Abstract
BACKGROUND: There is a high prevalence of psychoactive substance use among patients with mental health disorders. The optimal treatment of patients with mental health disorders requires an awareness of their history pertaining substance use. Several methods are used to assess the use of substance. Each of them embodies its limitations. This study aimed at assessing the diagnostic capability of a self-report psychoactive substance use among patients at the National Psychiatric University Hospital of Cotonou, Benin. METHODS: A cross-sectional survey was conducted from August 1, 2021 to November 24, 2021. A total of 157 consenting patients admitted to psychiatric consultations were successively enrolled in the ongoing study. They were screened for the use of psychoactive substance with Alcohol, Smoking and Substance Involvement Screening Test (ASSIST), followed by urine test using the NarcoCheck® kit for qualitative detection of substances or its metabolites. To assess the diagnostic capability, the participants' self-responses were compared with their urine test results. The sensitivity, specificity, positive and negative predictive values, and kappa coefficient were also calculated. RESULTS: The frequency of lifetime psychoactive substance use according to self-report was 81.5% (95% CI: 0.746-0.873), while over the past three months (recent use) was 52.2% (95% CI: 0.441-0.603) and 58.6% based on the urine test. Alcohol, tobacco and cannabis were the most prevalent psychoactive substance used. The overall concordance between self-reported psychoactive substance use and the urine test (gold standard) was moderate (sensitivity = 66%; kappa = 0.46). Self-report cocaine use compared with urine test showed the highest concordance (sensitivity = 100%; kappa = 79%), followed by tobacco (sensitivity = 58%, kappa = 41%). On an average 70% of urine test results were consistent with self-report (VPP). Participants' were more accurate when they were reporting no psychoactive substance use as suggested by the high negative predictive value (NPV). CONCLUSION: Diagnostic capability of self-reporting of psychoactive substance use among patients admitted to psychiatric consultations was moderate. Therefore self-reporting may not estimate the exact prevalence of psychoactive substance use. Optimal identification of psychoactive substances use in psychiatric patients requires both history and urine testing. The integration of these two approaches is an excellent method to find out the level, frequency and nature of drug used.