Vulnerability to homicide in Karachi: political activity as a risk factor.

BACKGROUND Previous studies analysing Karachi ambulance data from 1993 to 1995 identified neighbourhoods in Karachi disproportionately affected by homicide. As a step toward developing intervention programmes to curb violence, we conducted a study to identify risk factors for becoming a homicide victim in a high violence area of Karachi. METHODS We interviewed families of 35 cases, individuals intentionally killed through acts of violence between January 1994 and January 1997, and 85 community-based controls frequency matched by sex, from Orangi, a high violence area of Karachi. RESULTS Most of our cases and controls were male (97% and 92%, respectively) and had similar socioeconomic and ethnic backgrounds. All the victims were killed by firearms; 4 (11%) had been tortured prior to death. Most of the victims were killed in the streets (n = 25, 71%). Of these, 7 (36%) had been killed by law-enforcement officers, while 6 (24%) died from indiscriminate firing. People who were killed were 34 times more likely to have attended all political processions (29% versus 1%, odds ratio [OR] = 34; 95% CI: 4-749, P < 0.001), 19 times more likely to have attended political meetings (31% versus 2%, OR = 19; 95% CI: 4-136, P < 0.001), and 17 times more likely to have held an important position in a political party (29% versus 2%, OR = 17; 95% CI: 3-120, P < 0.001) than controls. CONCLUSIONS Homicide in Orangi was political. Efforts to improve trust between ethnic groups and to build legitimacy for non-violent forms of conflict resolution are important steps to limit future violence.