Studies of factors in relapse to alcohol, drug and nicotine use: a critical review of methodologies and findings.

OBJECTIVE: To critically review the methodologies used to study relapse in substance abusers and compare some of the findings these methodologies have produced.
METHOD: Representative studies that have used retrospective, prospective, or near real time methodologies to study antecedents or consequences of relapse are presented and the strengths and weaknesses of each methodology are reviewed. Limitations in the ability of substance abusers to accurately report internal states are considered. Degree of convergence in findings across methodologies is examined.
RESULTS: Retrospective reports can provide perhaps the most detailed accounts of circumstances and experiences leading up to relapse, but are also more likely to be inaccurate due to the limitations of memory. Prospective studies are less likely to be affected by memory problems, but may not capture factors that immediately precede relapse. The near real time methodology, which uses palm-top computers to gather data, has many advantages over other methodologies, but a number of limitations as well. The two studies in which retrospective reports were directly compared to either prospective or near real time reports yielded mixed results regarding degree of agreement. Findings from several areas of research suggest that some substance abusers may have difficulty recognizing and reporting internal states.
CONCLUSIONS: Despite memory problems and other limitations of self-report, all three methodologies have produced findings that implicate negative affect states, increased craving, cognitive factors, interpersonal problems, and lack of coping as factors in relapse. Future relapse studies should make use of all three methodologies to obtain data on relapse and consequences, and steps should be taken to enhance the accuracy of the self-reports that are gathered with each method.