Document Type


Publication Date



Summary—A study of female sex workers in Phnom Penh, Cambodia found self-reported condom use to be of questionable validity, particularly among amphetamine-type stimulant (ATS) users and those with multiple partners.

Background—Accurate measurement of unprotected sex is essential in HIV prevention research. Since 2001, the 100% Condom Use Program targeting female sex workers (FSW) has been a central element of the Cambodian National HIV/AIDS Strategy. We sought to assess the validity of self-reported condom use using the rapid prostate-specific antigen (PSA) test among Cambodian FSW.

Methods—From 2009 to 2010 we enrolled 183 FSW in Phnom Penh in a prospective study of HIV risk behavior. PSA test results from the OneStep ABAcard® were compared to self-reported condom use in the past 48 hours at quarterly follow-up visits.

Results—Among women positive for seminal fluid at the first follow-up visit, 42% reported only protected sex or no sex in the detection period. Discordant results were more likely among brothel and street-based FSW vs. entertainment (56% vs. 17%), recent (last 3 months) ATS users (53% vs. 20%), and those with >5 partners in the past month (58% vs. 13%). In multivariable regression models, positive PSA results were associated with recent ATS use (Adjusted Risk Ratio (ARR) = 1.5; 95% confidence interval (CI):1.1 – 2.2), having a non-paying last sex partner (ARR=1.7; CI: 1.2 – 2.5), and sex work venue (ARR=3.0; CI:1.4 – 6.5). Correspondingly, women with a nonpaying last sex partner were more likely to report unprotected sex (ARR=1.5; CI:1.1 – 2.2), but no associations were found with sex work venue or ATS use.

Conclusions—Results confirm the questionable validity of self-reported condom use among FSW. The PSA biomarker assay is an important monitoring tool in HIV/STI research including prevention trials.


This work is the Accepted Author Manuscript (not the final published version) of an article published by Lippincott, Williams & Wilkins

The published version is accessible at



Included in

Epidemiology Commons