How do EQ-5D-3L and EQ-5D-5L compare in a Swedish total hip replacement population?

Ted Eneqvist; Szilárd Nemes; Johan Kärrholm; Kristina Burström; Ola Rolfson

doi:10.1080/17453674.2020.1746124

Authors

Ted Eneqvist Swedish Hip Arthroplasty Register, Gothenburg; Department of Orthopedics, Institute of Clinical Sciences, the Sahlgrenska Academy, University of Gothenburg, Gothenburg; Department of Orthopedics, Sahlgrenska University Hospital Gothenburg
Szilárd Nemes Swedish Hip Arthroplasty Register, Gothenburg; Department of Orthopedics, Institute of Clinical Sciences, the Sahlgrenska Academy, University of Gothenburg, Gothenburg
Johan Kärrholm Swedish Hip Arthroplasty Register, Gothenburg; Department of Orthopedics, Institute of Clinical Sciences, the Sahlgrenska Academy, University of Gothenburg, Gothenburg; Department of Orthopedics, Sahlgrenska University Hospital Gothenburg
Kristina Burström Health Outcomes and Economic Evaluation Research Group, Stockholm Centre for Healthcare Ethics, Department of Learning, Informatics, Management and Ethics, Karolinska Institutet, Stockholm; Equity and Health Policy Research Group, Department of Public Health Sciences, Karolinska Institutet, Stockholm; Health Care Services, Region Stockholm, Stockholm, Sweden
Ola Rolfson Swedish Hip Arthroplasty Register, Gothenburg; Department of Orthopedics, Institute of Clinical Sciences, the Sahlgrenska Academy, University of Gothenburg, Gothenburg; Department of Orthopedics, Sahlgrenska University Hospital Gothenburg; Health Outcomes and Economic Evaluation Research Group, Stockholm Centre for Healthcare Ethics, Department of Learning, Informatics, Management and Ethics, Karolinska Institutet, Stockholm

DOI:

https://doi.org/10.1080/17453674.2020.1746124

Abstract

Background and purpose — To better detect small changes in postoperative outcome following total hip replacement (THR), the Swedish Hip Arthroplasty Register (SHAR) has decided to change from the EQ-5D-3L (3L) to the EQ- 5D-5L (5L). To enable comparison of results obtained with use of the 2 versions of EQ-5D, transferal of results between the questionnaires used is necessary. We assessed the measurement properties of the EQ-5D-5L compared with the EQ-5D-3L, preoperatively and 1-year postoperatively in a Swedish THR population.

Patients and methods — Patients eligible for elective THR during 2015 in Western Sweden were invited to the study. With a 2-week separation, the 3L and 5L questionnaires were administered to patients before and 1 year after surgery. Comparing the 2 versions of the EQ-5D, we investigated redistribution of responses, ceiling and floor effects, EQ VAS correlations (Spearman’s rank correlation coefficient, rs ), and EQ VAS scores for different severity levels by dimension (univariable ordinary least square regression).

Results — The additional severity levels of the 5L version were frequently used on both measurement occasions (preoperative mobility 5%, self-care 17%, usual activities 20%, pain 5% and anxiety 3%, postoperative mobility 6%, self-care 5%, usual activities 8%, pain 9%, and anxiety 5%). Ceiling effects of the 3L version diminished overall by 7% using the 5L version. The correlations between the 2 EQ VAS scores obtained with the 3L and 5L instruments were strong both pre- (rs = 0.71) and postoperatively (rs = 0.87). Estimated EQ VAS scores for different levels of severity were consistent for all dimensions except for the mobility dimension of the preoperative 5L version and the anxiety dimension in the postoperative 5L version.

Interpretation — Our findings support that the 5L has a higher resolution than the 3L version regarding description of health-related quality of life in patients undergoing THR in Sweden. The EQ VAS scores for different levels of severity agree well between the EQ-5D versions. This could potentially be used to develop a crosswalk value set for transforming 3L to 5L responses in this patient group.