The global rise of extensively drug-resistant tuberculosis: is the time to bring back sanatoria now overdue?

Before eff ective treatment for tuberculosis became available thousands of people spent time in sanatoria (fi gure) in the hope of a cure through fresh air, sunlight, adequate nutrition, and micronutrients such as calcium. 1 Surgical techniques to promote part or complete lung collapse (eg, artifi cial pneumothorax, plombage, and thoracoplasty) were also used. 1 Although no controlled studies have established the eff ectiveness of these interventions, the existence of sanatoria led to long-term removal of infectious patients from the community. The need for sanatoria lessened when therapy led to cure rates in outpatient settings similar to those in inpatient settings and when disease burden fell because of improved living conditions and eff ective national tuberculosis programmes. 4 After three to four decades of rifampicin use, cases of multidrug resistant (MDR) and extensively drug-resistant (XDR) tuberculosis have appeared. Worldwide, 5–10% of the roughly half a million prevalent cases of MDR tuberculosis are extensively drug resistant. 5 XDR tuberculosis in particular is a grave threat since mortality rates are substantially higher than for other forms of the disease; 6–8 it greatly increases the costs of running a tuberculosis programme and can thus weaken well or modestly functioning national treatment programmes. In South Africa, for example, the annual mortality rate in patients with XDR tuberculosis in a non-outbreak setting approaches 40%; 6 despite treatment of about half a million drug-susceptible cases and less than 10 000 MDR and XDR cases every year, MDR and XDR disease uses up more than 50% of the annual tuberculosis drug budget. Since the revised defi nition of XDR tuberculosis and widespread publicity that followed the Tugela Ferry outbreak in 2006, 7 drugs such as capreomycin and para-aminosalicylic acid and extended drug susceptibility testing have become available in several developing countries including South Africa. In that country, line-probe assays (and the Gene Xpert MTB-RIF assay) are now also available within the national treatment programme. Thus, since late 2006 an estimated 2000–3000 people with XDR tuberculosis and an equal or higher number of people with MDR disease in whom treatment has failed, have been detected and treated with second-line drugs. Consequently, treatment-related outcome data have become available in programmes using second-line drugs such as capreomycin and para-aminosalicylic acid in settings of intensive inpatient therapy. 6,9–11 However, data from retrospective studies have shown that outcomes in high-burden settings such as South Africa for both MDR 12 and XDR tuberculosis …