A core problem for health information seekers is locating relevant information among the vast amount of online content. Personalization has been suggested as a solution to effective Web information retrieval and also explored in the healthcare context, but then primarily with paper-based or locally available material. In this paper we report on the evaluation of a prototype system, STEPPS, which utilizes electronic patient record data to personalize the retrieval of health information from a variety of online resources, for the purposes of patient education. In a blinded clinician assessment of Web pages retrieved for 27 individual profiles of patients hospitalized in Dutch burn care units, STEPPS performed significantly better than random selection of material (p<0.001, N=270), thereby providing evidence of the 'proof of concept'. STEPPS also performed better than Google keyword searches for the specific patient test cases (p=0.010, N=270). Further research is necessary to optimize system performance, as well as to test its applicability and suitability for other clinical domains, particularly in the context of chronic disease management.

doi.org/10.1109/CBMS.2008.105, hdl.handle.net/1765/52117
21st IEEE International Symposium on Computer-Based Medical Systems, CBMS 2008
Department of Medical Informatics

Doupi, P., van Wijk, M., van Wyk, J., & van der Lei, J. (2008). Personalizing patient education using Internet health resources and EPR data: Pilot evaluation of the STEPPS prototype system. Presented at the 21st IEEE International Symposium on Computer-Based Medical Systems, CBMS 2008. doi:10.1109/CBMS.2008.105