Drs. Baker and Lewin Reply

To The Editor: Drs. Lan, Chiu, Wu, Hung, and Hu raise the important issue of the role of stage of change in adherence to treatment and treatment outcome in our smoking cessation trial, although evidence in support of the stage of change model is generally lacking (1) . In our analysis of the baseline characteristics of our cohort (2), we reported that precontemplators and contemplators had significantly fewer lifetime attempts to quit compared with those at the preparation stage of change. Among the 147 smokers with a psychotic disorder who were assigned to the treatment group, there were 14 precontemplators, 73 contemplators, and 60 at the preparation stage. On average, those at the precontemplation (mean=5.79 sessions) and contemplation (mean=5.74 sessions) stages attended one less treatment session than those at the preparation stage (mean=6.80 sessions) (F=4.15, df=2,144, p=0.02).

One of the key outcomes in our article was the smoking reduction status (i.e., whether or not participants had reduced their daily consumption of cigarettes by 50% or more, including abstinence, relative to baseline). Baseline stage of change was not significantly associated with smoking reduction status at 3, 6, or 12 months (for the cohort as a whole, nor among those receiving treatment). For example, among the treatment group, 28.6% of precontemplators, 31.5% of contemplators, and 31.7% of those at the preparation stage met our smoking reduction criterion at 12 months (χ ² =0.05, df=2, not significant). Moreover, all of the associations between treatment status and smoking reduction status reported in Table 1 of our article remained statistically significant after controlling for baseline stage of change.

Although nicotine replacement therapy was only provided to participants during the 10-week intervention period, we assessed self-reported use of nicotine replacement therapy between the 6- and 12-month follow-up interviews (N=272). Likelihood of using nicotine replacement therapy during this period was associated with baseline stage of change (precontemplators, 12.1%; contemplators, 24.8%; preparation, 38.2% [χ ² =10.04, df=2, p<0.01]). Additionally, there was no significant association between nicotine replacement therapy use during this period and intervention status (comparison group, 24.1%; attended fewer than five sessions, 31.3%; attended five to seven sessions, 35.9%; attended all sessions, 30.9% [χ ² =2.65, df=3, not significant]).

In summary, among our cohort of smokers with a psychotic disorder, we found the stage of change to be associated with the number of previous attempts to quit, the number of treatment sessions attended, and the subsequent use of nicotine replacement therapy, but not smoking reduction status. Thus, in our study, the stage of change does not account for the reduction in smoking reported, suggesting that the intervention (consisting of nicotine replacement therapy, motivational interviewing, and CBT) led to these changes. We propose that strategies to enhance engagement in treatment and the identification and delivery of effective treatments appear to be separate but important issues that need to be considered in parallel. Stage of change for smoking reduction may be a useful index of treatment preparedness, but it should not be the primary basis for initiating or changing treatment.