Reliability of self-reported use of amphetamine, barbiturates, benzodiazepines, cannabinoids, cocaine, methadone, and opiates among acutely hospitalized elderly medical patients.

BACKGROUND Undisclosed use of illicit drugs and prescription controlled substances is frequent in some settings. The aim of the present study was to estimate the reliability of self-reported use of amphetamine, barbiturates, benzodiazepines, cannabinoids, cocaine, methadone, and opiates among acutely hospitalized medical patients. METHODS Patients admitted to an acute medical department were interviewed about their drug use. Patients provided blood and urine samples for drug analysis. Results of a toxicology screen were compared to self-reported drug use. Toxicology screens positive for drugs not reported during the interview were only considered truly positive after verification by a substance specific analysis. RESULTS Five hundred patients were included. The median age was 72 years and 298 (60%) were female. In total, 103 patients (21%) reported use of opiates and 65 patients (13%) used benzodiazepines. Only 8 patients reported use of illicit drugs (cannabinoids, 2%). Toxicology analyses were performed in a randomly selected sub-sample of 100 patients. Among 27 patients (27%), the analyses indicated use of one or more drugs, mainly benzodiazepines (15 patients), morphine (12 patients) or cannabinoids (5 patients). Another 6 patients had screenings unexpectedly positive for opiates, but the verification analysis indicated use of codeine-containing drugs. The overall sensitivity of self-reports in detecting drug use was 66%. The negative predictive value of a patient not reporting use of a drug was over 90% for all 7 drug-types screened. CONCLUSION Among 100 randomly selected mainly elderly medical patients, undeclared use of illicit drugs was rare. However, some patients underreported use of benzodiazepines and cannabinoids.