WHO's headline indicator for drug-resistant tuberculosis is 'treatment success rate' -- the share of enrolled patients who are cured -- and in 2024 that number improved to 71%, up from 68%.

What that indicator does not report, anywhere in the standard fact sheet, is that only 42% of the estimated 390,000 people who developed multidrug- or rifampicin-resistant TB that year were ever enrolled in treatment at all. Multiply the two figures and the true population-level cure fraction is roughly 30%, not 71%.

This is not a WHO-specific failure -- it's a structural feature of how most global health programs report outcomes: success rate among the treated, not success rate among the sick. The gap matters because it changes the policy conversation entirely. A 71% success rate says 'the treatment works, keep funding regimen R&D.'

A 30% end-to-end cure fraction says 'the treatment works but almost nobody who needs it is getting it, so the marginal dollar belongs in case-finding and formulary access, not in another clinical trial.'

Countries and donors currently have no standardized, mandatorily-reported indicator that combines incidence, diagnostic yield, treatment initiation, and treatment success into one number. Each piece is measured by a different part of the health system, on a different reporting cycle, and nobody is institutionally responsible for multiplying them together.

The result is that genuinely excellent treatment science (BPaLM cuts DR-TB therapy from ~18 months to 6, all oral) gets celebrated in isolation from the fact that formulary and procurement barriers mean only 3 of 18 surveyed European countries had full pretomanid availability as of the most recent survey.

The same blind spot almost certainly exists in HIV, hepatitis C, and cervical cancer screening programs, anywhere a 'success among the treated' metric is reported without its companion 'share of the sick who got treated' metric sitting next to it on the same dashboard.