Home » DHS-7 Questionnaire Review - Archived » HIV/AIDS » Capturing Transactional Sex
|Capturing Transactional Sex [message #1707]
||Tue, 01 April 2014 11:50
Registered: March 2014
Location: Washington D.C.
STRIVE case for including questions on Transactional sex in DHS questionnaire for sub-Saharan Africa|
The recommendations below follow those offered by Tisha Wheeler of USAID concerning the importance of distinguishing sex work from informal transactional sex (TS). We strongly agree with the importance of making such a distinction, both on epidemiological and intervention related grounds, and we have specific recommendations regarding how TS as differentiated from sex work can be measured.
I am writing on behalf of the STRIVE Working Group on Transactional Sex and HIV (TS-WG), for which I am a co-chair, to make specific recommendations on ensuring better measurement and therefore understanding of the role of transactional sex in HIV risk, acquisition and transmission.
STRIVE is a five-year DFID-funded Research Programme Consortium (RPC) addressing structural drivers and pathways to HIV (http://strive.lshtm.ac.uk/). The STRIVE Working Group on Transactional Sex and HIV (TS-WG) brings together researchers, programme implementers and research uptake experts with a shared interest in understanding and curtailing the role that informal sexual exchange plays in adolescent girls' heightened vulnerability to HIV in sub-Saharan Africa (SSA).
The STRIVE TS-WG recently convened a two-day meeting in London focused exclusively on developing greater clarity about how to conceptualise TS; a paper and measurement brief, specifically laying out recommendations on the measurement of TS; and a systematic review of current evidence of the association between TS and HIV in sub-Saharan Africa, and the methodological lessons to be drawn from this research. This issue was identified by STRIVE as being of critical importance for HIV prevention, especially within sub-Saharan Africa, as there is strong epidemiological evidence that adolescent girls (ages 15-19) are at heightened risk of HIV, and identifying the causal factors associated with this risk is of utmost importance. A growing body of evidence, both qualitative and quantitative, points to the role that transactional sex, or informal sexual exchange, contributes to adolescent girls' heightened risk of HIV. In order to capture the role that transactional sex plays in young women's HIV risk, it is essential that we measure the prevalence of this practice rigorously, and over time. TS was measured, somewhat inadequately, for women in earlier DHS surveys across SSA; and the topic was not included in more recent survey waves.
Ensuring that the DHS captures valid and reliable data on transactional sex would be an invaluable contribution to HIV prevention, and we urge its serious consideration.
We recently completed a systematic literature review of published evidence on TS in sub-Saharan Africa, with the aim both of compiling current evidence, and lessons on how to better conceptualise and measure transactional sex in surveys. Our findings emphasize the importance of differentiating TS both from sex work and from economic exchange in marriage (or other primary relationships), with a need to document the extent that women may have either "one off" or longer term sexual relationships that are primarily financially motivated.
Based upon our extensive review of different approaches to conceptualising and measuring transactional sex, and our discussions at the STRIVE TS-Working Group meeting, we propose the following working definition for TS:
Transactional sex refers to a sexual relationship or act(s), outside of marriage or sex work, motivated primarily by the expectation of material gain where love and trust are sometimes present.
This working definition guides the questions we propose below for measurement purposes, and draws heavily on definitions provided by Mojola, Hunter, Dunkle and Jewkes. The measurement questions we recommend below are conceptually aligned with the above definition, and are similar to those used in repeated surveys by Dunkle and Jewkes.
We would like to emphasize that we are at the beginning stages of this body of work. We will be conducting cognitive interviews to test and improve on the specific wording of the questions we recommend below to generate indicators on TS. We would be more than happy to work with ICF-Measure DHS as we go forward with this effort.
The STRIVE Consortium recommend adding questions to the woman questionnaire and revising and adding questions to the male questionnaire, to ensure a better understanding of the role of transactional sex in HIV risk, acquisition and transmission.
RECOMMENDED ADDITION: WOMEN'S QUESTIONNAIRE
1. Information needed: The prevalence of women who report having practiced transactional sex, the frequency with which it is practiced, and the magnitude of the practice (i.e. number of partners with whom they have practiced transactional sex).
2. Question that will elicit this information: We recommend adding the following questions to measure transactional sex.
Question to come after question 619 (on history of most recent sexual partners, would be asked of each partner):
• Did you have sex or become sexually involved with this person because they provided you with, or you expected that they would provide you with material support of any kind? (then would be asked of every partner)
o 1 Yes
o 2 No
Question to come after 626:
• In the past 12 months have you had sex or become sexually involved with someone because they provided you with, or you expected that they would provide you with material support of any kind? (skip pattern, if yes)
o 1 Yes
o 2 No
• In the last 12 months, how many different persons have you had sex or become sexually involved with because they provided you with, or you expected that they would provide you with material support?
o __ __ (insert number)
o 98 Don't Know/Don't remember
3. How will the resulting information be used: The resulting information will be used to capture the prevalence of having had transactional sex in the last 12 months, and with how many partners; as well as the percent of recent partnerships that could be categorized as "transactional sexual partners." Together this information will provide a valid understanding of the practice of TS; and one that can then be compared across and within countries, age groups, as well as used as an outcome measure in multivariate models.
Indicators generated from these questions (either overall or within specified age ranges):
• 1.Percent of sexually active women reporting a transactional sexual relationship with a recent sexual partner in the past 12 months :
- [Numerator]No. of women reported transactional sexual relationships in the last 12 months with any of up to 3 recent sexual partners / [Denominator] total number of sexually active women
- This question can also then be used to look at differences in partner characteristics for those partners who are reported as TS partners (age gap, type of relationship acquaintance, boyfriend; condom use with partner); as well as differences in sexual risk behaviors with these as compared to other partners (e.g. condom use)
- In addition, in multivariate analysis, one could then assess the extent to which women reporting 2 or more transactional sexual relationships should be considered as a 'risk group'
• 2. Percent of sexually active women reporting having had transactional sex in the last 12 months:
- [Numerator] No. of sexually active women reporting having had transactional sex in the last 12 months/ [Denominator] No. of sexually actively women
• 3. Average number of transactional sexual partners among women who report having practiced transactional sex in the last 12 months:
- [Numerator] aggregate number of transactional sexual partners reported in last 12 months/ [Denominator] number of women reporting TS in last 12 months
- In addition, in multivariate analysis, one could then assess the extent to which women reporting 2 or more transactional sexual partners should be considered as a 'risk group'
4. What is the priority of suggested additions compared with what is already in the questionnaires? The suggested questions hold higher priority than Questions 622 and 623. While 622 is important, it relies on respondent's recall, and there are additional questions in the DHS assessing concurrency (616, 625). For question 623 (frequency of sexual intercourse with a given partner in the last 12 months), this question would be difficult to respond to accurately, from the point of view of the respondent, and may therefore not be as informative as intended. We feel understanding the prevalence of transactional sex, and having opportunity to assess the relationship between transactional sex and other risky outcomes, takes precedence over these questions.
5. If suggesting more than one addition, what is the priority among the suggested additions?
• First priority: In the past 12 months have you had sex or become sexually involved with someone because they provided you with, or you expected that they would provide you with material support such as, clothes, food, cash, or some other benefit? (skip pattern, if yes)
• Second priority: In the last 12 months, how many different persons have you had sex or become sexually involved with because they provided you with, or you expected that they would provide you with material support?
• Third priority: Did you have sex or become sexually involved with this person because they/he provided you with, or you expected that they/he would provide you with material support such as, clothes, food, cash, or some other benefit?
6. Should the additional data be collected in all countries or only in selected types of countries?
We would recommend that these data are included for all sub-Saharan African countries, where the practice of TS corresponds to the definition that has guided the construction of the question, and where the literature suggests that TS is putting adolescent girls and young women at heightened risk.
RECOMMENDED REVISION: MEN'S QUESTIONNAIRE
In agreement with recommendations from Tisha Wheeler, we would urge DHS to consider suggestions for more accurately assessing men's sexual behavior with sex workers. We would strongly recommend replacing the wording of questions 430 - 432 which prompt men to discuss their experience with 'paid sex,' with questions that more clearly differentiate sex with a sex worker from transactional sex (or informal exchange). As it stands, the question conflates sex work with transactional sex and therefore the responses to this question will undoubtedly vary significantly depending on whether or not this phrasing is interpreted as 'sex work' by the respondent, or as a spectrum of sexual exchange including more informal exchange within relationships. Indeed, the DHS reports on this question are themselves inconsistent, at times describe this measure as capturing transactional sex, while in others, describing this measure as capturing sex with a sex worker.
The following recommended changes to the wording of questions 430-432 are meant to restrict the interpretation of the question to those sexual encounters wherein a man paid a person he identified as a sex worker in immediate exchange for sex. (Also, this wording omits sexual relationships where the man is a regular partner of a sex worker; or informal sexual exchange within the context of relationships or that is characterized as outside of sex work).
• In lieu of question 430, we would recommend a question specifically focused on sex work, such as:
o In the last 12 months, did you pay to have sex with a sex worker (or prostitute)?
• In lieu of question 431, we would recommend:
o Have you ever paid to have sex with a sex worker (or prostitute)?
• In lieu of question 432, we would recommend s:
o Was a condom used during sexual intercourse every time you paid to have sex with a sex worker in the last 12 months?
RECOMMENDED ADDITION: MEN'S QUESTIONNAIRE
We would also recommend, in parallel with our above suggested added questions concerning transactional sex for women, that the DHS consider including a similar set of questions for men. Below, we provide the parallel phrasing for men for questions that would focus on transactional sex as we define it above. These questions still require cognitive testing to determine their validity; however, we would suggest that it's equally important to identify the percentage of men who reporting engaging in transactional sex as it is to identify the percentage of women who report this practice. As the justifications and indicators for men would resemble those given above for women, we do not repeat these components, but do list our recommended added questions below:
• (To be added following question 425) Did you provide any kind of material support in order to have sex or become sexually involved with this person?
o 1 Yes
o 2 No
• (To be added in between questions 433 and 434) Have you ever provided any kind of material support in order to have sex or become sexually involved with a woman other than a sex worker?
o 1 Yes
o 2 No
• (To follow the above question) In the past 12 months, have you provided a woman, other than a sex worker, with any kind of material support in order to have sex or become sexually involved with her?
o 1 Yes
o 2 No
• (To follow the above question) In the past 12 months, how many different women have you provided with material support in order to have sex or become sexually involved with them?
o __ __ (insert number)
o 98 Don't Know/Don't remember
We would suggest that parallel measures and indicators as to those for women could then be developed for men. Likewise, we would suggest that these questions would take priority over parallel questions present in the women's questionnaire (Questions 423 and 424). Again, we would suggest the same prioritization of these questions as for the women's questionnaire, with the highest priority going to the following question: In the past 12 months, have you provided a woman (other than a sex worker) with any kind of material support in order to have sex or become sexually involved with her? Finally, we would recommend that these questions be included for all countries in sub-Saharan Africa, in particular.
Kirsten Stoebenau, PhD
Gender and Population Specialist
Current Time: Sun May 9 03:33:20 Coordinated Universal Time 2021