Table 2.

Outcome data in newborn babies weighing <2000 g assigned to receive BCG Danish and OPV within 48 hours of admission (intervention) or delayed at least to discharge (control)

	Mortality rate (deaths/total person years); No		Adjusted hazard ratio (95% CI)	P value*
	Early BCG-OPV	Control
Deaths ≤28 days (primary outcome)	1.29 (238/184); 2714	1.50 (273/182); 2706	0.83 (0.69 to 0.98)	0.03
Deaths from infection	0.40 (73/184); 2707	0.73 (133/182); 2702	0.53 (0.40 to 0.70)	<0.001†
Deaths from non-infectious cause	0.89 (164/184); 2707	0.77 (139/182); 2702	1.10 (0.88 to 1.39)	0.41†
Deaths in hospital	2.33 (237/102); 2714	2.69 (269/100); 2706	0.83 (0.70 to 0.99)	0.04†
Deaths ≤28 days; controls censored when BCG-OPV given ≤28 days (original primary outcome)	1.29 (238/184); 2714	2.25 (273/121); 2706	0.78 (0.65 to 0.93)	0.006‡
Per protocol analysis	1.28 (236/184); 2702	2.26 (275/122); 2709	0.77 (0.65 to 0.92)	0.004§
Time of death after enrolment (days):
0-2	2.53 (56/22.1); 2714	3.41 (75/22.0); 2706	0.72 (0.51 to 1.03)	0.07¶
3-6	3.48 (99/28.4); 2658	4.13 (116/28.1); 2631	0.84 (0.64 to 1.10)	0.21¶
7-27	0.62 (83/134); 2553	0.62 (82/132); 2512	0.91 (0.67 to 1.24)	0.55¶

CI=confidence interval; OPV=oral polio vaccine.

^*Cox regression stratified by neonatal intensive care unit and adjusted for age, birth weight, gestational age, and sex. Six babies in the early BCG-OPV group and three in the control group were lost to follow-up before 28 days of age and censored from the time of last contact.

^†Secondary outcome not adjusted for multiplicity.

^‡Biased estimate because low risk time from discharge to 28 days was removed from vaccinated controls.

^§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.

^¶Prespecified descriptive outcome not adjusted for multiplicity.