Decentralisation of multidrug-resistant-tuberculosis care and management.

Multidrug-resistant (MDR) tuberculosis is an important medical and public health challenge, affl icting an estimated 500 000 new patients each year. Recent progress in the development of new molecular diagnostic techniques and the fi rst new antituberculosis drug registration in almost 40 years provide reasons for optimism, but still, globally, less than 10% of individuals with MDR-tuberculosis receive treatment of known quality. Country data for detection and enrolment onto treatment reported to WHO over the past 3 years paint a varied picture of progress. Some high burden countries report steady increases in the numbers of patients treated over the past 3 years, notably India, Russia, and South Africa. However, less than a quarter of the countries providing complete data showed a linear increase in case detection, and less than half showed a year-on-year increase in the numbers of patients enrolled onto treatment. No country that treated more than 500 patients achieved a treatment success rate of greater than 55%. There has been progress in places, but not nearly enough. Impediments towards universal access to MDRtuberculosis treatment are many, ranging from insuffi cient donor funding to poor laboratory diagnostic capacity and health system challenges, including the need to encourage ambulatory care models. This last point is crucial, because although advances in diagnostics and improved treatment regimens will go a long way to improving access, the models of care in which such technologies are delivered will strongly aff ect patient access and retention in care. Historically, MDR-tuberculosis treatment has been provided through small, individualised programmes with specialist clinical support. Such models of care might work where patient numbers are small, but in settings like South Africa and Russia, which both detected more than 10 000 cases in 2011 alone, specialised, centralised programmes are unlikely to meet the need. Moreover, larger MDR-tuberculosis treatment programmes are associated with poorer rates of retention in care. Increased eff orts in global tuberculosis control have resulted in strong basic tuberculosis programmes in many settings that have developed local systems for supporting patients on treatment for tuberculosis and ensuring adherence. Integration of MDR-tuberculosis care and management into routine tuberculosis programmes in such settings is feasible and allows patients with MDR-resistant tuberculosis to receive treatment close to where they live. There is increasing evidence from several large-scale MDR-tuberculosis programmes of improved access to care and management through decentralisation, without compromising treatment outcomes, including programmes that have task shifted initiation of MDR-tuberculosis treatment to trained nurses. Decentralisation has been a crucial strategy for expanding access to treatment for HIV and has been associated with better patient outcomes than with hospital-managed care, mainly due to improved retention; there is a broad consensus that to improve early health-seeking behaviour, promote adherence to medication, and minimise defaulting, HIV care is best provided as close as possible to the patient’s home and community. These lessons for patient support are clearly applicable to MDR-tuberculosis because default rates from care commonly exceed 20%. One concern about decentralising MDR-tuberculosis care is the potential increased risk of community transmission, but in most high-burden settings, the MDR-tuberculosis epidemic is driven primarily by direct transmission of MDR-tuberculosis strains. The low proportion of estimated cases that are treated appropriately means that most transmission is occurring among undiagnosed and untreated cases. Therefore, eff orts to reduce transmission should be directed at diagnosing and treating as many cases as possible and as early as possible. After decades of neglect, much needed improvements in diagnostics and therapeutics are becoming available. However, their impact will be small without parallel eff orts to improve the service delivery model. Decentralisation is likely to off er the best opportunity for early initiation of treatment at the scale required, Published Online June 4, 2013 http://dx.doi.org/10.1016/ S1473-3099(13)70151-8