Page 55 - PhD GT

withdrawal symptoms) were rated using a fifteen item subset of the 62 item Last Six Months of Drinking Questionnaire (Hesselbrock et al. 1983). The authors report their reliability analysis for this 15 item set which was used to measure recent dependence symptoms (last six months). A separate measure was used to measure lifetime dependence symptoms, namely the alcohol dependence / abuse section of the National Institute of Mental Health Diagnostic Interview Schedule (Robins et al. 1981). Three measures of reinstatement of the syndrome were used to test their hypothesis that degree of dependence at admission to treatment predicted reinstatement of the syndrome at twelve month follow-up where alcoholic patients had attempted to drink following a period of abstinence. Among these was a repeat of the measure of dependence described above, a measure of the amount of time between first drink and daily drinking and the number of hospital treatment episodes for alcohol problems since the index treatment in alcoholics but not in opiate addicts. Rapidity of reinstatement as measured by dependence at follow-up was predicted by lifetime and recent dependence at intake for both males and females, but when reinstatement was measured using time between first and daily drinking and number of hospital treatments, a gender difference emerged. These measures did predict rapidity of reinstatement for the male group but not for the female group. They also found that severity of recent dependence at intake predicted intensity of craving for abstinent males and females at follow-up.
Different measures of dependence and of reinstatement were used to examine the hypothesis in opiate users. Drug and alcohol severity composites of the Addiction Severity Index (McLellan et al. 1980) were used as “provisional surrogates of dependence syndrome measures” (Babor et al. 1987a p. 401) and reinstatement was measured by a dichotomous measure of return to daily heroin use and by the amount of time between first use of heroin following a period of abstinence and daily heroin use. Drug severity at intake predicted severity at follow-up for only one of the three samples investigated but none of the indicators of prior dependence at intake predicted reinstatement of the syndrome following a period of abstinence.
The sheer diversity of the instruments used to measure dependence and other behaviours and outcomes with which it is compared render the drawing of conclusions difficult. This appears as much the case now as it was in the mid-eighties when Duckitt et al. (1985) proposed that there was emerging evidence that dependence was an important mediating variable between different outcome measures (level of consumption, social adjustment, mental health) but that “imperfect and post hoc measures of dependence would be insufficient for establishing these relationships with any certainty” (Duckitt et al. 1985 p. 161).
Noting the variety of instruments used to measure dependence, Edwards (1986 p. 181) 43