Mesure objective des rapports sexuels non protégés et caractérisation du sous-rapportage des rapports non protégés chez les travailleuses du sexe au Bénin à l'aide de marqueurs biologiques de l'exposition au sperme

Authors: Giguère, Katia
Advisor: Alary, Michel
Abstract: Unprotected sex (UPS) is a major risk factor for human immunodeficiency virus (HIV) infection and must be measured as validly as possible in HIV prevention studies. To date, the questionnaire is the most commonly used tool to assess sexual behaviours. However, self-report of sexual behaviours is subject to recall and desirability biases. The use of biomarkers of recent semen exposure might help to overcome these biases. Prostatespecific antigen (PSA) and Y chromosomal DNA (Yc-DNA) are the most characterized biomarkers of semen exposure. PSA and Yc-DNA can be detected up to two and 14 days following UPS. Over the course of an early antiretroviral therapy (E-ART) and pre-exposure prophylaxis (PrEP) demonstration study that was conducted among professional female sex workers (FSW) in Cotonou, Benin, under-reporting of UPS was expected, as well as a change in UPS in the PrEP group. Our objectives were thus to validate self-report of UPS by the means of PSA and Yc-DNA detection; to compare under-reporting of UPS over the last two and 14 days; and to assess trends in UPS in the PrEP group. We also aimed to compare the UPS detection capability of a novel screening test of Yc-DNA, a nested polymerase chain reaction targeting the testis-specific protein Y-encoded family of homologous genes (n-TSPY), to six other commonly used methods to detect recent semen exposure. At baseline of the E-ART/PrEP study and over a 24 months period of follow-up, UPS from the last two and 14 days were assessed every six months by questionnaire and by PSA and Yc-DNA screening. Under-reporting of UPS in the last two or 14 days was defined as reporting no UPS in the last two or 14 days while testing positive for PSA or Yc-DNA, respectively. A robust Poisson regression was used to compare under-reporting over the last two and 14 days. Trends in UPS as measured with the different tools were assessed by the means of a log-binomial regression. Generalized estimating equations (GEE) were used to account for dependence between observations. At baseline, we observed about 20% of under-reporting of UPS among FSW from Cotonou. However, we observed no statistically significant difference between under-reporting in the last two days and under-reporting in the last 14 days. Some of our results suggest that the relative performances of each biomarker to detect UPS over its corresponding recall period are not equal, which might have prevented us to detect a difference in under-reporting over the two recall periods. No trend in UPS was observed in the PrEP group. Finally, using n-TSPY as a reference test, each of six commonly used methods to detect recent semen exposure lacked sensitivity in the detection of RSNP. In conclusion, self-report of UPS is biased and must be cautiously interpreted. A better characterization of the clearance of PSA and Yc-DNA is required in order to better evaluate the potential effect of the recall period length on under-reporting of UPS. The absence of any evidence of a trend in UPS in the PrEP group might suggests that there was no risk compensation over the PrEP demonstration study and that PrEP might be a suitable HIV prevention method to use among FSW. Our trends in UPS analyses also pointed out the necessity to objectively assess UPS by the means of biomarkers and to correct for the potential selection bias when assessing trends in UPS over the course of a longitudinal study with high attrition. Finally, the n-TSPY might be of great utility to detect UPS in observational studies where many factors might accelerate the clearance of the biomarkers.
Document Type: Thèse de doctorat
Issue Date: 2018
Open Access Date: 14 March 2019
Permalink: http://hdl.handle.net/20.500.11794/34010
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
SizeFormat 
34932.pdf2.68 MBAdobe PDFView/Open
All documents in CorpusUL are protected by Copyright Act of Canada.