Cambodia has the highest HIV prevalence of any Asian country, and over the last decade has experienced the most serious HIV/AIDS epidemic in Southeast Asia . Heterosexual contact is the major route of HIV transmission, and female sex workers (FSW) remain the group at highest risk . Although crucial progress has been made in reducing risky sexual behavior, including widespread condom use and promotion of reduced number of sexual partners, HIV prevalence among FSW remains high, ranging from 11% to 26% [2–5]. Poverty , low literacy , a high prevalence of sexually transmitted infections (STI) , and a highly mobile workforce  are contributing factors to the epidemic. Recent research has also identified drug use and, in particular, amphetamine-type stimulant (ATS) use as a serious emerging problem associated with HIV risk among FSW [3, 9, 10], which threatens to reverse downward trends in HIV infection rates in the region.
Amphetamine-type stimulants include a range of synthetic psychostimulants, including methamphetamine, amphetamine, and ecstasy, which can be injected, smoked, or taken orally. Effects of these drugs include euphoria, alertness, arousal, increased libido, increased sympathetic nervous responses, (heart rate, respiratory rate, blood pressure), and perceived increases in confidence, energy and physical strength . In Cambodia, a pill form of methamphetamine known as “yama” is widely produced, trafficked, and used. The tablets generally contain about 25% methamphetamine. “Crystal” (also known as “ice”) is generally about 85% methamphetamine and more addictive. Although yama pills are swallowed, both forms are usually melted and the vapors inhaled, resulting in rapid neurologic effects [11, 12]. Use of ATS has been associated with elevated HIV risk behavior in many countries and in several population subgroups [3, 9, 13–19]. The United Nations Office on Drugs and Crime (UNODC) reports that use of these drugs is widespread in Asia and increasing rapidly in Cambodia . In Cambodia, ATS are highly available both in pill and crystalline form and are generally ingested or smoked; injection use is uncommon . The Cambodia National Authority for Combating Drug Abuse (NACD) estimated that 70% of all drug users in Cambodia use ATS . The drug accounts for the majority of all drug seizures by authorities, and, in pill form, has been ranked as the leading drug of abuse for the past nine years with consistent increases since 2006 , at which time it was estimated that 30,000 tablets of yama were consumed orally or smoked there daily. Use is particularly high among vulnerable populations, including FSW [3, 9], men who have sex with men (MSM), and street children [12, 20, 21].
Measuring drug use in epidemiological research studies poses challenges. Self-reported measures of drug use have the advantage of being noninvasive and permit evaluation over longer time periods compared with biochemical assessments . However, study participants may misrepresent drug use due to social desirability bias, stigma, poor recall, poorly worded questions, or poorly worded response categories in surveys and interviews [23, 24], all of which could result in misclassification of measured exposures. Although studies have shown that the use of Audio Computer-Assisted Self Interview (ACASI) increases reporting of sensitive and stigmatized behaviors [25, 26], research suggests that the validity of self-reported drug use varies by population , race/ethnicity [28–30], mental health , and drug treatment status [22, 31, 32], although not by gender [30, 33]. Accuracy has varied in studies of arrestee populations [34, 35] but have been reported as higher in groups sampled in emergency department and STI clinics . Those that report more frequent drug use, compared to infrequent use, are more likely to self-report recent drug use [27, 36]. Urine toxicology assessments provide sensitive and valid measures of many drug types; but some, like ATS, are restricted to a short timeframe due to rapid metabolization. The detection window may also depend on the physical condition of the individual (e.g., degree of hydration), route of drug ingestion (e.g., oral, intranasal, or intravenous), frequency of use, and drug-related factors such as purity .
To explore the validity of self-reported ATS use among young FSW in Phnom Penh, Cambodia, we compared self-reported ATS use with results from concurrently collected urine toxicology tests. We also examine whether sociodemographic, sex-work venue, and HIV status were associated with validity of self-reported ATS use.