Measuring the use and career histories of drug users in treatment: reliability of the Lifetime Drug Use History (LDUH) and its data yield relative to clinical case notes.

INTRODUCTION AND AIMS There is no generally accepted clinical or research instrument available for recording the longitudinal course of a drug-using 'career'. This paper reports on an initial examination of the properties of the Lifetime Drug Use History Questionnaire (LDUH), built around monthly mapping of drug use patterns in relation to other life events. DESIGN AND METHODS Forty heroin and cocaine users completed structured interviews at two treatment sites. Twenty subjects were interviewed on two occasions separated by a 3-day interval, using either the same interviewer (n = 10) or two different interviewers (n = 10) as assessments of inter-rater and test - retest reliability. RESULTS Very good inter-rater agreements were observed, demonstrated by Cronbach's alpha and intraclass correlation coefficients generally higher than 0.8 and 0.7, respectively. Additionally, concordance with clinical notes was assessed for four drug use history variables, resulting in poorer rates of agreement. An exact matching with clinical records was obtained for the variable 'age of first use of heroin' in 47.2% (n = 17) of the heroin users, while a good agreement (only 1 or 2 years' difference) was found in 36.1% of cases (n = 5). DISCUSSION AND CONCLUSIONS The LDUH method resulted in high reliability for heroin and cocaine and suggests an effective, clinically applicable method for history-taking. The paucity and inconsistency of similar information in the clinical notes would further justify the use of a standardised method for recording drug histories.