Cash payments to stay in school reduce HIV incidence in girls and young women, eSwatini study finds

Financial incentives to remain in school reduced HIV incidence among adolescent girls and young women in eSwatini (Swaziland) by 21%, and participants exposed to both financial incentives and a lottery open only to those who remained free of sexually transmitted infections (STIs) were 37% less likely to acquire HIV infection, a trial has found.

Study participants not in education or training were 190% more likely to acquire HIV during the follow-up period.

The findings differ from those of two large randomised studies which reported results in 2015, each showing that financial incentives for school attendance did not reduce HIV incidence in adolescent girls and young women in southern Africa.

Glossary

control group

A group of participants in a trial who receive standard treatment, or no treatment at all, rather than the experimental treatment which is being tested. Also known as a control arm.

trend

In everyday language, a general movement upwards or downwards (e.g. every year there are more HIV infections). When discussing statistics, a trend often describes an apparent difference between results that is not statistically significant. 

syphilis

A sexually transmitted infection caused by the bacterium Treponema pallidum. Transmission can occur by direct contact with a syphilis sore during vaginal, anal, or oral sex. Sores may be found around the penis, vagina, or anus, or in the rectum, on the lips, or in the mouth, but syphilis is often asymptomatic. It can spread from an infected mother to her unborn baby.

trial design

How a clinical study or trial is structured to answer the questions being asked, e.g., open-label or double-blind, comparative or observational.

The findings of the Sitakhela Likusasa study were presented at the Tenth International AIDS Society Conference on HIV Science (IAS 2019) in Mexico City last month.

The study was designed to investigate whether female vulnerability to HIV infection in eSwatini could be reduced by improving school attendance and the financial independence of adolescent girls and young women. Only a third of adolescent girls were enrolled in secondary school in 2014, few girls return to school once they drop out, and only 40% of young women have jobs. Transactional sex with older men is common; 15% of women aged 15-24 reported sex with a partner at least ten years older in 2014.

Study design and results

The study recruited 4389 adolescent girls and young women aged 15-22 from 266 enumerator areas, 80% rural. While half of participants were already in school or another form of education, half were not enrolled in any form of education. Participants were assigned equally to four study sub-arms:

  • Educational incentives and raffle interventions
  • Educational incentives intervention, no raffle
  • Raffle intervention, no educational incentives
  • No interventions.

Participants assigned to education incentives received approximately US$100 a year for enrolling in and attending school in each school year (2016-17, and 2017-18). In 2018, tuition fees of up to US$200 were paid for out-of-school participants. Young women could also receive up to US$100 a year for enrolling in, and completing, tertiary education or vocational short courses.

In each raffle round, 400 participants were randomly selected for testing for syphilis and trichomonas vaginalis. Those who were free of the STIs were entered in the raffle and 80 prizes were awarded in each round. A total of 560 prizes were awarded throughout the study and 19% of all study participants won at least one prize. One participant won a raffle prize five times.

Recruitment began in November 2015 and final study data collection, including HIV antibody testing, took place between November 2018 and April 2019. Eighty-six per cent of study participants remained enrolled in the study up to the final data collection period.

Participants in the education incentives arm were less likely to acquire HIV during the study. HIV incidence in the education incentive arm was 21% lower than in the control arm (8.08 vs 6.34%, p = 0.036) and 23% less likely to become HIV positive (OR 0.77, 95% CI 0.60-0.98, p = 0.037). Participants who were not enrolled in education were 190% more likely to acquire HIV. Participants who showed a propensity for risk taking in a financial decision-making game were 145% more likely to acquire HIV.

Education participation in years 1-5 of school appeared to be more protective than participation in education in years 1-6, (59% reduction in risk for girls in years 1-5 compared to 29% reduction in girls in years 1-6) suggesting that education is especially protective for younger adolescent girls and that the age range studied in trials of educational incentives is important for understanding the generalisability of the results.

An 'on treatment' analysis, which looked at HIV incidence according to the number of cash payments received (including raffle payments), showed a strong trend towards lower incidence in those who received more payments, up to eight payments, but no additional effect during the three-year follow-up for those who received a higher number of payments (p = 0.003). HIV incidence was 2% in those who received 8-12 payments compared to 8% in those who received two payments.

The study also found that exposure to both educational incentives and raffle entry was associated with the largest reduction in HIV incidence. Participation in the raffle-only sub-arm was not associated with reduced HIV incidence but participants in the raffle-plus-educational incentives arm were 37% less likely to acquire HIV (OR 0.63, 95% CI 0.44-0.91, p = 0.013).

Participants in the educational incentives arm were significantly more likely to be enrolled in school at the end of the study than participants in the control arm (p < 0.001).

Differences between studies

In the HPTN 068 study in Mpumalanga Province, South Africa, cash transfers were paid to the households of adolescent girls. Cash transfers had no impact on school attendance or HIV incidence but did reduce the likelihood of unprotected sexual intercourse. School attendance, regardless of cash transfer receipt, was associated with a reduced risk of HIV acquisition.

In the CAPRISA 007 study in KwaZulu-Natal province, South Africa, a cash transfer was conditional on annual HIV testing, pass performance in 6-monthly school tests and participation in HIV educational interventions. CAPRISA 007 recruited both adolescent girls and boys. The study was unable to detect an impact on HIV incidence because incidence was lower than expected, and cash transfers did not affect school attendance.

In the Sitakhela Likusasa study, financial incentives were paid directly to the girls, the study population was larger than either HPTN 068 and CAPRISA 007, and HIV incidence was higher (8.08% in the control arm over three years, compared to approximately 1.5% per year in the previous studies). However, further analysis will be needed to explore the impact of the interventions at different ages and to provide more insight into how promotion of school attendance affected sexual behaviour in study participants.

References

Gorgens M et al. Sitakhela Likusasa Impact Evaluation: results of a cluster randomized control trial (cRCT) of financial incentives for HIV prevention among adolescent girls and young women (AGYW) in Eswatini. 10th IAS Conference on HIV Science, Mexico City, abstract TUAC0205LB, 2019.

View the abstract on the conference website.