Better treatment of XDR tuberculosis needed in South Africa - Author's reply.

Elize Pietersen and colleagues (April 5, p 1230) show the very poor outcomes for treatment of extensively drug-resistant (XDR) tuberculosis in South Africa. The study suggests that large numbers of patients, for whom treatment was ineffective, pose a risk of ongoing community transmission. Although we support the provision of community-based care minimising transmission risk after treatment failure, we would also like to put these numbers into context. More than 1500 XDR tuberculosis cases are diagnosed in South Africa annually, likely representing a fraction of the actual burden. In view of that less than half of people diagnosed receive treatment, the combination of undiagnosed and untreated XDR-tuberculosis poses a much larger risk in our communities than the small number of patients discharged with treatment failure Pietersen and colleagues describe. Rather than emphasising the risk from these few patients, the focus should be directed towards early diagnosis and rapid initiation of effective treatment for all drug-resistant tuberculosis cases, including individuals unfortunate enough to have contracted XDR-tuberculosis. South African guidelines describe individualised X D Rt u b e r c u l o s i s t r e a t m e n t , including the use of linezolid and clofazamine. Unfortunately, access to both these drugs is restricted; clofazamine due to supply issues and linezolid due to high costs charged by Pfizer in the absence of a registered generic alternative. Early XDR-tuberculosis diagnosis, treatment with more drugs (including linezolid and new drugs via expanded access), and particularly timely initiation of antiretroviral HIV Authors’ reply We thank Aurélien Dupré and colleagues for their interest in our Article. So far, most research into adhesions has focused on the consequences and prevention of small bowel obstruction, which is shown by the results presented in our meta-analysis. However, there is growing evidence suggesting that difficulties and subsequent iatrogenic injuries in reoperations are an even larger health problem in terms of morbidity and costs. Dupré and colleagues justly remark that the burden of adhesiolysis is only expected to rise further with the increasing number of reoperations in oncological surgery. Although somewhat outside the scope of our systematic review, these new epidemiological data warrant a change in the understanding of adhesion-related morbidity and strategies for adhesion prevention. Previous cost-effectiveness models for adhesion barriers that focused on prevention of adhesive small bowel obstruction demonstrated that these agents might be cost eff ective for only selected patients. A more complete cost-effectiveness model would also account for the additional costs made in reoperations and fertility and chronic visceral pain treatments. A more comprehensive model is expected to show that barriers are cost effective in most patients who undergo abdominal surgery because reduction of adhesion formation would already provide a benefi t. For explanation, the difficulty in prevention of adhesive small bowel 1 ten Broek RPG, Stommel MWJ, Strik C, van Laarhoven CJHM, Keus F, van Goor H. Benefi ts and harms of adhesion barriers for abdominal surgery: a systematic review and meta-analysis. Lancet 2014; 383: 48–59. 2 ten Broek RP, Strik C, Issa Y, Bleichrodt RP, van Goor H. Adhesiolysis-related morbidity in abdominal surgery. Ann Surg 2013; 258: 98–106. 3 ten Broek RP, Schreinemacher MH, Jilesen AP, Bouvy N, Bleichrodt RP, van Goor H. Enterotomy risk in abdominal wall repair: a prospective study. Ann Surg 2012; 256: 280–87. obstruction is that small bowel obstruction can be caused by just one adhesive band and prevention of small bowel obstruction therefore requires total adhesion prevention in the whole peritoneal cavity, which is relatively difficult to achieve. The complications of adhesiolysis during reoperation are correlated to the extent and severity of adhesion formation. Thus, although the use of adhesion reducing agents might not completely prevent adhesion formation in operations with extensive peritoneal damage and have only a modest effect on the incidence of small bowel obstruction, reducing the extent and severity of adhesions is likely to have a benefi cial effect on the outcomes of future operations. These new data have consequences for the design of future trials on adhesion prevention. Dupré and colleagues were one of the first to study the eff ect of adhesion barriers on adhesiolysis time in two-stage oncological liver surgery. Whether this reduction of adhesiolysis time is indeed correlated to a reduction in iatrogenic bowel injury and serious adverse events of the reoperation needs validation in larger trials.