Efficacy of BCG vaccination of the newborn: evaluation by a follow-up study of contacts in Bangui

Abstract

Background: The efficacy of Bacillus of Calmette and Guérin (BCG) vaccination given at birth is still controversial. We therefore conducted a study in Bangui (Central African Republic) to estimate the protection afforded over the first seven years of life by BCG administered at birth.

Methods: One thousand children who had lived in contact with a recently diagnosed case of contagious tuberculosis were followed up for a period of 6 months in order to detect the occurrence of tuberculosis. Diagnosis of tuberculosis was made through a scoring system. Vaccine efficacy (VE) was calculated on the basis of the relative risk of contracting tuberculosis according to vaccination status.

Results: The efficacy of BCG was estimated to be 71% (95% confidence interval: 56-81%). This result remained practically the same after changing the definition used for tuberculosis cases (VE = 75% for a threshold with a score of 15 instead of 6, VE = 74% when only confirmed cases were considered). There was no difference between the two groups in the variables measuring intensity of contact with the source of contamination, but there was a difference in age distribution. Vaccine efficacy adjusted for this factor was the same as the crude VE.

Conclusion: This study, based on a methodology that controls for most of the risks of bias inherent to field efficacy measurement, confirms the protective capacity of neonatal BCG against childhood tuberculosis. Therefore BCG vaccination at birth must remain a public health priority especially in countries with high incidence of the disease.

PIP: The efficacy of Bacillus of Calmette and Guerin (BCG) vaccination given at birth is still controversial, therefore a study was conducted in Bangui, Central African Republic, to estimate the protection afforded over the first 7 years of life by BCG administered at birth. 1000 children who had lived in contact with a recently diagnosed case of contagious tuberculosis (TB) were followed up from May 1989 to February 1991 in order to detect the occurrence of TB. 896 of them were considered as vaccinated. Diagnosis of TB was made through a scoring system endorsed by the World Health Organization. Contact children with a score of or= 6 were considered to have TB. Four groups of children were formed in order to calculate the risk of TB in relation to vaccination status: children with TB who had been vaccinated, children with TB who had not been vaccinated, healthy children who had been vaccinated, and healthy children who had not been vaccinated. Vaccine efficacy (VE) was calculated on the basis of the relative risk of contracting TB according to vaccination status. Of the 1000 contact children, 91 had a score of or= 6. The TB incidence rate was 7.3% in vaccinated children and 25% in nonvaccinated children, which corresponded to a 0.29 relative risk of contracting TB or efficacy of BCG of 71% (95%, confidence interval: 56-81%). This result remained practically the same after changing the definition used for TB cases (VE = 75% for a threshold with a score of 15 instead of 6, VE = 74% when only confirmed cases were considered). There was no difference between the 2 groups in the variables measuring intensity of contact with the source of contamination, but there was a difference in age distribution. BCG vaccination at birth must remain a public health priority, especially in countries with high incidence of the disease in view of the protective capacity of neonatal BCG against childhood TB.