Vaccinating small newborns on day one cut neonatal deaths 17% in an Indian trial

A vaccine that works against a disease the babies did not have is a difficult result to write about honestly. The BLOW2 trial, published in The BMJ on 22 September, reports that giving BCG and oral polio vaccine to newborns weighing under 2,000 g within 48 hours of admission reduced all-cause neonatal mortality — and that the reduction came from fewer deaths due to infections other than tuberculosis [s1].

No deaths from tuberculosis occurred in either group [s1].

What was tested

The trial was multicentre, open-label and randomised, conducted in three tertiary neonatal intensive care units in southeast India [s1]. Of 7,067 newborns assessed for eligibility, 5,420 were randomised — 2,714 to early vaccination and 2,706 to control [s1].

Babies in the early group received 0.1 mL of BCG Danish intradermally plus oral polio vaccine within 48 hours of NICU admission; the control group had the same vaccines delayed until at least the time of discharge [s1]. Randomisation was stratified by NICU, sex, and birth weight in three bands: under 1,000 g, 1,000–1,499 g, and 1,500–1,999 g [s1].

This is a timing trial, not a vaccination-versus-no-vaccination trial. Every baby was intended to be vaccinated; the question was whether doing it on day one rather than at discharge changes survival.

Median age at randomisation was 0.9 days and median birth weight 1,560 g [s1]. Nine babies were lost to follow-up [s1].

The result

Deaths occurred in 238 of 2,714 (8.8%) in the early vaccination group and 273 of 2,706 (10.1%) in the control group [s1]. In an 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 group and 1.50 in the control group — an adjusted hazard ratio of 0.83 (95% CI, 0.69 to 0.98; P=0.03) [s1].

The number needed to treat to prevent one death was 21, with a 95% confidence interval running from 10 to 245 [s1]. That interval is the most important number in the paper and the one least likely to be quoted. The point estimate says one death averted for every 21 babies vaccinated early; the data are also compatible with one death averted for every 245.

The secondary outcome — added a third of the way through the trial — was mortality due to infection [s1]. Infection-related neonatal mortality per person-year was 0.40 in the early group and 0.73 in the control group, an adjusted hazard ratio of 0.53 (95% CI, 0.40 to 0.70) [s1]. No serious adverse effects were associated with vaccination [s1].

The off-target effect problem

The trial's authors describe the finding as a non-specific, or off-target, effect: a reduction in deaths from infections other than tuberculosis [s1]. This is a contested area of vaccinology with a long observational literature and a much thinner randomised one.

The most directly relevant prior randomised evidence comes from Guinea-Bissau, where low-birth-weight children are not given BCG at birth. A trial recruiting 2,320 low-birth-weight children between 2004 and 2008 randomised them to early BCG versus delayed BCG [s2]. Infant mortality was reduced insignificantly by 17% in the primary analysis (mortality rate ratio 0.83; 95% CI, 0.63 to 1.08) [s2]. In secondary analyses, early BCG was associated with a mortality rate ratio of 0.49 (95% CI, 0.21 to 1.15) after three days and 0.55 (95% CI, 0.34 to 0.89) after four weeks, with the reduction attributed mainly to fewer cases of neonatal sepsis, respiratory infection and fever [s2]. The effect on infant mortality was most marked in children weighing under 1.5 kg (mortality rate ratio 0.43; 95% CI, 0.21 to 0.85) [s2].

The Guinea-Bissau trial's authors were careful about what they had shown: early BCG did not reduce infant mortality significantly, but might have a beneficial effect in the neonatal period [s2].

BLOW2 is a larger trial in a different setting reaching statistical significance on all-cause neonatal mortality where the earlier one did not — and reproducing the earlier trial's key secondary signal, that the benefit sits in infections rather than tuberculosis. Two randomised trials pointing the same way is a stronger position than one. It is not the same as a settled mechanism.

What is not established

Two components were given together. BLOW2 cannot attribute the effect to BCG, to oral polio vaccine, or to their combination [s1].

The trial was open-label [s1]. Clinicians knew which babies had been vaccinated, and in a NICU population where care decisions are continuous and discretionary, that is a live source of bias for an all-cause mortality endpoint.

The setting is specific: three tertiary NICUs in southeast India, with a control group vaccinated at discharge rather than never [s1]. The result speaks to timing within a system that already vaccinates, not to adding vaccines where there are none.

The authors' own framing of the implication is conditional: a substantial reduction in neonatal mortality could be achieved if a skilled administrator vaccinated a high proportion of newborns in high-mortality settings on the day of, or soon after, birth [s1]. "Skilled administrator" is doing work in that sentence — intradermal BCG in a 1,000 g infant is not a trivial procedure.

The trial is registered with the Clinical Trials Registry India as CTRI/2017/01/007676 [s1].

This article is informational and is not medical advice.

Sources

• [s1] "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)," BMJ, 22 September 2025. https://doi.org/10.1136/bmj-2025-084745
• [s2] "Randomized trial of BCG vaccination at birth to low-birth-weight children: beneficial nonspecific effects in the neonatal period?", The Journal of Infectious Diseases, July 2011. https://doi.org/10.1093/infdis/jir240