Improving Quality With Outpatient Decision Support

The evidence base for practicing medicine continues to improve. However, abundant data show that gaps exist between best evidence and practice. Moreover, health care costs are climbing at an alarming rate. We propose to ask three related questions: 1) how effective are computer decision-support systems for improving compliance with evidence-based guidelines and costs in the ambulatory setting; 2) what is the impact on guideline compliance of applications that allow clinicians to track and follow-up test results; and 3) what are the main barriers to acceptance of guidelines delivered via real-time clinical decision-support systems. Our work and that of others has shown that computerized decision-support in the form of alerts and reminders can improve outcomes and reduce costs in the inpatient setting. However, fewer data are available in the outpatient setting. An elegant series of studies from Regenstrief found that certain computer-based interventions, such as displaying charges for tests, prior test results, and the likelihood that a particular test would be abnormal, all reduced outpatient utilization, and that reminders to perform health maintenance procedures improved compliance. However, such systems are still not used broadly and the full potential of computer-based technology remains to be tested. Also, there is ample evidence that physicians do not always act optimally on the results of patient studies and often are remiss at communicating satisfactorily with patients about the results of these studies. This situation may be exacerbated by increasing patient volumes in the face of managed care. The ability of the computer to assist in the tracking and follow-up of test results as well as communication with patients remains to be evaluated. Even though some benefits of computer-based decision-support systems have been documented, such systems are slow to be adopted. Moreover, even when computerized guidelines have resulted in demonstrable improvements, often this improvement has been smaller than anticipated. This proposal aims to better understand the barriers to guideline acceptance so that the benefits of computer based decision-support can be realized. Our organization, Brigham and Women's Hospital, is in a particularly good position to study these issues. We have in place a highly developed clinical information system including an outpatient electronic medical records (EMR) application that has been an active part of the clinical workflow since 1999. The EMR application currently is used by primary care physicians at one of our major medical centers to track their patients' problems, medications, allergies, and health maintenance data. We are developing a new EMR that will be used more broadly across our network, and that features a new interface with added functionality. The new EMR will allow us to evaluate the state of the patient at the time of the visit and generate reminders if the patient is out of compliance for certain guidelines. It also includes outpatient order entry that allows physicians to enter medication and laboratory orders directly into the computer. Decision-support in order entry will allow us to guide physician decision making at the most opportune time, and then evaluate the result of that guidance. For automated decision-support applications to be widely adopted, it is critical that their benefits be demonstrated in a wide variety of situations. We plan to implement several different types of interventions targeted at various phases of the clinical workflow to determine which strategies can achieve the greatest benefit. Specific Aims: 1. To evaluate the effectiveness of paper-based and interactive computer-based alerts and reminders for improving compliance with guidelines and reducing costs in the ambulatory setting. 2. To evaluate the impact of computer-based tracking and follow-up reminder systems on guideline compliance. 3. To identify and address patient, clinician, and system barriers to the effective use of computer-based clinical decision-support strategies in a diverse array of clinical settings.