Effect of a lottery intervention on gender-based violence among female sex workers in Dar es Salaam, Tanzania: results from a randomised trial.

INTRODUCTION

Financial incentives are a promising approach for HIV prevention. Some studies have also shown positive spillover effects on gender-based violence (GBV); however, little is known about impacts of financial incentives on GBV among female sex workers (FSW). We investigated the impact of lottery-based incentives on GBV among FSW in Dar es Salaam, Tanzania.

METHODS

The RESPECT II study was a parallel-arm trial (n=2206) evaluating the effect of lottery-based incentives on HIV/sexually transmitted infections (STIs) among FSW. Participants were randomised 1:1 to (1) the basic test group (control), including testing and counselling for HIV/STIs and bi-weekly text messages on safe sex or (2) the lottery group, including the basic test group intervention plus entry into a weekly lottery for an award of 100,000 Tanzanian Shilling (US$50) conditional on negative syphilis and Trichomonas vaginalis tests. We conducted an intent-to-treat analysis using additive risk models (binomial distribution, identity link) to compare physical and sexual GBV, intimate partner violence (IPV) and non-partner violence between study arms at endline, with estimates expressed as prevalence differences (PDs) and 95% CIs. Causal, population-level impacts were estimated using g-computation.

RESULTS

All forms of violence declined over the study period in the endline sample (n=1117; 49.4% lost to follow-up). There were no significant differences between arms at endline in the primary analysis: GBV - 34.1% control versus 30.7% lottery, PD: -0.03, 95% CI -0.09 to 0.02; IPV - 18.9% control versus 15.9% lottery, PD: -0.03, 95% CI -0.07 to 0.01; non-partner violence - 21.1% control versus 20.4% lottery, PD: -0.01, 95% CI -0.06 to 0.04, nor in the g-computation analysis.

CONCLUSION

This study did not find any significant impact of low-probability, high-reward lottery incentives on any form of GBV, including IPV and non-partner violence, suggesting that this intervention does not influence risk of violence for FSW in the context of HIV prevention. Nonetheless, given the high attrition in the study sample, effects of the intervention cannot be ruled out, and additional investigation is warranted among a larger sample to further assess the potential benefits and harms of this approach for FSW.

TRIAL REGISTRATION NUMBER: AEA RCT registry: AEARCTR-0002677.