Information seeking features of a PICO-based medical question-answering system

A survey was conducted among 20 health care providers to evaluate the information seeking features in a prototype question-answering system called CliniCluster. The respondents were required to complete a 16-item questionnaire while examining 25 therapy questions processed and answered by CliniCluster. The respondents were instructed to rate the usability of two seeking features in the system user interface, and the satisfaction gained by them during the search of relevant documents. The majority of the respondents agreed that the hierarchy of medical interventions and the P-O and I/C elements in the answer field assist them in narrowing down the search results and in identifying relevant documents. Respondents with higher previous knowledge on the 25 questions gained a higher satisfaction, when compared to those with moderate previous knowledge. The respondents gained new knowledge and better understanding of the therapy questions, and are satisfied with the ease of completing the search tasks using the features provided by the system. To conclude, the overall results support the use of CliniCluster for a more efficient search of clinical evidence.