O’Neil JT, et al. Application of computerized adaptive testing to the foot & ankle ability measure. Presented at: American Orthopaedic Foot & Ankle Society Annual Meeting; Sept. 10-12, 2020 (virtual meeting).
A computerized adaptive testing system of the foot and ankle ability measure may provide a more accurate and efficient assessment of patient-reported outcome measures for foot and ankle pathology, according to recently presented research.
“Patient-reported outcome measures (PROMs) are becoming an increasingly important tool for orthopedic surgeons, as they assess the impact of the treatments we render for patients,” Joseph T. O’Neil, MD, a foot and ankle specialist at the Rothman Orthopaedic Institute, said at the virtual American Orthopaedic Foot & Ankle Society Annual Meeting. “These measures can create a burden for both patient and physician.”
For patients, PROMs can be time consuming and arduous, ultimately leading to poor compliance and less accuracy, O’Neil said. For physicians, PROMs can impact clinical workflow and be financially prohibitive for practices, he added.
O’Neil presented findings of a study on a computerized adaptive testing (CAT) version of the foot and ankle ability measure (FAAM). The researchers won the 2020 Roger A. Mann Award presented by AOFAS.
“CAT has evolved in light of these issues,” he said. “It allows for questions to be tailored to specific traits and responses of an individual patient. It allows for accurate assessments to be made with fewer questions asked, and it allows for the creation of the most concise version possible of an outcome instrument.”
O’Neil and colleagues developed a CAT version of the FAAM, retroactively applying the system to 30,246 patient responses from both the FAAM - activities of daily living (ADL) subscale and the FAAM - sports subscale.
While the full FAAM consists of 32 questions (22 from the ADL subscale and 10 from the sport subscale), O’Neil’s CAT FAAM consisted of approximately 11 to 12 questions from the CAT ADL subscale and five to six questions from the CAT sports subscale.
After plotting differences between the full FAAM and the CAT FAAM, O’Neil found the CAT ADL and CAT sports subscale had a distribution of score differences clustered around 0 and an almost identical frequency distribution of scores.
Researchers also created bland-Altman plots, which demonstrated the difference between the full and CAT forms of the FAAM was independent of the overall score and that there was no bias for the CAT scores in either a positive or negative direction, he said.
“This CAT system can be used interchangeably with full form FAAM scores which can be important for both clinical as well as research purposes,” O’Neil said. “It allows for a significant reduction of question burden, and it provides a more disease- and anatomy-specific outcome measure to complement the PROMs systems.”