Objectives: To demonstrate the usefulness of an effective and timely information model, underscore the seriousness of the problem of homicides, and point out the need to apply this type of model as well as comprehensive prevention projects, such as Desarrollo, Seguridad y Paz (DESESPAZ). From 1993 to 1998, 11 457 homicides were registered in Cali, Colombia, through an epidemiological surveillance model established under DESESPAZ by the mayor's office in Cali.
Methods: Beginning in January 1993, a work group organized by DESESPAZ reviewed and standardized the variables that different institutions gathered about the victims, their assailants, and the facts surrounding each case, and issued a weekly summary bulletin for the mayor and other local authorities.
Results: Between 1983 and 1994, the homicide rate increased from 23 to 124 per 100 000 inhabitants. Subsequently, rates went down in 1995, 1996, and 1997 to 112, 102, and 86,1 per 100 000, respectively, and again rose slightly in 1998 to 88 per 100 000. Even though people of all ages, including children under 5, have been victims of violence, the most affected group is that of men between the ages of 20 and 34. The ratio of men to women has varied from 14.3:1 to 9.2:1. In terms of numbers, percentages, and rates, low-income groups are the most seriously affected, although the highest-income groups have had rates as high as 160 per 100 000. A firearm was used in over 80% of homicides, and the crime was most often committed at night and on a weekend. A suspect was identified in only a few cases (8% to 21%). The bivariate analysis revealed a positive association with alcohol consumption by the victim, as well as with the use of firearms by the assailant (OR: 3.1; 95% CI: 2.6 to 3.6). Cases that occurred during a fight between individuals or during group fighting showed an association with the use of a sharp weapon and with alcohol consumption by the victim (OR: 1.9; 95% CI: 1.4 to 2.6).
Conclusions: A map shows the homicide distribution by neighborhood, and the benefits of a population-based surveillance model are discussed, particularly their usefulness for identifying risk factors and the measures that can be applied to prevent and control this form of violence.