A high proportion of patients demonstrate recall bias in the retrospective collection of patient-reported outcomes following hip arthroscopy.

Abstract

Purpose: The aim of this study is to assess agreement between retrospectively and prospectively collected patient-reported outcome measures (PROMs) following hip arthroscopy for femoroacetabular impingement syndrome (FAIS).

Methods: Patients undergoing hip arthroscopy from 2021 to 2023 for FAIS completed preoperative PROMs, including the modified Harris Hip Score (mHHS) and the Non-Arthritic Hip Score (NAHS). Post-operatively, patients were surveyed and asked to recall their preoperative hip function. Paired two-sample t tests were used to compare baseline and recalled baseline PROMs and the difference between scores was compared to previously published minimally clinically important difference (MCID) values. Intraclass correlation coefficients (ICCs) were calculated to test the reliability between scores based on a single-rater, two-way mixed-effects model. Multivariable regression, accounting for age, sex and preoperative baseline scores, was used to evaluate the relationship of time elapsed since surgery with recall accuracy.

Results: A total of 116 patients (age: 37.6 ± 11.8 years; 61.2% female) were included. The mean time elapsed for recalled data was 13.1 months (range: 1-27 months). Overall, patients' recalled scores were significantly lower than those prospectively collected (mHHS: 52.9 ± 20.1 vs. 61.5 ± 18.5, p < 0.0001; NAHS: 54.7 ± 20.0 vs. 58.8 ± 19.1, p < 0.0001). Frequency distribution found 68.1% of recalled mHHS and 61.2% of NAHS scores to have a greater difference (between baseline and recalled scores) than the MCID. The ICC was moderate for both mHHS (ICC = 0.559, 95% confidence interval [CI] = [0.420-0.672], p < 0.001) and NAHS (ICC = 0.612, 95% CI = [0.484-0.714], p < 0.001). Multivariate regression analysis did not find time elapsed since surgery to be associated with the difference between baseline and recalled mHHS (n.s.) or NAHS (n.s.).

Conclusion: There are significant differences between retrospective and prospectively collected PROMs in patients undergoing hip arthroscopy that are not predicted by time to recall. These findings should impact the interpretation of the existing literature, support the routine collection of prospective data and inform patient counsel regarding their perceived post-operative outcomes.

Level of evidence: Level IV.