Outcome metrics in pediatric sports medicine: Do PROMIS computer-adaptive testing metrics correlate with pedi-IKDC?

Background: A modified version of the International Knee Documentation Committee (IKDC), the Pedi-IKDC, is a validated patient-reported outcome measure in pediatric patients with knee pain. However, this questionnaire is lengthy and can fatigue patients, leading to inconsistent outcome collection. Thus, we sought to compare more easily attainable Patient-Reported Outcomes Measurement Information System (PROMIS) computer adaptive testing patient reported outcomes to the gold standard Pedi-IKDC. Methods: We prospectively collected PROMIS scores and Pedi-IKDC scores in 100 new, consecutive patients presenting with knee pain to a pediatric sports medicine practice. Patients were excluded if they provided an incomplete Pedi-IKDC, had prior ipsilateral lower extremity surgery, or significant associated medical comorbidities. PROMIS domains including Mobility, Pain Interference, and Upper Extremity (control) were compared with the Pedi-IKDC with Pearson correlations. The number of questions in each metric was analyzed. Floor and ceiling effects of each test were also assessed. Results: The average age of the study cohort was 14 years (range, 7 to 18 y) with 53% female and 47% male. 70% of patients completed the Pedi-IKDC questionnaire, compared with 100% with PROMIS tests. The average Pedi-IKDC score was 48.8±22.3 (range, 5.4 to 100). Mean scores for Mobility, Pain Interference, and Upper Extremity domains were 38.4±10.1, 53.5±10.3, and 49.7±8.7, respectively. All tests demonstrated similar and acceptable floor and ceiling effects (<15%). The length of the Pedi-IKDC (22 questions) was roughly double that of combined PROMIS Pain Interference and Mobility tests (11.9±2.3 questions). Pedi-IKDC scores correlated with tested PROMIS measures (Mobility/Pain Interference, r=0.42/-0.49). When 7 highly functional patients with significant pain symptoms were removed for a secondary analysis, Mobility and Pain correlations improved to 0.69 and -0.67, respectively. Conclusions: PROMIS Mobility and Pain scores demonstrate moderate correlations with the Pedi-IKDC, highlighting these tests are not capturing the same patient experiences. These correlations are weakened by a small group of painful yet highly functioning patients. The Pedi-IKDC was significantly longer and had a much lower completion rate than PROMIS tests, highlighting a need for a validated computer adaptive testing in evaluating pediatric patients with knee pain. Levels of Evidence: Level II.