Using Electronic Medical Records to Measure & Improve Adherence to Tobacco Treatment Guidelines in Primary Care

Although tobacco remains the leading cause of preventable morbidity and mortality in the United States, implementation of proven tobacco-cessation treatments developed specifically for primary care settings has been disappointingly slow.

In response to the implementation problem, a clear set of evidence-based practice guidelines have been developed for primary care providers (the "five As"). These guidelines have been strongly endorsed by health care organizations, many of which have conducted dissemination programs. However, evaluation of guideline implementation, and routine assessment of the quality of tobacco-cessation services in primary care, has been hampered by imprecise and costly patient questionnaire methods currently used for measuring treatment delivery.

The rapidly growing use of electronic medical records (EMRs) offers an attractive alternative for evaluating guideline implementation and quality of care. This study will begin with the development of a method for coding tobacco-cessation activities (the five A's) in four HMOs using EMRs. Data from coded fields (check boxes, standardized diagnosis & treatment codes, and prescriptions) will be taken directly from the EMR, and information entered in free-text fields (e.g. progress notes) will be coded using a natural language processing program. (Such automated methods can be comprehensively applied to large populations at a much lower cost than reviewing a sample of records by hand.)

Data extraction and coding procedures will be standardized across the four participating health care systems, with validity assessed by comparing automated coding to coding by medical record technicians. Next, a random sample of primary care providers will be selected for an effectiveness trial in which half will be provided regular feedback on their adherence to the tobacco control guidelines and the other half will not receive feedback. Feedback will also include information about patient follow-through on treatment recommendations as recorded in the EMR. The effect of this feedback on practice patterns will be assessed over a 2-year follow-up period.

Study results will help to improve methods for monitoring quality of care using EMRs, help improve programs designed to increase adherence to national tobacco treatment guidelines, and provide useful information for EMR developers.

Project Status: This study was funded as part of the CRN Competing Continuation (Ed Wagner, PI). The project began March 1, 2003. The project leader is Vic Stevens, PhD, of the Kaiser Permanente Northwest Center for Health Research.

Return to top