Feasibility of 4 patient-reported outcome measures in a registry setting

Abstract

Background and purpose Feasibility is an important parameter when choosing which patient-reported outcomes (PRO) to use in a study. We assessed the feasibility of PROs in a hip registry setting.

Methods Primary total hip arthroplasty (THA) patients (n = 5,747) who had been operated on 1–2, 5–6, or 10–11 years previously were randomly selected from the Danish Hip Arthroplasty Register and sent 2 PRO questionnaires: 1 generic (EuroQoL-5D or SF-12 health survey) and 1 disease-specific (hip dysfunction and osteoarthritis outcome score (HOOS) or Oxford 12-item hip score). We compared response rates, floor and ceiling effects, missing items, and the need for manual validation of forms.

Results 4,784 patients (mean age 71 years, 57% females) were included (83%). The response rates ranged from 82–84%. Statistically significantly different floor and ceiling effects ranged from 0% to 0.5% and from 6.1% to 46%, respectively. Missing items ranged from 1.2% to 3.4%, and 0.8–4.3% required manual validation (p < 0.009). A hypothetical repeat study found that group sizes from 51 to 1,566 are needed for subgroup analysis, depending on descriptive factor and choice of PRO.

Interpretation All 4 PROs fulfilled a priori set criteria, with the exception of ceiling effects. The high ceiling effects were attributed to postoperative administration and good outcome for THA. We conclude that all 4 PROs are appropriate for administration in a hip registry.

AP, ABP, SO, and EMR participated in the design of the study, analysis of data, and in writing of the manuscript. AP prepared the raw data.

This study was funded by Region Syddanmark, Gigtforeningen, Syddansk Universitet, Familien Hede Nielsens Legat, Bauers Legat, and Ryholts Legat, none of which played a role in the investigation.

EMR is a member of the HOOS development group. The other authors have no competing interests.