Validity of retrospective patient reported assessment of pre-surgical hip pain and disability following hip arthroscopy

Abstract

PURPOSE: The purpose of this study was to assess the validity of pre-surgery PROMS completed retrospectively 6 and 24 months after hip arthroscopy compared with prospective completion. Validity was judged in relation to preoperative test-retest reliability.

METHODS: Three separate groups of hip arthroscopy patients were recruited. In two retrospective groups, patients who had undergone a hip arthroscopy and had completed preoperative PROM questionnaires were contacted 6 or 24 months after surgery and asked to recall their preoperative state. In a third test-retest group, patients completed prospective PROMs twice prior to surgery, 2–8 weeks apart. PROMs included Non-Arthritic Hip Score (NAH), Hip Disability and Osteoarthritis Outcome score (HOOS), International Hip Outcome Tool (IHOT- 12), and Visual Analogue Scale (VAS) for Medical Care, Feeling and Pain.

RESULTS: In the 6-month, 24-month and test-retest groups, patient sample sizes were 52, 40 and 61, respectively. Retrospective 6- and 24-month and test-retest PROMs demonstrated ‘moderate’ to ‘very large’ correlations (ICC=0.42–0.79), except for 6-month VAS-Medical- Care and VAS-Feeling (‘small’) and 24-month VAS-Feeling (‘trivial’). PROMs for the 24- month group reduced, from prospective to retrospective, for NAH-Total (MD -4.10 [-7.93–- 0.27; (95%CI)]; p=0.03), HOOS Symptoms (-8.21 [-13.44–-2.97]; p=0.003), HOOS-Daily- Living (-7.53 [-12.71–-2.36]; p=0.005), HOOS-Sports (-6.49 [-12.97–0.00]; p=0.05), and HOOS-Short-Form (-5.16 [-9.83–-0.49]; p=0.03). Similar non-significant reductions were observed for 6-month retrospective PROMs, except for VAS-Feeling which increased retrospectively (7.47 [0.13–14.81]; p=0.05). Three test-retest PROMs also showed statistically significant reduction, with worse outcomes approaching surgery: HOOS-Daily- Living (-.3.87 [-7.73–0.00]; p=0.05), HOOS-Short-Form (-4.49 [-8.41–-0.53]; p=0.03), and IHOT-12-Total (-4.91 [-8.7–-1.23]; p=0.01).

CONCLUSION: Correlations for retrospective versus prospective PROM scores spanned a similar range to test-retest correlations. Therefore, PROM scores relating to pre-surgery condition, but completed 6 or 24 months after surgery, showed agreement with PROM scores completed before surgery which were consistent with the day-to-day variability expected before surgery. This consistency supports the use of retrospective PROMs completed up to 2 years after hip arthroscopy.