefixed the key-facts sentence with a photo caption from the page chrome. The number (1.23 million deaths in 2024) is correct and verbatim; the leading caption text is page noise, not evidence.

Search log (32 searches)

BCG vaccine efficacy tuberculosis meta-analysis Colditz · PubMed → Found PMID 8309034 (Colditz 1994 JAMA) and 7596718 (Colditz 1995 Pediatrics, infants); compiled as bcg-eff-001 to 007. Also 7795103 (Wilson, latitude; no effect estimate in abstract, not filed).
Mangtani BCG protection against tuberculosis systematic review latitude · PubMed → Found PMID 24021245 (Abubakar 2013 HTA duration review); compiled as bcg-eff-014, 015, 022.
BCG vaccine tuberculous meningitis miliary tuberculosis children meta-analysis Trunz · PubMed → Found PMID 16616560 (Trunz 2006 Lancet); abstract reports cases prevented per vaccination, not an efficacy ratio, so used only as context (not filed as a record).
Mangtani 2014 protection by BCG vaccine against tuberculosis systematic review of randomized controlled trials · PubMed → Found PMID 24336911; compiled as bcg-eff-008 to 013.
BCG vaccination duration of protection Alaska 60 years Aronson · PubMed → Found PMID 15126436; compiled as bcg-eff-017 and bcg-bur-022.
Chingleput BCG trial fifteen year follow up · PubMed → Found PMID 10573656; compiled as bcg-eff-016 and bcg-bur-021.
BCG non-specific effects all-cause mortality randomised trial Guinea-Bissau low birth weight · PubMed → Returned DTP/OPV papers only; refined below.
Aaby BCG low birth weight neonatal mortality randomised trial Guinea-Bissau · PubMed → Found PMIDs 21673035, 29579158, 29216358, 20215360; compiled as bcg-eff-029 to 031.
tuberculous meningitis children treatment outcomes systematic review meta-analysis mortality neurological sequelae · PubMed → Found PMID 25108337 (Chiang 2014); compiled as bcg-bur-026, bcg-pa-001, bcg-pa-002.
Dodd global burden of childhood tuberculosis mortality mathematical modelling · PubMed → Found PMID 39706662 (du Preez 2025 TBM burden); compiled as bcg-bur-023 to 025 and bcg-pa-003.
Dodd global burden tuberculosis mortality children 2015 mathematical modelling Lancet Global Health · PubMed → Found PMID 28807188; compiled as bcg-bur-017, 018.
Dodd burden of childhood tuberculosis estimates 22 high burden countries mathematical modelling · PubMed → Found PMID 25103518; compiled as bcg-bur-019.
tuberculosis case fatality untreated natural history systematic review · PubMed → Found PMID 21483732 (Tiemersma 2011); compiled as bcg-bur-012, 013.
tuberculosis case fatality ratio treated children global estimates · PubMed → Found PMID 27964822 (Jenkins 2017); compiled as bcg-bur-011, 014, 015. Also 33640076 (TB survivors) compiled as bcg-pa-010.
BCG revaccination adolescents randomized trial Brazil efficacy · PubMed → Found PMIDs 16214599, 21616115, 16229231; compiled as bcg-eff-021, 023 to 025.
BCG vaccine effectiveness infants school age United Kingdom MRC trial · PubMed → Found PMID 39709975 (Pelzer 2025 IPD); compiled as bcg-eff-028.
Medical Research Council BCG vole bacillus vaccines trial adolescents · PubMed → Found PMIDs 326347 and 4537855; compiled as bcg-eff-018 to 020 and bcg-bur-020.
Roy BCG effect on Mycobacterium tuberculosis infection children systematic review BMJ · PubMed → Found PMID 25097193; compiled as bcg-eff-026, 027.
Martinez infant BCG vaccination household contacts individual participant data meta-analysis · PubMed → Found PMID 32199484 (2020 Lancet exposure IPD); compiled as bcg-bur-016. Target BCG-by-age IPD paper not found (see gaps).
miliary tuberculosis children case fatality · PubMed → Found PMID 1749696 (Hussey 1991); compiled as bcg-bur-027, 028.
tuberculous meningitis children long-term neurodevelopmental outcome · PubMed → Found PMIDs 26888419 and 36525135; compiled as bcg-bur-029, bcg-pa-004, bcg-pa-005.
BCG vaccination non-specific effects mortality systematic review · PubMed → Found PMIDs 36679966 (Trunk 2023) and 27737834 (Higgins 2016); compiled as bcg-eff-032, 033.
disseminated BCG disease HIV-infected infants South Africa incidence · PubMed → Found PMIDs 19649364 (Hesseling 2009) and 19861003 (Azzopardi 2009 review); compiled as bcg-saf-001.
post-tuberculosis lung disease prevalence systematic review meta-analysis survivors · PubMed → Found PMID 37076175 (Ivanova 2023); compiled as bcg-pa-008, 009.
tuberculosis survivors mortality after treatment completion systematic review meta-analysis · PubMed → Found PMID 31324519 (Romanowski 2019); compiled as bcg-pa-006, 007.
Marais natural history childhood intra-thoracic tuberculosis pre-chemotherapy era · PubMed → Found PMID 15141729; abstract has no figures, not filed (see gaps).
BCG vaccination protection against tuberculosis in HIV-exposed infants · PubMed → Only immunogenicity/safety trials returned; no efficacy estimate (see gaps).
BCG vaccination tuberculosis children case-control effectiveness England 1990s · PubMed → No relevant case-control study returned; a preterm/LBW BCG safety meta-analysis (30476973) surfaced but was not fetched.
childhood tuberculosis hospitalisation outcomes cohort South Africa infants · PubMed → No usable hospitalisation-rate study surfaced.
WHO Global tuberculosis report 2024 landing page, section 1.1 TB incidence, section 1.2 TB mortality, TB fact sheet · websource → Cached who-int-142037f5b4 (landing, no figures), who-int-39a13a321d (incidence), who-int-d1cfc27960 (mortality), who-int-64be06f908 (fact sheet); compiled as bcg-bur-001 to 010.
WHO BCG position paper February 2018 (IRIS bitstream, apps.who.int, cdn.who.int, publication landing page) · websource → PDF unreachable (DSpace stub / 404); landing page cached as who-int-0d52b23290 with abstract only. See gaps and flags.
CDC MMWR 1996 Role of BCG vaccine in the prevention and control of tuberculosis in the United States (ACET/ACIP) · websource → Cached cdc-gov-1af285e85a; compiled as bcg-saf-002. CDC BCG clinician page cached as cdc-gov-f1efca4863 and TB clinical overview as cdc-gov-b1668c50b4 (no numeric estimates used).