High recall bias in retrospective assessment of the pediatric International Knee Documentation Committee Questionnaire (Pedi-IKDC) in children with knee pathologies

Purpose The increasing incidence of knee injuries among children is well known by sports physicians. Papers dealing with this topic have often collected patient-reported outcomes measures (PROMs) in a retrospective manner; this limitation could lead to a misinterpretation of the results, because pediatric patients might not remember their preoperative conditions adequately. This study aims to evaluate the reliability and the reproducibility of the IKDC pediatric score when administered retrospectively at a 12-month follow-up. Methods From September 2018 and June 2019, all patients aged 7-18 scheduled for surgery due to different knee pathologies in a single center were considered eligible. Parents were contacted by phone for consent. An open-source platform was implemented to collect the responses: two surveys were created (Q1, Q2). They included general information and the Pedi-IKDC score. Q1 was completed prospectively, while Q2 was completed 12 months after surgery. The two questionnaires were identical, and patients were carefully advised to complete Q2 recalling their health status before surgery. ICC and the concordance correlation coefficient (rho(c)) were used to assess the reproducibility between the prospective and recalled scores. Results Sixty-six patients responded to Q1 and Q2, and the mean age was 12.9 +/- 2.2 years at Q1 and 14.1 +/- 2.2 years at Q2. The mean time between Q1 and Q2 was 14.1 +/- 2.1 months. Between prospective-IKDC and recall-IKDC, the ICC coefficient was "poor" at 0.32 (CI 0.09 to 0.5) and the rho(c) was "poor" at 0.4 (CI 0.29 to 0.51). Mean prospective-IKDC was 76.8 +/- 23.52 mean recalled-IKDC was 60.4 +/- 11.5 (P < 0.0001), while mean difference was -16.3 +/- 2.09. Simple linear regression models showed that Delta-IKDC is independently associated with age at Q1 (R-2 = 0.2676; P0.0001) and prospective-IKDC (R-2 = 0.653; P < 0.0001). Conclusions Retrospective collection of the Pedi-IKDC score is not reliable and has high recall bias. This should be avoided in children with knee conditions.