A web-based program was shown to assist clinicians in making a differential diagnosis for knee pain, according to study data.
Researchers evaluated the program’s ability to collect pain history from 527 consecutive ambulatory patients with knee pain and make a differential diagnosis for them through the use of an internet questionnaire during a 3-month period. The program made its diagnosis by selecting from 26 common diagnoses relating to knee pain.
Following completion of the questionnaire, the patients were examined by an orthopedic surgeon who generated his own differential diagnosis entirely independent from that generated by the program. The diagnoses from surgeon and program were compared via a third party to independently determine the program’s accuracy.
The program generated a median of 4.8 diagnoses per patient. Overall, the surgeon’s diagnosis was one of those generated by the program 89% of the time with a 27% specificity, according to the researchers.
Disclosure: This study was funded by the Ralph and Mary Wilson Fund.