Cost-effectiveness of alternative strategies for tuberculosis screening before kindergarten entry

OBJECTIVE. We undertook a decision analysis to evaluate the economic and health effects and incremental cost-effectiveness of using targeted tuberculin skin testing, compared with universal screening or no screening, before kindergarten. METHODS. We constructed a decision tree to determine the costs and clinical outcomes of using targeted testing compared with universal screening or no screening. Baseline estimates for input parameters were taken from the medical literature and from California health jurisdiction data. Sensitivity analyses were performed to determine plausible ranges of associated outcomes and costs. We surveyed California health jurisdictions to determine the prevalence of mandatory universal tuberculin skin testing. RESULTS. In our base-case scenario, the cost to prevent an additional case of tuberculosis by using targeted testing, compared with no screening, was $524 897. The cost to prevent an additional case by using universal screening, compared with targeted testing, was $671 398. The incremental cost of preventing a case through screening remained above $100 000 unless the prevalence of tuberculin skin testing positivity increased to > 10%. More than 51% of children entering kindergarten in California live where tuberculin skin testing is mandatory. CONCLUSIONS. The cost to prevent a case of tuberculosis by using either universal screening or targeted testing of kindergarteners is high. If targeted testing replaced universal tuberculin skin testing in California, then $1.27 million savings per year would be generated for more cost-effective strategies to prevent tuberculosis. Improving the positive predictive value of the risk factor tool or applying it to groups with higher prevalence of latent tuberculosis would make its use more cost-effective. Universal tuberculin skin testing should be discontinued, and targeted testing should be considered only when the prevalence of risk factor positivity and the prevalence of tuberculin skin testing positivity among risk factor-positive individuals are high enough to meet acceptable thresholds for cost-effectiveness.