Impact evaluation of MSM HIV/AIDS prevention activities in Mexico – the use of quasi experimental methods for measuring behavioral change
C. Gayet1, A. Ospina2, A. Madrigal1, R. Roman2, J.C. Garcia-Benitez3, I. Ahumada-Lobo1
1FLACSO Mexico, Mexico D.F., Mexico, 2Population Services International / Mexico, Mexico D.F., Mexico, 3Ahumada-Lobo Assoc., Mexico D.F., Mexico
Background: MSM is the most affected population in Mexican HIV/AIDS epidemics (10% HIV prevalence). PSI-Mexico has developed behavior change communication campaigns and condom social marketing for MSM. The objective of this paper is to evaluate their impact on condom use, carrying of condoms, testing for HIV and number of sexual partners.
Methods: A quasi experimental method was applied based on a 2009 MSM RDS survey conducted in five Mexican cities (ethical procedures were implemented). 827 MSM voluntarily answered a questionnaire about sexual behavior and exposition to PSI-Mexico prevention activities. We estimated four alternatives of the Propensity Score Matching (PSM) method to compare behavior of MSM exposed to prevention activities (exposed group) with that of MSM not exposed (control group). High, medium and no exposure were considered.
Results: The main results (with variations depending on the PSM alternative) are:
* In the previous year, the use of condoms in the exposed group was between 30% and 60% higher than in the control group.
* The probability of carrying a condom in the exposed group was between 19% and 45% higher than in the control group. However, the probability of carrying condoms in the exposed group is still low (5% to 9%).
* The program increased between 15% and 76% the number of times that an individual has been tested for HIV.
* In the previous year, the exposure group had approximate 3 anal sex partners less than the control group.
It is acknowledged that limitations of the PSM method should be discussed.
Conclusions: In homophobic societies it is difficult to evaluate the impact of HIV prevention activities in MSM population with experimental methods. The use of Propensity Score Matching was useful for identifying successful prevention interventions and assessing the magnitude of behavioral changes. PSI-Mexico prevention activities have improved MSM prevention behavior.
Download the e-Poster