Modelling the cost-effectiveness of a new infant vaccine to prevent tuberculosis disease in children in South Africa

Background

Tuberculosis remains the leading cause of death in South Africa. A number of potential new TB vaccine candidates have been identified and are currently in clinical trials. One such candidate is MVA85A. This study aimed to estimate the cost-effectiveness of adding the MVA85A vaccine as a booster to the BCG vaccine in children from the perspective of the South African government.

Methods

The cost-effectiveness was assessed by employing Decision Analytic Modelling, through the use of a Markov model. The model compared the existing strategy of BCG vaccination to a new strategy in which infants receive BCG and a booster vaccine, MVA85A, at 4 months of age. The costs and outcomes of the two strategies are estimated through modelling the vaccination of a hypothetical cohort of newborns and following them from birth through to 10 years of age, employing 6-monthly cycles.

Results

The results of the cost-effectiveness analysis indicate that the MVA85A strategy is both more costly and more effective – there are fewer TB cases and deaths from TB than BCG alone. The South African government would need to spend an additional USD 1,105 for every additional TB case averted and USD 284,017 for every additional TB death averted. The threshold analysis shows that, if the efficacy of the MVA85A vaccine was 41.3% (instead of the current efficacy of 17.3%), the two strategies would have the same cost but more cases of TB and more deaths from TB would be prevented by adding the MVA85A vaccine to the BCG vaccine. In this case, the government chould consider the MVA85A strategy.

Conclusions

At the current level of efficacy, the MVA85A vaccine is neither effective nor cost-effective and, therefore, not a good use of limited resources. Nevertheless, this study contributes to developing a standardized Markov model, which could be used, in the future, to estimate the potential cost-effectiveness of new TB vaccines compared to the BCG vaccine, in children between the ages of 0–10 years. It also provides an indicative threshold of vaccine efficacy, which could guide future development.

Antiretroviral therapy and early mortality in South Africa

OBJECTIVE

To describe province-wide outcomes and temporal trends of the Western Cape Province antiretroviral treatment (ART) programme 5 years since inception, and to demonstrate the utility of the WHO monitoring system for ART.

METHODS

The treatment programme started in 2001 through innovator sites. Rapid scaling-up of ART provision began early in 2004, located predominantly in primary-care facilities. Data on patients starting ART were prospectively captured into facility-based registers, from which monthly cross-sectional activity and quarterly cohort reports were aggregated. Retention in care, mortality, loss to follow-up and laboratory outcomes were calculated at 6-monthly durations on ART.

FINDINGS

By the end of March 2006, 16 234 patients were in care. The cohort analysis included 12 587 adults and 1709 children. Women accounted for 70% of adults enrolled. After 4 and 3 years on ART respectively, 72.0% of adults (95% confidence interval, CI: 68.0-75.6) and 81.5% (95% CI: 75.7-86.1) of children remained in care. The percentage of adults starting ART with CD4 counts less than 50 cells/microl fell from 51.3% in 2001 to 21.5% in 2005, while mortality at 6 months fell from 12.7% to 6.6%, offset in part by an increase in loss to follow-up (reaching 4.7% at 6 months in 2005). Over 85% of adults tested had viral loads below 400 copies/ml at 6-monthly durations until 4 years on ART.

CONCLUSION

The location of care in primary-care sites in this programme was associated with good retention in care, while the scaling-up of ART provision was associated with reduced early mortality.