Directly Observed Therapy for TB

throughout developing countries. There is a paucity of literature on the complications of medical procedures in developing countries. Most hospital infection outbreaks within underdeveloped countries generally are not reported or investigated. There is a shortage of trained personnel, records tend to be poor, and there is little incentive for this kind of investigation by hospital administrators. To our knowledge, this investigation is the first of this kind done in hospital A. It was the unusual commitment to quality care from the physician in charge and a willingness to address weaknesses in this system that made this investigation possible. Despite the difficulty in collecting this information, the information on complication rates for various procedures is crucial for appropriate decision making. Even within developed countries, patient selection for discectomy remains varied and controversial. For example, the incidence of discectomies in the United States ranges from 83 per 100,000 patients in the Western region compared to 143 per 100,000 in the South9 versus 36 per 100,000 in Finland,1 and even within the United States rates vary markedly. Discectomies are performed twice as commonly on the West Coast compared to the East Coast.10 In Washington State, there was a 15-fold difference in rates of discectomy among different counties,2 with a significant relationship to the level of insurance reimbursements. Indications for surgery that are developed in wealthy countries should not be accepted uncritically in more resource-constrained settings where complication rates may be higher. Clearly, at hospital A, in the presence of such high complication rates, operations should be performed only if the potential benefit exceeds the high risk of discitis. At a minimum, as part of informed consent, the complication rate should be discussed explicitly with potential laminectomy patients. It is unclear why women were more likely to develop postoperative discitis than men. Possible explanations include obesity among women that made the surgical procedure more difficult and prolonged or the surgical staff’s attempt to preserve the modesty of women, resulting in a less complete skin preparation than for male patients. Because we did not observe skin preparation, nor did we have a measure of height and weight, we cannot assess whether these factors were important. Surveillance of postoperative infections and feedback to care providers is possible even in the setting of severe resource constraints. Unless outcomes are monitored and critically considered, developing-country healthcare facilities risk wasting the scarce resources allocated to health.