concomitant cisplatin and radiotherapy on inoperable non-small-cell lung cancer. N Engl J Med 1992; 326: 524-30. 7 Furuse K Fukuoka M,Takada Y, Nishikawa H, Katagami N, Ariyosahi Y. A randomized phase III study of concurrent versus sequential thoracic radiotherapy (TRT) in combination with mitomycin (M), vindesine (V), and cisplatin (P) in unresectable stage III non-small-cell lung cnacer 9NSCLC): preliminary analysis. Proc Am Soc Clin Oncol 1997; 16: 1649a. 8 Sause WST, Scott C,Taylor S, Johnson D et al. RTOG 8808 ECOG 4588, preliminary analysis of a phase III trial in regionally advanced unresectable non-small cell lung cancer with minimum three year follow-up (abstract). Proc 37th Annual ASTRO meeting. Intl J Radiat Oncol Biol Phys 1995; 32 (Suppl 1): 195. 9 Vokes EE, Leopold KA, Herndon II JE, et al. A CALGB randomized phase II study of gemcitabine or paclitaxel or vinorelbine with cisplatin as induction chemotherapy (Inc CT) and concomitant chemoradiotherapy (XRT) in stage IIIB non-small-cell lung cancer (NSCLC): feasibility data (CALGB study #9431). Proc Am Soc Clin Oncol 1997; 16: 455a. 10 Tannehill SP, Mehta MP, Froseth C, et al. Phase II trial of hyperfractionated accelerated radiation therapy (HART) for unresectable non-small-cell lung cancer (NSCLC): preliminary results of ECOG 4593. Proc Am Soc Clin Oncol 1997; 16: 446a. Vol 350 • July 19, 1997 157 THE LANCET COMMENTARY NIDDM and breastfeeding See page 166 The study by David Pettitt and co-workers in today’s Lancet shows an association between breastfeeding and reduced risk of non-insulin-dependent diabetes mellitus (NIDDM). The relation was consistent by age cohort, by degree of breastfeeding, and by degree of obesity. Prospective randomised controlled studies investigating the relation between breastfeeding and NIDDM could theoretically be piggy-backed onto those underway for insulin-dependent diabetes. However, such studies would need to be extended for several decades since they are being conducted among ethnic groups with much lower prevalences of NIDDM and a higher age at diagnosis than Pimas. In view of the lack of markers with a high positive- predictive value for NIDDM and the length of time needed to complete randomised trials, it is essential that the findings of this new study are put into perspective. Although breastfeeding is now almost universally accepted as the ideal means of infant nutrition, one thing that has become clear is that those who choose to breastfeed are different from those choosing to bottlefeed. Studies in the USA over the same time period as the study showed that breastfeeding rates plummeted but that women who breastfed their infants were better educated and more likely to participate in health-promoting behaviours than other women. 1 Breastfeeding rates among Pimas ran in parallel with national rates over the study period, such that by 1970–77 under 20% were exclusively breastfed for at least 2 months. 2 In this cohort, those who wholly breastfed were initially less likely to be 100% Pima Making DOTS succeed See page 169 Shortly before World Tuberculosis Day, March 24, 1997, WHO announced a “breakthrough in tuberculosis control that will make it possible to save millions of lives”, namely DOTS (Directly Observed Treatment, Short Course). This announcement caused surprise and concern to many tuberculosis workers—surprise because the concept of DOTS is far from new, and concern that this strategy has not been adequately evaluated. 1,2 The report from Bangladesh by Mushtaque Chowdhury and colleagues in this issue of The Lancet shows that DOTS can indeed lead to a high cure rate and low drop-out and relapse rates. The key lesson from this study, however, is that a DOTS programme is much more than the mere tactic of supervised medication. Since few countries have adopted DOTS as a national strategy and only 10% of tuberculosis patients worldwide have access to it, it is important to examine the factors that contributed to success in the Bangladesh study and to ask whether these can be applied universally. These factors include the existence of an effective non-governmental organisation capable of securing technical and financial support from several donor agencies. Another is the network of diagnostic laboratories with technicians trained in sputum microscopy, a factor that itself is no small achievement. A key element is the use of community health workers recruited from village organisations that are involved in other aspects of health care, education, and improvement of socioeconomic conditions. In other tuberculosis-endemic regions, there may be a lack of public motivation, political factors that discourage international support, inadequate health-care infrastructures, and quite different community structures. Belief patterns about disease show considerable regional variation and profoundly affect case-finding and case- holding. HIV-related tuberculosis seems to be uncommon in Bangladesh but in regions with a high prevalence of HIV infection stigmatising factors are proving to be a serious barrier to detection and effective therapy. Thus, many local factors must be taken into consideration in the design of a control programme. Perhaps the most important factor contributing to success of a DOTS programme is the nature of the “therapeutic encounter.” Much has been written on the prevalent model of medical care with authoritarian prescribers and obedient, unquestioning patients. 3 The success of the Bangladesh programme may well stem from a partnership of equals, with well-informed patients being guided (rather than commanded) by health workers selected from their own community. In this context, it has been suggested that health-care workers should abandon the concept of patient “compliance” in favour of “concordance”. 4 Owing to these human and other factors, the implementation of the DOTS strategy calls for a revolution of thinking in many aspects of health-care provision. In many regions of the world this change may be slow in coming, if it comes at all, and the success of tuberculosis control may then be seriously compromised by the HIV pandemic and multidrug resistance. Thus, despite some successful DOTS programmes, it is essential that research into newer, more effective vaccines and novel therapeutic approaches should continue, since a combination of these with DOTS will be required for the eventual eradication of tuberculosis. John M Grange, Alimuddin Zumla Imperial College School of Medicine at the National Heart and Lung Institute, London, and University College London Medical School, London W1P 6DB, UK 1 Brown P.TB claims slammed as dangerous. New Sci 1997: April 19: 4. 2 Volmink J, Garner P. Directly observed therapy. Lancet 1997; 349: 1399–400. 3 Sumartojo E. When tuberculosis treatment fails. A social behavioural account of patient adherence. Am Rev Respir Dis 1993; 147: 1311–20. 4 Fox R. Are you a commander or a guide? J R Soc Med 1997; 90: 242–43.