A controlled evaluation of diagnostic criteria in the development of a rheumatology expert system.

As an intermediate stage in the development of an expert system to support undergraduate teaching in rheumatology, a decision tree incorporating the diagnostic criteria to be used in the expert system was produced by a team of rheumatologists. In a controlled trial, 119 final-year medical students each diagnosed 10 rheumatology cases, drawn from a pool of 96 cases, with or without the aid of the decision tree. Students who used the decision tree correctly diagnosed the following conditions more frequently than the control group: polymyalgia rheumatica (p less than 0.05), myopathies (p less than 0.01), systemic lupus erythematosus (p less than 0.05), pyrophosphate arthropathy (p less than 0.05), seronegative spondylarthropathies (p less than 0.01), intra-articular bleeding (p less than 0.05) and traumatic synovitis (p less than 0.05). The overall diagnostic accuracy of the students who used the decision tree was 81% compared with 68% for the control group (p less than 0.001).