SOMWeb : Supporting a Distributed Clinical Community of Practice Using Semantic Web Technologies

This thesis concerns supporting the collaboration and knowledge sharing of distributed clinicians of oral medicine, a sub-discipline of dentistry. The Swedish Oral Medicine Network (SOMNet) holds monthly telephone conferences where a group of clinicians discuss interesting and difficult cases, which distinguishes it from one-to-one teleconsultations. SOMNet can be seen as a distributed community of practice, that is, a group of people sharing a concern and who interact regularly to extend their individual and collective expertise. Related to this, several topics need further investigation: How can geographically distributed clinical collaboration be characterized? What is appropriate functionality for a Web-based system supporting such collaborations? What are the impacts of such systems on collaboration? Further, Semantic Web technologies, such as the Web Ontology Language (OWL), have been proposed as a means of enhancing knowledge sharing. What are benefits and limitations of using these technologies to encode domain knowledge in oral medicine and to support clinical collaboration, and what practical issues face developers? The developed system, SOMWeb, focuses on functionality for meetings and structured cases, and has been regularly used for three years. Interviews, observations, a questionnaire, system log analysis, and case analysis were used to study SOMNet's collaboration and identify system impacts. The documentation of the forms of collaboration in SOMNet can serve as a model for other groups of clinicians wishing to establish a distributed collaboration. SOMNet's meetings provide a necessary rhythm for the community and the cases give context to the clinicians' learning which point toward that the centrality of meetings and cases in a tool will benefit collaboration. Impacts on SOMNet's collaboration include enabling the participation of a wider range of clinics. Factors influencing this are the more accessible submission process as well as the increased tangibility of the collaboration. The thesis also provides recommendations for developers of systems supporting clinical collaboration and knowledge sharing. The use of OWL in examination descriptions has enabled reasoning over cases in the system to provide improved case browsing. At the same time, limitations were found in using OWL for examination templates. Based on the lessons learned in this development, the thesis provides recommendations for using Semantic Web technologies, which can be of value for other developers and to guide future research.