Page 51 - PhD GT

some debate (for example see Shaw 1979) it was shown to have explanatory and predictive utility. Describing dependence as the consequence of operant learning wherein morning after
withdrawal symptoms are alleviated by further alcohol consumption, the two item scale with three possible response choices described above was used by Edwards et al. (1971) to demonstrate a relationship between dependence and drinking behaviour, offences of drunkenness, and violent offences in a sample of prisoners. The questions asked refer to lifetime experience of withdrawal tremor and relief drinking: ‘Have you ever woken up with your hands very shaky?’ and ‘Have you ever taken a morning drink to steady yourself after a hard night’s drinking?’ Responses were awarded one point for ‘once or twice’ and two points for ‘quite often’. Responses were divided into categories of no dependence (0), moderate dependence (1-2) and severe dependence (3-4). With this early attempt at measuring dependence (Edwards et al. 1971 p. 395), these authors asserted the importance of distinguishing degrees of dependence by demonstrating that relationships with other factors such as drinking behaviour and social consequences varied not only with the presence or the absence of dependence but also with the degree of the dependence.
Drummond (1990) demonstrated that while dependence as measured by the Severity of Alcohol Dependence Questionnaire (SADQ) (Stockwell et al. 1979) was shown to correlate highly with both measures of consumption and of alcohol related problems, a highly significant relationship between dependence and problems was also shown which was independent of the quantity of alcohol consumed. Due to the relative ease of measuring dependence compared to gaining an accurate measure of consumption Raistrick et al. (1994) argued that it may be preferable for use in the busy and routine clinical setting. An interesting question would be whether the measure of dependence would constitute a good proxy measure, as Raistrick et al. (1994 p.570) argued, or whether the measurement of dependence would in fact improve on the measurement of consumption in its ability to measure the presence and degree of a condition that is shown to persist during a period of abstinence and has been shown to predict relapse (see below, next section). Since the business of the addiction clinic is to target the problems caused by substance misuse and to minimise or to prevent relapse, it could be argued that it is expedient on both counts to have a measure of dependence in routine clinical evaluation.
2.5 Predictive validity of dependence
A series of studies conducted at the Maudsley Hospital during the nineteen-seventies provided evidence that degree of dependence could predict the speed of drinking and the quantity of
39