The management of relapse has become a central theme in research concerning the addictive behaviours. However, this attention is relatively recent and until the late 1970's relapse was rarely, if ever, mentioned in the literature and there were few major attempts to develop techniques to investigate and respond to this phenomenon. The main focus of recent studies and models of relapse has been on identifying relapse precipitants and equipping clients with appropriate skills. Such a focus either ignores or downplays the importance of the individual's decision making processes: both the quality of the initial decision to change and the role of decision making in any return to addiction behaviour are essential ingredients in understanding the relapse process. The current study involved reviewing the relevant literature on treatment and treatment outcome, the process of giving up addictive behaviour and the process of relapse.
A number of models of the relapse process have been developed and these were critically evaluated. On the basis of these reviews a new model of the process of giving up and relapse was proposed. This model involved four stages: Resolution, Commitment, Action and Maintenance. In addition to the four stages, it was proposed that a number of background factors, drug effects and individual characteristics, such as degree of alcohol dependence, cognitive functioning and self-efficacy had influence on the process of giving up and relapse. The implications of this model for intervention were explored. The model was then employed in the development of a treatment program for problem drinkers. The next stage of the investigation had two phases. The first phase involved assessing the impact of the experimental relapse prevention (RP) program compared to a discussion and no-additional treatment control group.
The second phase involved assessing the influence of treatment, post-treatment self-effficacy, alcohol dependence, cognitive functioning, level of depression and level of pre-treatment drinking on the relapse process. It was found that the experimental RP group was associated with significantly greater increases in self efficacy compared to the discussion control but not the no-additional control group. Survival analyses indicated that the experimental RP group was associated with longer time to lapse (Log-Rank Chi Square=6.92, df=2, p<0.04) and relapse (Log Rank Chi-Square=7.34, df=2, p<0.03) than the controls and a significantly greater probability of being in the category “good outcome” at six month follow-up (Log Regression Chi-Square=5.83, df=2, p<0.02). Significant differences were not evident at twelve month follow-up.
The factors found to contribute to the relapse process were treatment group allocation, post-treatment self-efficacy rating and visuo-perceptive functioning. Those who had higher post-treatment self-efficacy scores took longer to lapse (survival analysis chi square=5.99, df=l, p<0.015) while lower Digit Symbol Test scores were associated with more rapid lapse (survival analysis chi-square=4.8, df=1, p<0.03). Higher post-treatment self-efficacy scores were associated with lower hazard rates for relapse (survival analysis chi-square=8.8, df= 1, p c 0.003). Higher post-treatment self-efficacy scores were also associated with longer time intervals between an initial lapse and relapse (survival analysis chi-square=4.5, df=1, p<0.04). The implications of these results for the model of giving up and relapse and intervening in drinking problems were explored.