Screening and assessment for unhealthy substance use offers clinicians the opportunity to identify harmful substance use or disorders and provides the opportunity to address such use. However, few studies have reported on screening for substance use as part of HIV primary care. The goal of our study was to characterize patterns and severity of substance use through two different screening and assessment approaches in a large, urban public HIV clinic providing primary care to PLWH and to describe gender and racial differences in alcohol, tobacco, and other substance use. Descriptive statistics, including frequencies, percentages, means, and standard deviations were performed to characterize the sample. Pearson’s chi-square and Fischer’s exact test analyses were used to determine differences in risk stratification of SSIS scores between male and female genders, and across the four race/ethnicity categories and substances detected via urine toxicology. All analyses were conducted with STATA 14. Urine toxicology screens provides information that will tell you if a substance is present in the urine or not, however, urine tests cannot be used for diagnosis of a substance use disorder. We enrolled 208 HIV primary care clinic patients from an urban public clinic. The analysis presented here is based on the 168 participants who completed both the ASSIST questionnaire and urine drug screening procedure. The participants were primarily male ; and more than one third were African American . There were no significant demographic differences between the entire sample of 208 and the analytic sample of 168. The average age was 45.66 years with an average of 12.40 years living with HIV. The majority of the participants had an undetectable HIV viral load . More than two thirds of the study sample reported using tobacco or other non-prescribed substances in the previous 3 months. Forty-one percent of our participants reported alcohol use for the same time period.
As described in the Methods section, we determined Single Substance Involvement Scores for each substance reported and stratified these scores into low , moderate , and high risk for all substances except alcohol ,flood table following the validated ASSIST scoring guidelines. More than half of our participants’ SSIS scores indicated moderate risk for tobacco, and cannabis . The three drug classes with the greatest number of participants exhibiting high-risk scores were for tobacco , cocaine , and amphetamine . The SSIS for alcohol use indicated that over one-third of study participants reported a moderate risk level for alcohol and 7.7% had a high-risk score for alcohol use. When comparing the SSIS score for each substance by gender and race, we observed differences in reported substance use. Compared to females, males in this sample reported greater levels of moderate risk cannabis use and moderate risk amphetamine use . There were also significant differences for cocaine use with Hispanic or Latino/a participants reporting lower risk use than African American, White/Anglo, or Other race participants . Finally, more African American participants reported low or no risk amphetamine use as compared to Hispanic or Latino/a, White/Anglo, or Other race participants . More than half of the sample submitted urine specimens that tested positive for cannabis , nearly one third tested positive for cocaine, and almost a quarter tested positive for benzodiazepines. Significant gender differences in urine toxicology were also present . Male gender was significantly associated with positive urine toxicology for amphetamine and methamphetamine . Female gender was significantly associated with positive urine toxicology for cocaine , methadone , and opiates . Significant racial differences were also observed in urinetoxicology. Those of Other race or ethnicity screened positive for cannabis use more frequently . Both Hispanic or Latino/a participants and White/Anglo participants screened positive for cocaine less frequently. African American race or ethnicity was associated with lower levels of positive urine toxicology for both amphetamine and methamphetamine . In this study of patients in an HIV primary care clinic-based urban population, we found high rates of self-reported substance use, which were confirmed by urine toxicology testing. The SSIS risk scores for all substances, excepting inhalants and hallucinogens, demonstrated that moderate and high-risk substance use was highly prevalent in this sample of patients.
Reported substance use in this HIV clinic sample was higher than in other studies of both HIV and non-HIV primary care patient samples for most substances reported except for tobacco use. In the United States, approximately 19% of the adult population smokes cigarettes . When compared to the U.S. general population, a number of studies have documented considerably higher rates of smoking in PLWH , which is of grave concern given the now well-documented increased mortality associated with smoking among PLWH due to cardiovascular disease and non-AIDS related cancers . For other substances such as cannabis, our sample exhibited levels of use similar to other primary care settings where the ASSIST measure was used. However, in another study of an HIV clinic-based sample, the reported use of cannabis was 18% , which was considerably lower than what we found in our study. When examining other substances reported by participants in our study, we saw similarities compared to other clinic samples of HIV-infected and uninfected patients, for example with stimulant use . A large number of participants in our sample reported moderate or higher ASSIST scores for cocaine and amphetamine-type stimulants . There have been a multitude of studies on stimulant use and HIV, ranging from stimulants as a risk factor for HIV transmission and as a method of managing mental health symptoms and the experience of discrimination, to the manner in which they impacted adherence to ART; however, very few of these samples were drawn solely from clinic settings where HIV care was delivered. In the studies that have been conducted in HIV primary care settings, a range of stimulant use has been reported. Skeer et al. studied HIV-infected men who have sex with men in a large primary care setting in Boston, MA. and reported that 21% of their sample used amphetamines. In an earlier study of a nationally representative probability sample of PLWH, 40% of the subjects reported using an illicit drug other than cannabis. In a more recent study of the Women’s Interagency HIV Study, investigators did not solely recruit samples from HIV primary clinics; however, nearly one third of the HIV infected women in the sample reported crack cocaine use within the previous 3 months . The participants in our study also reported a high prevalence of moderate-severe SSIS for alcohol . In comparison, the 2013 National Survey on Drug Use and Health determined the national rate of alcohol use disorders was 7% . While in studies conducted in general outpatient settings site the prevalence of unhealthy alcohol use ranging from 7 and 20% . The methods used in these studies vary, however, the prevalence of alcohol use in general medical settings is much lower than what we measured in this sample.
Alcohol like other substance use can complicate HIV care and treatment outcomes and continues to be a major driver of HIV acquisition. Substance use patterns can differ between women and men. In the literature, many studies of HIV and substance use conducted with MSM have focused on alcohol or amphetamine use , while studies of HIV-infected women have been more focused on crack cocaine and heroin use . In our study, we observed gender differences in SSIS scores and in urine toxicology results. Males in our sample had a significantly higher proportion of moderate or high-risk SSIS scores for amphetamine and for cannabis , while women had significantly higher levels of cocaine, methadone, and opiate positive urines when compared to men . This differed from what we observed in the self-report SSIS scores. While women were marginally more likely than men to report moderate or high-risk cocaine use, this difference was not statistically significant. Many studies in the HIV literature have focused on men, MSM, or women and substance use. To our knowledge, however, no studies analyzed gender differences between men and women in an HIV-infected sample. One more general study found that women were more likely to have a substance use disorder combined with other mental illness compared to men; however, there were no gender differences in the presence of a substance use disorder in the absence of mental illness . Urine toxicology in our study looked different from self-report responses using the ASSIST. Urine drug screening is limited to the detection of drug use within a few days before the test and, as in most tests,indoor plant table false positives and false negatives as well as technical problems can occur. Although objective, the use of biomarkers is not without limitation. The literature has indicated that, in some persons who use drugs, self-report, when compared to urine toxicology verifies under reporting of illicit substance use, although it is not known how widespread this is. Also, some clinicians may conduct urine screening as evidence of therapeutic adherence and evidence of use or non-use of illicit drugs. . In our sample, women had more methadone and opiates in their urine when compared with men; however, opiates and methadone are both commonly prescribed in medical settings for both pain management and opiate agonist therapy and we did not systematically ask participants if they were being prescribed opiates. As reported by Robinson-Papp, Elliott, Simpson, and Morgello , singular reliance on self-reports for implementation of substance use screening and brief interventions has limitations. In addition, more stigmatized drugs, such as cocaine, methamphetamine, or heroin, may be under-reported using self-report but could be documented with urine toxicology tests . In this study participants were paid for urine testing which might not happen in a primary care setting so motivation to provide a urine sample may be different. While we are not advocating urine screening as the initial step for screening in a clinical setting, some clinicians may use it as a tool to work with patients with a history of substance use to validate their reported use and not as a test, which could penalize the patient . Although substance use levels differed by screening modality in our study, the evidence clearly pointed to high levels of substance use in this HIV clinic sample.
High amounts of reported substance use found in our study and others highlights a critical problem that HIV clinicians may be overlooking and that could be addressed by universal substance use screening. Based on the evidence of efficacy for screening and offering a brief intervention for alcohol and tobacco use, the U.S. Preventive Services Task Force has recommended universal preventive substance use screening in primary carefor adolescents and adults . While screening and brief intervention has shown promise for harmful alcohol use and smoking , the efficacy of universal BI for illicit drug use and prescription drug misuse has not been universally recommended for primary care settings . However, because of the overwhelming evidence that illicit drug use negatively impacts health, research to determine the efficacy of screening and brief intervention for drug use is ongoing. SBIRT has emerged as an important model for identifying and addressing substance use problems in health care settings . Brief intervention approaches are typically delivered on site, and individuals with more severe substance use problems also may be offered referrals to specialized treatment. Brief intervention for non-treatment seeking samples has strong support in the alcohol literature and some promising effects have been observed with respect to other substance use . Substance use screening followed by a brief intervention conducted by an individual trained in motivational interviewing has been extensively examined in adolescents and young adults using drugs and alcohol. These studies have revealed significant reductions in marijuana use ; decreases in alcohol use, binge drinking, and days of drug use ; lower alcohol, tobacco, and cannabis use ; and reductions in illicit drug use . To our knowledge, few studies of SBIRT have been conducted in HIV settings. Cropsey et al. conducted an SBIRT feasibility and acceptability study in an HIV primary care clinic to address the high rates of smoking by PLWH; the findings of Cropsey’s study indicated that SBIRT was feasible and acceptable to staff and patients in the HIV primary care setting. Using SBIRT as an approach for SBI was feasible and acceptable for many participants in our study .