The evaluation project tested two strategies to change provider behavior and improve patient outcomes: financial incentives, and the availability of a registry and proactive telephonic support for smoking cessation.

Forty primary care clinics in a large medical group were randomly assigned to one of three arms. One arm received a financial incentive for achieving a predetermined rate of documentation of tobacco use status and advice to quit. A second arm received the same financial incentive and also had access to a centralized patient registry and proactive telephonic support for smoking cessation. The third arm served as the control. The primary outcome measures targeting providers were rates of tobacco use identification and rates of providing advice to quit. Patient outcome measures were sustained quit attempts, intention to quit within 30 days, and use of any medication for cessation. Data were collected through a series of patient surveys and chart abstraction.

Rates of tobacco user identification, provision of advice to quit, and assistance in quitting improved in the experimental conditions as compared to the control condition. Statistical significance was achieved in the rates of tobacco user identification between experimental conditions (14.1% improvement in the financial incentive arm compared to 8.1% improvement in the registry arm, p=.009).

Quit rates (self-reported 7-day point prevalence) ranged from 19.2% (control) to 22.4% (incentive) and did not differ statistically significantly between experimental conditions. Overall, 10.8% of respondents reported intention to quit within 30 days (range 9.4%-12.8%), 20.8% reported using any medication for smoking cessation (range 19.8%-21.6%), and 1.6% (range 1%-3.3%) reported use of counseling. The only rate that differed statistically significantly between the two experimental arms was the use of counseling, as patients in the registry arm reported using counseling more often (p=.001).

The study found promising but modest support for the use of financial incentives as structured in this study to improve clinical practice. Possible explanations for the outcomes seen include significant cost pressures in the health care marketplace at the time of the study and possible lack of provider knowledge of the incentive program. It should also be noted that a secular trend toward improved practice patterns was noted across experimental conditions during the study. Possible explanations of this trend include a potential Hawthorne effect on all clinics due to the establishment of practice monitoring systems and exit surveys as part of the study protocol, and the state of Minnesota’s settlement with the tobacco industry raising both patients’ and providers’ awareness of the harms of tobacco use and the importance of smoking cessation.