For personal use. Only reproduce with permission from The Lancet. The importance of conflict-related mortality in civilian populations Debarati Guha-Sapir, Willem Gijsbert van Panhuis Civil conflict affects the health of individuals in many countries, and draws a substantial amount of international humanitarian aid. The most widely used indicator of the effect of conflict is the rate of civilian death during conflict. We aimed to assess mortality estimates from conflicts in Sudan, Somalia, the Democratic Republic of Congo, and Afghanistan by calculating the relative risk of death during and after conflict compared with that in preconflict peacetime. Katale, in the Democratic Republic of Congo, had the highest relative risk of death during conflict (11·2 [9·1–13·8] and 103·3 [94·7–112·6], for children younger than 5 years and the whole population, respectively). Our results suggest that high rates of civilian mortality are determined more by the pre-existing fragility of the affected population than the intensity of the conflict. In many instances, a high rate of civilian deaths during conflict shows that international development aid before the conflict was grossly inadequate. Lancet 2003; 361: 2126–28 In 2002, more than 39 countries were engaged in or were recovering from a civil conflict. Most conflicts last several years and profoundly affect individuals in part or all of the country. In the past decade, the international community has provided substantial resources for humanitarian aid (mostly medical RESEARCH LETTERS Conflict of interest statement None declared. Acknowledgments We thank Ying Liu and William Stewart for support, and Li Zhen, Fen Ding, Qin Zhang, and Ying Shao for research assistance. This research was supported by National Institute of Mental Health (NIMH grant MH42459 [Center for AIDS Prevention Studies]). The funding source had no role in the study design, data collection, data analysis, data interpretation, or writing of this report. 1 The UN Theme Group on HIV/AIDS in China. HIV/AIDS: China’s titanic peril, 2001 update of the AIDS situation and needs assessment report. Geneva, Switzerland: UNAIDS, 2002. 2 Beach M. China responds to increasing HIV/AIDS burden and holds landmark meeting. Lancet 2001; 358: 1792. 3 Zhang K, Ma S. Epidemiology of HIV in China: intravenous drug users, sex workers, and large mobile populations are high risk groups. BMJ 2002; 324: 803–04. 4 Zhang B, Liu D, Li X, Hu T. A survey of men who have sex with men (MSM): mainland China. Am J Public Health 2000; 90: 1949–50. 5 Kumta S, Setia M, Jerajani HR, et al. Men who have sex with men and male-to-female transgender in Mumbai: a critical emerging risk group for HIV and STI in India. XIV International Conference on AIDS. Barcelona, Spain: July, 2002 (abstr TuOrC1149). Center for AIDS Prevention Studies, University of California, 74 New Montgomery, Suite 600, San Francisco, CA 94105, USA (K-H Choi PhD, L Han PhD, J S Mandel PhD, G W Rutherford MD); Beijing Municipal STD Clinic, Beijing, China (H Liu MD); and Beijing Tongzhi Hotline Programme, Beijing, China (Y Guo) Correspondence to: Dr Kyung-Hee Choi (e-mail: khchoi@psg.ucsf.edu) 2126 THE LANCET • Vol 361 • June 21, 2003 • www.thelancet.com Variable Number HIV-1 positive Odds ratio (95% CI) Adjusted odds ratio* (95% CI) Overall 481 15 (3·1%) ·· ·· Age (years) 18–39 448 11 (2·5%) 1·00 1·00 40–69 33 4 (12·1%) 5·48 (1·64–18·27) 4·48 (1·31–15·33) Marital status Never married 409 10 (2·4%) 1·00 ·· Ever married 72 5 (6·9%) 2·98 (0·99–8·98) Education (age range in years) Junior high school or less (12–14) 95 3 (3·2%) 1·00 ·· Senior high school (15–17) 187 5 (2·7%) 0·84 (0·20–3·60) College graduate (18–21) 199 7 (3·5%) 1·10 (0·55–2·20) Had Beijing residence card No 312 9 (2·9%) 1·00 ·· Yes 169 6 (3·6%) 1·24 (0·43–3·54) Self-reported sexual orientation† Homosexual 282 9 (3·2%) 1·00 ·· Bisexual 162 5 (3·7%) 1·06 (0·37–3·03) Heterosexual or undecided 37 0 (0·0%) Type or place of recruitment‡ Personal contact 246 10 (4·1%) 1·95 (0·66–5·78) ·· Bars 163 5 (3·1%) 1·00 Parks and bath-houses 72 0 (0·0%) AIDS knowledge score§ 1–5 70 3 (4·3%) 1·00 ·· 6–7 127 2 (1·6%) 0·36 (0·06–2·19) 8–9 284 10 (3·5%) 0·82 (0·22–3·04) Ever had STD No 371 11 (3·0%) 1·00 ·· Yes 109 4 (3·7%) 1·25 (0·39–4·00) Ever tested for HIV-1 No 392 10 (2·6%) 1·00 ·· Yes 88 5 (5·7%) 2·30 (0·77–6·91) Ever had sex with women No 171 3 (1·8%) 1·00 ·· Yes 308 12 (3·9%) 2·27 (0·63–8·16) Lifetime number of male sex partners 1–20 352 7 (2·0%) 1·00 1·00 21 or more 123 8 (6·5%) 3·43 (1·21–9·66) 3·00 (1·04–8·61) *Only age entered in model because highly correlated with marital status ( 2 =66·12, p<0·0001). †Bisexual and heterosexual or undecided combined to calculate odds ratios. ‡Bars and parks and bath-houses combined to calculate odds ratios. §Nine true-or-false statements about HIV-1 transmission used to assess AIDS knowledge; median=8. HIV-1 seroprevalence by participants’ characteristics