Objective. Test whether physicians' counseling patients for smoking cessation with an autonomy supportive rather than controlling style would increase patients' active involvement in the counseling session and increase maintained abstinence.
Design. Randomized trial of 27 community-based physicians using two interview styles, with observer ratings of patient active involvement and assessments of patient smoking status at 6 months, 12 months, and 30 months.
Patients. Adult smokers: 336 recruited; 249 for final analyses.
Intervention. Physicians used an autonomy- supportive or controlling interpersonal style, randomly assigned within physician, to briefly counsel patients about smoking cessation, using the National Cancer Institute's 4-A's model.
Measurement. Patient active involvement was rated from audio tapes of the interviews. Continuous abstinence came from self-reports at 6 months, 12 months, and 30 months, CO validated at 6 months or 12 months and at 30 months.
Results. Physician style did not have a significant direct effect on smoking cessation but did significantly increase patient active involvement in the interview. Active involvement, in turn, increased smoking cessation. Structural equation modeling confirmed a theoretical model in which the intervention positively predicted patient active involvement after controlling for patient reports of wanting to stop smoking, and active involvement significantly predicted continuous abstinence after controlling for previous quit attempts.
Conclusions. Although physicians' autonomy- supportive style while counseling smokers to quit did not have a direct effect on smoking cessation, it increased patients' active involvement in the counseling session which in turn increased continuous abstinence over 30 months. Further research should clarify the direct effects of physician interpersonal style on health outcomes.