The focus and conduct of clinical trials

This paper discusses some issues regarding how putative stop-smoking treatments should be evaluated. Preliminary evaluations of treatments have used withdrawal symptoms as a surrogate for abstinence in an attempt to decide those treatments on which to carry out expensive full-scale clinical trials. However, withdrawal symptom severity has not been shown to relate to long-term abstinence in any prospective studies. Until a better understanding of the relationship between withdrawal symptoms and abstinence is obtained withdrawal symptoms cannot be considered an appropriate surrogate for abstinence. Research is needed into what pattern of short-term abstinence provides the best predictor of long-term abstinence. Evidence from carbon monoxide validation of short-term abstinence raises doubts about whether minor lapses back to smoking are likely to lead to full blown relapse. Further research into the use of expired-air CO in validation of claims of abstinence is required. Finally, the traditional clinical trials format is not well suited to evaluation of treatment programmes with a large behavioural component. An alternative based on the principles underlying biological evolution is proposed.