The Reliability of Reports of Recent Psychoactive Substance Use at the Time of Admission to an Acute Mental Health Unit

Abstract

The aim of this study was to examine the accuracy of patients’ accounts of their psychoactive substance use in the week before admission to an acute mental health unit. Fifty consecutively admitted patients undertook a semistructured clinician-administered questionnaire for recent substance use. The results of the interview were compared to the results of gas chromatography mass spectrometry/ mass spectrometry. No patient refused to participate, 46 patients were able to complete the interview, and 48 patients provided a blood sample. Six patients had unreported cocaine or amphetamine in their blood. Cannabis and opiates were not detected in the blood of most of the patients who reported using these drugs. The self-reports of the use of prescription sedatives were inaccurate, but self-reports of recent use of caffeine, nicotine, antidepressants, antipsychotics, and carbamazepine had good overall reliability. Patient reports of recent illicit substance use at the point of admission were found to be unreliable, which may result in incorrect diagnosis and suboptimal treatment of both mental illness and substance-related disorders.

KEYWORDS:

• Mass spectrometry
• substance abuse
• stimulants
• cannabis
• reliability
• mental disorder
• psychosis
• schizophrenia

The research was supported by a grant in aid from the Royal Australian and New Zealand College of Psychiatrists. The authors would also like to thank the Department of Pharmacy, University of Sydney, for allowing access to the TSQ46 GCMS/MS instrumentation and associated laboratories and the medical officers who performed the initial assessment and took blood samples.

Notes

^† N = 48; 2 patients refused blood collection. Four patients who refused the drug interview were assumed to have denied substance abuse.

∗Kappa scores: 0 to 0.2, poor reliability; 0.2 to 0.4, fair reliability; 0.4 to 0.6, moderate; 0.6 to 0.8, good; 0.8 to 1.0, very good (Landis & Koch, 1977).

^§All patients reported using at least one substance, and every patient had at least one substance detected in their blood using GCMS/MS.