Investigación original / Original research Physical and sexual violence, mental health indicators, and treatment seeking among street-based population groups in Tegucigalpa, Honduras Javier Rio Navarro,1 Julien Cohen,2 Eva Rocillo Arechaga,3 Edgardo Zuniga,3 and Médecins Sans Frontières Evaluation and Action Team–Honduras 3 Suggested citation Rio Navarro J, Cohen J, Rocillo Arechaga E, Zuniga E, Médecins Sans Frontières Evaluation and Action Team–Honduras. Physical and sexual violence, mental health indicators, and treatment seeking among street-based population groups in Tegucigalpa, Honduras. Rev Panam Salud Publica. 2012;31(5):388–95. ABSTRACT Objective. To establish the prevalence of exposure to physical and sexual violence, mental health symptoms, and medical treatment-seeking behavior among three street-based subpopu- lation groups in Tegucigalpa, Honduras, and to assess the association between sociodemo- graphic group, mental health indicators, and exposure to violence. Methods. An anonymous, cross-sectional survey among randomly selected street-based adolescents, adults, and commercial sex workers (CSWs) was undertaken at the end of 2010 in Tegucigalpa. Médecins Sans Frontières (MSF) mapped places where the study population gathers. Stratified probability samples were drawn for all groups, using two-stage random sampling. Trained MSF staff administered on-site standardized face-to-face questionnaires. Results. Self-reported exposure to severe physical violence in the previous year was 20.9% among street-based adolescents, 28.8% among adults, and 30.6% among CSWs. For the physical violence event self-defined as most severe, 50.0% of the adolescents, 81.4% of the adults, and 70.6% of the CSWs sought medical treatment. Their exposure to severe sexual violence was 8.6%, 28.8%, and 59.2%, respectively. After exposure to the self-defined most se- vere sexual violence event, 14.3% of adolescents, 31.9% of adults, and 29.1% of CSWs sought treatment. Common mental health and substance abuse symptoms were highly prevalent and strongly associated with exposure to physical (odds ratio 4.5, P < 0.0001) and sexual (odds ratio 3.7, P = 0.0001) violence. Conclusions. Exposure to physical and sexual violence reached extreme levels among street-based subpopulations. Treatment-seeking behavior, particularly after severe sexual violence, was limited. The association of mental health and substance abuse symptoms with exposure to violence could lead to further victimization. Medical and psychological treatments targeting these groups are needed and could help decrease their vulnerability. Key words Violence; sexual violence; mental health; vulnerable populations; urban health; Honduras. Honduras has reported the highest Central. These departments’ main cit- 1 Médecins Sans Frontières, México DF, México. homicide rates in Central America since ies, San Pedro Sula and Tegucigalpa, Send correspondence to: Javier Rio Navarro, 2008 and reached a national homicide concentrate 23.0% of the Honduras [email protected] rate of 77.5 per 100 000 population in population and 33.6% of the country’s 2 Institut National de la Santé et de la Recherche Médicale, U912 (SESSTIM), Marseille, France. 2010 (1–3). Violence is concentrated in homicides—17.0% and 16.6%, respec- 3 Médecins Sans Frontières, Tegucigalpa, Honduras. two departments: Cortes and Distrito tively (3). Socioeconomic factors associ- 388 Rev Panam Salud Publica 31(5), 2012 Rio Navarro et al. • Violence among street-based population groups Original research ated with raising levels of violence in alization due to their main source of in- plained the study purposes and proce- urban settings are a reality in Honduras come: selling drugs, petty crime, selling dures. No differences between subpopu- (4–9). About one-third of the people are sex, and recycling waste. Marginaliza- lation strata were found for the refusal underemployed, more than three in five tion was used as the main selection cri- rate, which was low, at 9%. A total of live below the poverty line, and 30% are terion among street-based people for its 283 street-based people were included in between the ages of 15 and 30 years (10, documented association with increased the analysis. 11). In this context, 7 of 10 Hondurans exposure to violence (27–29). A qualita- reported the cities’ streets as the nation’s tive study was conducted among street- Data collection procedures and most dangerous place (12). based adolescents, adults, and elderly instruments Violence in urban settings is a public and street-based male, female, and trans- health problem (13) with considerable vestite commercial sex workers (CSWs) At each data collection location, in- social consequences (14). In 2001, the (referred to in this paper collectively as terviewers explained the study proce- main referral hospital in Tegucigalpa CSWs); it was decided that quantitative dures, answered questions, and sought treated 1 228 nonfatal injuries caused by surveys should be conducted among informed consent before enrolling par- violence (15). By 2009, the number had three subpopulations: adolescents (10 to ticipants. A study supervisor answered risen to 5 421, a 4.4-fold increase (16). 18 years old), adults (18 to 72 years old), any remaining questions about informed Although the actual number of people and CSWs. consent and the study’s purposes and suffering the consequences of violence in Médecins Sans Frontières (MSF) medi- oversaw the data collection procedures. Honduras is not available, it is certain to cal street outreach teams mapped the The study was anonymous, with indi- exceed the number of people attending locations where these populations gath- vidual questionnaires identified by a medical facilities, as only a small percent- ered to meet their peers and where unique code. Interviewers administered age of victims seek medical care (17–19). commercial sex, drug selling, and recy- a pretested, structured questionnaire Furthermore, public health service provi- cling activities took place. These places that needed an average of 30 minutes sion in Honduras is limited by a lack of were visited at least once a week for for completion and included successive trained health professionals (20). one month; basic medical services were modules on sociodemographics, mental Street-based populations are particu- provided and health information and health, exposure to physical and sexual larly vulnerable to violence and abuse individual codes were assigned to each violence, and medical treatment-seeking (21–23). Increased risks faced by street- street-based person, specifying the sub- behavior. based population groups toward experi- population group each person belonged Exposure to physical violence during encing violence have been documented to. Sample frames were drawn for 25 the previous year and medical treatment- in different settings (21, 24–26). In Te- locations, weighted for peak gathering seeking behavior were assessed by the gucigalpa, high levels of distress were times, and stratified by subpopulations. World Health Organization’s question- documented among street-based ado- The sampling universe was estimated naire on violence surveillance at the lescents and youth (22), but no data on through capture–recapture methods, community level, adapted for an in- the factors associated with exposure to facilitated by the individual codes re- dividual self-report (31). Exposure to violence in the population groups stud- corded during the mapping phase. This severe physical violence was defined ied were available in Honduras. This method is widely used for population as having suffered at least one physical information would allow for identifying estimates of hard-to-reach population assault in which the victim was stabbed, the health services these most at-risk groups (30). The final sampling frame hit with objects, shot, or burned by one population groups require. consisted of 1 743 street-based people or more aggressors. This paper reports on a quantitative (highest estimate 2 499, lowest estimate Previous year exposure to sexual vio- survey measuring exposure to physi- 987). The study’s sample size was maxi- lence and medical treatment-seeking be- cal and sexual violence, the sociodemo- mized with an estimated prevalence of havior were explored by using the ques- graphic and mental health indicators 50%, a 5% significance level, and an tionnaire on sexual- and gender-based associated with such exposure, and the estimated precision of 5%. The initial cal- violence produced by the Reproductive medical treatment-seeking behavior of culation of 384 participants was adjusted Health Response in Crisis Consortium three subpopulations of street-based with a correction factor of 1.22 due to the (32). Participants who reported having people in Tegucigalpa. limited target population size, resulting been raped, as defined by the World in a final sample size of 314. The overall Health Organization (33), or forced to MATERIALS AND METHODS sample was divided proportionally to observe somebody being sexually as- each subpopulation stratum size, with saulted at least once during the previous Mapping and sampling 92 adolescents, 167 adults, and 55 CSWs. year were categorized as having suffered Fifteen locations were randomly se- severe sexual violence. In August and September 2010, forma- lected, with probability of selection pro- Data on mental health indicators tive research and medical street outreach portional to the population size. In each were collected with the