Long-term Effects of Male Circumcision on Risky Sexual Behaviors and STD Infections: Evidence from Malawian Secondary Schools

How to cite

Kim, Hyuncheol; Jung, Jaehyun; Kim, Booyuel; Pop-Eleches, Cristian. 2016. "Long-term Effects of Male Circumcision on Risky Sexual Behaviors and STD Infections: Evidence from Malawian Secondary Schools." AEA RCT Registry. AEARCTR-0001335. https://doi.org/10.1257/rct.1335-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Hyuncheol Kim* , Jaehyun Jung , Booyuel Kim , Cristian Pop-Eleches

* Primary investigator

Registry ID
AEARCTR-0001335
DOI
https://doi.org/10.1257/rct.1335-2.0
First registered
Last updated
Status
On going
Trial dates
2011-10-03 – 2016-08-31
Countries
Private
JEL codes
C93, I12, I18
Registry record
http://www.socialscienceregistry.org/trials/1335

Abstract

Medical male circumcision has drawn substantial attention as one of most effective HIV/AIDS prevention strategies since three studies proved that male circumcision can reduce HIV transmission risk by up to 50 percent (Auvert et al., 2005; Bailey et al., 2007; Gray et al., 2007). However, although male circumcision can reduce the infection risk in the short run, it does not eliminate an individual’s risk of infection. That is, circumcised men may participate in riskier sexual behaviors if they perceive reduction in risk following the procedure, and thus mitigate the direct preventive effect of male circumcision.
The purpose of this study is to investigate the causal relationship between medical male circumcision for secondary school students in Malawi and their risky sexual behaviors and infection of STDs after 3-4 years. We randomly provided free male circumcision surgery and transportation subsidies to 3,974 boys (9th-11th grade) at 124 classrooms of 33 public schools in Malawi. We can conduct a comprehensive examination of men’s behavioral change and the mechanism on how they compensate their reduced risk after circumcision by exploiting biomarkers and a variety of novel measures for sexual responses in the long-term.

Intervention

After the baseline, students in the treatment group received a free male circumcision offer. The male circumcision offer consists of free surgery at the assigned hospital and transportation subsidies. In addition, there are two complication check-ups (3-day and 1-week after surgery) at the student’s school. For transportation subsidies, students can choose either direct pick-up service or transportation voucher which is reimbursed after circumcision surgery at the hospital. Since this program was implemented as a phase-in randomized controlled trial, the students in the control group received the same treatment one year later.

Experimental design

We used the two-step randomization design for the sample of 3,974 boys (9th-11th grade) at 124 classrooms of 33 public schools. In the first round, we randomly assigned 41 classrooms (across 24 schools) to 100% Treatment, 41 classrooms (across 25 schools) to 50% Treatment, and 42 classrooms (across 28 schools) to the No Treatment group. A free male circumcision offer with transportation subsidies was provided for all the students in the100% Treatment classroom and randomly selected half of the students in 50% Treatment classroom.

Design details

Randomization method
Randomization of classrooms was done in office by computer random number generator
Randomization unit
Classroom and individual level
Clusters
2,661 completed baseline surveys for men (baseline 9th and 10th grade) who will be targeted in this 3-year follow-up study coming from 33 schools and 124 classes.
Observations
We expect to successfully interview about 91% (Effective Survey Rate) of men from the initial baseline for a total of 2,661 men (excluding baseline 11th grade students).
Arms
100% treatment: 41 classrooms (1,293 male students) 50% treatment: 41 classrooms (679 male students in treatment group, 679 male students in control group) Control: 42 classrooms (1,323 male students)
Primary outcomes
Main outcome variables include 1) risky sexual behaviors, and 2) HIV and HSV2 infection; 1.Risky sexual behaviors: We can exploit a variety of measures in the survey to analyze whether the boys who we induced to get circumcised by the intervention are changing their sexual behaviors. These measures encompass not only the conventional information on risky sexual practices like condom use and the number of sexual partners, but also innovative measures which are robust to potential measurement error. To be specific, demand for condoms can act as a valid proxy for demand for safe sex (Thornton, 2008), and the item count technique can increase the robustness of self-reported answers to sexual behavior questions to the social desirability bias (Coffman et al., 2013). 2. HIV and HSV2 infection: Our three measures of STD infections are HIV infection and HSV2 infection by serum IgG and IgM. IgG captures lifetime HSV2 infection while IgM detects very recent one. Based on results from the pilot, we expect that we have more statistical power to detect difference in HSV2 infection when we measure it by IgG than by serum IgM. We may be underpowered to detect statistical differences in the HIV infection given that the HIV prevalence rate is relatively low. According to Malawi AIDS Response Progress Report 2015, percentage of women aged 15-24 who are living with HIV is 3.6%. As a result, the main outcome variable for this hypothesis will be the HSV2 infection rate measured by serum IgG. The secondary outcome will be the HIV infection rate, and the HSV2 infection rate measured by serum IgM will be tertiary.

Cited by

No citation data available for this registration yet.

Keywords