Making wider use of the world's most widely used vaccine: Bacille Calmette-Guerin revaccination reconsidered

Abstract

Approximately 100 million newborn children receive Bacille Calmette-Guérin (BCG) annually, because vaccination is consistently protective against childhood tuberculous meningitis and miliary TB. By contrast, BCG efficacy against pulmonary TB in children and adults is highly variable, ranging from 0% to 80%, though it tends to be higher in individuals who have no detectable prior exposure to mycobacterial infections, as judged by the absence of delayed-type hypersensitivity response (a negative tuberculin skin test, TST). The duration of protection against pulmonary TB is also variable, but lasts about 10 years on average. These observations raise the possibility that BCG revaccination, following primary vaccination in infancy, could be efficacious among TST-negative adolescents as they move into adulthood, the period of highest risk for pulmonary disease. To inform continuing debate about revaccination, this paper assesses the effectiveness and cost-effectiveness of revaccinating adolescents in a setting with intense transmission-Cape Town, South Africa. For a cost of revaccination in the range US$1-10 per person, and vaccine efficacy between 10% and 80% with protection for 10 years, the incremental cost per year of healthy life recovered (disability-adjusted life years, DALY) in the vaccinated population lies between US$116 and US$9237. The intervention is about twice as cost-effective when allowing for the extra benefits of preventing transmission, with costs per DALY recovered in the range US$52-$4540. At 80% efficacy, revaccination averted 17% of cases. Under the scenarios investigated, BCG revaccination is cost-effective against international benchmarks, though not highly effective. Cost-effectiveness ratios would be more favourable if we also allow for TB cases averted by preventing transmission to HIV-positive people, for the protection of HIV-negative people who later acquire HIV infection, for the possible non-specific benefits of BCG, for the fact that some adolescents would receive BCG for the first time, and for cost sharing when BCG is integrated into an adolescent immunization programme. These findings suggest, subject to further evaluation, that BCG revaccination could be cost-effective in some settings.

Keywords: Bacille Calmette–Guérin; adolescents; cost-effectiveness; immunization; revaccination; tuberculosis.

Figure 1.

Distributions of tuberculin skin-test responses in children in (a) North India [30] and (b) Cape Town, South Africa [31]. The solid line in (a) is an estimate of the number of children infected with M. tuberculosis, as distinct from those showing cross reactions with non-tuberculous mycobacteria or primary BCG vaccination at smaller induration sizes.

Figure 2.

(a) The proportion of people (p) with a positive TST by age, linearized with the transformation –ln(1−p). (b) The number of HIV-negative TB cases reported from Cape Town in 2009 (black), the number estimated by fitting model (2.1)–(2.5) (white), and the number obtained when BCG revaccination has efficacy 0.6 (grey). Adapted from Wood et al. [32,33].

Figure 3.

The estimated numbers of new TB cases that arise from new infections, from reinfection and by reactivation in adolescents and adults, Cape Town, South Africa. The numbers are a breakdown of the white bars in figure 2b.

Figure 4.

The impact of revaccination measured in terms of (a) the proportion of cases averted in the cohort, calculated with (black) and without (white) the effect of reduced transmission, and (b) the cost per DALY regained (contours) in terms of the cost of revaccination per person (US$, horizontal axis) and vaccine efficacy (vertical axis), and allowing for the effects of reduced transmission. The darkest area also includes values in excess of US$1000 per DALY regained.