INDIVIDUALIZED FEEDBACK TO IMPLEMENT CLINICAL GUIDELINES

Compliance with accepted practice guidelines is an important quality issue in health care and quality management programs often provide feedback information. However, the use of guidelines to develop quality improvement programs remains difficult; individual differences in information needs and appropriate actions are not recognized; and results of the few clinical trials testing computerized feedback intervention are controversial. The aim of this project is to assist health care providers in translating accepted guidelines and practice variation data into measurable individual quality improvement objectives and to assess the clinical effect of this intervention on clinical practice patterns in diverse areas of primary care. In a randomized controlled clinical trial, an individualized quality feedback intervention will be tested. The feedback will be administered by a rule-based expert system supporting a) the specification of corporate and individual quality improvement objectives based on the recommended guidelines and b) the follow-up of previously set objectives by highlighting discrepancies between individual objectives and actual- practice patterns. The sites of the trial will be the outpatient centers of Humana Health Care Plans, a health maintenance organization in Kansas City. The new individualized quality feedback method will be used to implement various guidelines distributed by AHCPR (HIV positive asymptomatic patients, depressed outpatients) and developed by Humana (antibiotic prescription, management of diabetes requiring medication). Forty three physicians and 13 nurse practitioners will be randomly assigned to participate in the new quality management program or to receive usual feedback information on quality. Balanced stratification will occur by provider site. Patients will be followed up six months after their first visit which makes them eligible for the care recommended by the guidelines. Observed major variables will include the number of identified eligible visits, frequency of recommended procedures ordered during eligible visits, specialist referrals, and use of not recommended procedures. A sample size of the eligible patients is estimated to be adequate to test the primary hypothesis. The new quality management intervention and the results of our controlled trial should have practical applicability in a wide variety of actively managed health care organizations (e.g., HMO, PPO, IPA).