BACKGROUND: The long list of ophthalmologic findings and symptoms at a baseline examination of a patient with uveitis can have an impact on further laboratory workup and differential diagnosis. METHODS: Based on publications and previous work, a decision tree was transformed into the expert system shell D3. A web-based Java Servlet was then programmed and published on our website (http://www.uveitiscenter.de at DiagnoseFinder). Ophthalmologic findings and symptoms of the patient can be checked with this online questionnaire, and recommendations for further laboratory tests and further care of the patient received. RESULTS: Retrospective analysis of 62 patients with characteristic features of secondary uveitis showed a good correlation between the actual diagnosis of the patient and the diagnosis found by the program. In 74% of cases, the correct diagnosis was found among other options. In 37%, only the correct diagnosis was found, while in 37% the correct diagnosis was found among other probable diagnoses. Not all forms of secondary uveitis were found easily by the software. CONCLUSION: This program is meant to be an easily accessible and simply employable help for the care of patients with a secondary form of uveitis. It is intended to be a support not only for residents in particular, but also for ophthalmologists in private practice who rarely treat patients with uveitis.
BACKGROUND: The long list of ophthalmologic findings and symptoms at a baseline examination of a patient with uveitis can have an impact on further laboratory workup and differential diagnosis. METHODS: Based on publications and previous work, a decision tree was transformed into the expert system shell D3. A web-based Java Servlet was then programmed and published on our website (http://www.uveitiscenter.de at DiagnoseFinder). Ophthalmologic findings and symptoms of the patient can be checked with this online questionnaire, and recommendations for further laboratory tests and further care of the patient received. RESULTS: Retrospective analysis of 62 patients with characteristic features of secondary uveitis showed a good correlation between the actual diagnosis of the patient and the diagnosis found by the program. In 74% of cases, the correct diagnosis was found among other options. In 37%, only the correct diagnosis was found, while in 37% the correct diagnosis was found among other probable diagnoses. Not all forms of secondary uveitis were found easily by the software. CONCLUSION: This program is meant to be an easily accessible and simply employable help for the care of patients with a secondary form of uveitis. It is intended to be a support not only for residents in particular, but also for ophthalmologists in private practice who rarely treat patients with uveitis.