PROMIS physical function item bank v. 1.0: Validity of physical function assessment for athletic training clientele

Context: As the practice of evidence-based medicine grows among athletic trainers, the need for reliable outcome instruments increases substantially. Currently, the field of athletic training has adopted standard orthopedic assessment tools, such as the Foot and Ankle Disability Index (FADI), to obtain necessary patient reported outcomes (PRO). Unfortunately, these tools were developed around concepts of classical test theory, which limit effectiveness for high ability measurements. Within the last decade, researchers have shown that item response theory and computer adaptive tests produce valid, precise measures of functional status. The ability to administer large item banks in a minimal amount of time looks promising for health care professionals, like athletic trainers, who are looking to avoid potential ceiling effects during PRO measurements. The universal item banks created by the Patient-Reported Outcome Measurement Information System (PROMIS) provide practitioners with the tools to administer computer adaptive tests within their own practice. Researchers have found that the PROMIS item banks adequately measure lower extremity, central body functions, and activities of daily living for general populations. However, its’ ability to measure physical function in athletic training clientele is yet to be determined. Objective: Determine the validity of the PROMIS physical function item bank v. 1.0 as an assessment tool for individuals with high levels of physical function by measuring item difficulty and specificity. Design: Non-experimental, observational design. Participants: A heterogeneous sample of 216 physically active people participating in collegiate (n = 151) and intramural (n = 65) sports. Methods: Participants were divided into groups depending on sport type and current health status. Mean PROMIS physical function scores were compared between the groups to assess item specificity and the presence of ceiling effects. Main Outcome Measures: Mean physical function scores calculated for each group using the PROMIS physical function item bank v. 1.0 in computer adaptive form. Results: A one-way ANOVA found a significant difference in mean PROMIS physical function scores between the four sport groups (F = 48.80, p \u3c 0.001). Post hoc analysis found that a significant difference in mean scores existed between the three college groups (healthy, injured and participating, injured and not participating;