Improving care of vulnerable elders through computerized clinical decision support

This thesis contains the groundwork for the Improving Care of Vulnerable Elders (ICOVE) project, which aims to improve the quality of care for older patients by providing computerized clinical decision support to both clinicians and patients. To this end, we present a general model of information flow in decision support systems that can be used to teach, review, and report the factors that may influence the success of systems in supporting decisions and improving care. We found that clinicians value systems that help them manage their responsibilities and avoid errors, and are concerned about interruption. We addressed the quality of evidence underlying the system by reviewing the coprescription of proton pump inhibitors with NSAIDs, and found that although the evidence is as yet insufficient, it points towards support of the recommended practice of coprescription. To transfer this knowledge to a clinical decision support system, we developed the LERM method for formalizing clinical rules. For informing older patients, we found that the internet appears to be a preferred channel for health information, although further research is needed on how to reach older patients who do not actively seek health information. SnelleCor offers a case study demonstrating that decision support systems can be both well-liked and effective, and despite evidence from recent reviews, interruptive systems are not the only recipe for success. We hope that the studies in this thesis will help guide the development of more useful and usable systems for patients and their clinicians.