I read with great interest this article by Vassall and colleagues [1] describing the cost-effectiveness of rapid TB diagnosis with Xpert MTB/RIF in high-burden countries. It should be noted that the methods used can be extended to show that TB case-finding is, in general, highly cost-effective. In Table 1, the authors show that detection of a single case of (otherwise-undiagnosed) TB saves over 10 disability-adjusted life years (DALYs). Since the costs of treating false-positive cases are dwarfed by this benefit of treating undiagnosed cases [2], expending $300-$600 per additional case detected (e.g., 10-30 Xpert MTB/RIF tests, but also attainable by other means) yields highly-favorable cost-effectiveness ratios of $30-60/DALY averted. To generate cost-effectiveness ratios similar to per-capita GDP in low-income countries (e.g., $490 in Uganda), TB case-finding would need to be exceedingly inefficient (e.g., >$4900 spent per TB case detected). Thus, any intervention that increases the number of TB cases detected and treated is likely to be highly cost-effective by traditional metrics [2].

Thus, while the appropriately-stated outcome of this manuscript is that Xpert MTB/RIF is likely to be cost-effective, the true messages of this manuscript are twofold. First, as the authors state, we need pragmatic trials to determine which TB case-finding methods are most effective and efficient, in terms of reducing TB incidence and mortality. Of note, a recent trial in Zambia and South Africa [3] suggests that clinic-based case-finding does not reduce TB prevalence, even though more cases are detected. If scale-up of Xpert MTB/RIF does not lead to reduced TB incidence when implemented in the field (e.g., if cases are detected by other means, or if positive Xpert results do not lead to treatment), then the actual cost-effectiveness might be much less favorable than expected. The authors have valiantly attempted to account for such realities, but no modeling study can conclusively determine the “counterfactual” fate of undiagnosed TB cases. Second, and more important, this analysis calls attention to the urgent need for increased funding to TB control efforts. Despite offering generally much-better value, only $1 of global donor funding is spent for TB control for every $10 spent on HIV [4]. We should not speak of highly effective and cost-effective measures such as TB case-finding as “unaffordable,” when the reason for this lack of affordability is a lack of global commitment to TB control. What is unaffordable is our indifference – as demonstrated, in part, by funding commitments – in the fight against this global scourge.

References
1. Vassall A, van Kampen S, Sohn H, Michael JS, John KR, et al. (2011) Rapid diagnosis of tuberculosis with the xpert MTB/RIF assay in high burden countries: A cost-effectiveness analysis. PLoS Med 8(11): e1001120.
2. Dowdy DW, Cattamanchi A, Steingart KR, Pai M. (2011) Is scale-up worth it? challenges in economic analysis of diagnostic tests for tuberculosis. PLoS Medicine 8(7): e1001063.
3. Beyers N. Taking TB control beyond the clinic: Impact of household and community interventions in southern africa. 42nd Union World Conference on Lung Health Symposium 39 Lille, France, 2011 .
4. [Anonymous]. (2011) Global funding for infectious diseases: TB or not TB? Lancet 378(9801): 1439.