Background: Around 0.4 million cases of multidrug-resistant tuberculosis (MDR-TB) occur each year. Only a small fraction of these cases are treated according to international guidelines. Evidence relevant to decisions about whether to scale-up treatment for MDR-TB includes cost and cost-effectiveness data. Up to 2010, no systematic review of this evidence has been available.
Objective: Our objective was to conduct a systematic review of the cost and cost effectiveness of treatment for MDR-TB and synthesize the available data.
Methods: We searched for papers published or prepared for publication in peer-review journals and grey literature using search terms in five languages: English, French, Portuguese, Russian and Spanish. From an initial set of 420 studies, four were included, from Peru, the Philippines, Estonia and Tomsk Oblast in the Russian Federation. Results on costs, effectiveness and cost effectiveness were extracted. Assessment of the quality of each economic evaluation was guided by two existing checklists around which there is broad consensus. Costs were adjusted to a common year of value (2005) to remove distortions caused by inflation, and calculated in two common currencies: $US and international dollars (I$), to standardize for purchasing power parity. Data from the four identified studies were then synthesized using probabilistic sensitivity analysis, to appraise the likely cost and cost effectiveness of MDR-TB treatment in other settings, relative to WHO benchmarks for assessing whether or not an intervention is cost effective. Best estimates are provided as means, with 5th and 95th percentiles of the distributions.
Results: The cost per patient for MDR-TB treatment in Estonia, Peru, the Philippines and Tomsk was $US10 880, $US2423, $US3613 and $US14 657, respectively. Best estimates of the cost per disability-adjusted life-year (DALY) averted were $US598 (I$960), $US163 (I$291), $US143 (I$255) and $US745 (I$1059), respectively. The main influences on costs were (i) the model of care chosen (the extent to which hospitalization or ambulatory care were relied upon) and (ii) the second-line drugs included in the treatment regimen. When extrapolated to other settings, the best estimate of the cost of treatment varied from US3401 to US195 078, depending on the region and model of care. The cost per DALY averted was lower than GDP per capita in all 14 WHO sub-regions considered, with better cost effectiveness for outpatient versus inpatient models of care.
Conclusions: Treatment for MDR-TB can be cost effective in low- and middle-income countries. Evidence about the relative cost effectiveness of outpatient versus inpatient models of care is limited and more data are needed from Africa and Asia--especially India and China, which have the largest number of cases. Unless there is strong evidence that hospitalization is necessary to achieve high rates of adherence to treatment, patients with MDR-TB should be treated using mainly ambulatory care.