Making patient-reported outcomes relevant: Two examples measuring health-related quality of life in primary care

Context: Patient-reported outcome measures (PROMs) are increasingly incorporated into primary care practices; however, the broad characteristics of this population pose unique implementation challenges. Generic PROMs (e.g., EQ-5D-5L) measure health-related quality of life (HRQL) can contribute toward a more comprehensive evaluation of health programming by integrating the patient’s perspective.

Objective: The purpose of this study was to describe the integration of the EQ-5D-5L within a program evaluation framework in primary care in Alberta, Canada.

Study design and analysis: This retrospective observational study, uses routinely collected data, examining HRQL as measured by the EQ-5D-5L. Data was analyzed descriptively, by dimension, index, and visual analogue scale (VAS) scores. Minimally important differences were used to interpret index (0.04) and VAS scores (7.0), and the Pareto Classification of Health Change was applied to the dimensions. Effect size of change in index and VAS scores and population comparisons were examined.

Population studied: The Edmonton O-Day’min Primary Care Network (EOPCN) routinely collects the EQ-5D-5L from patients entering the Kinesiology (KIN) and Dietitian (DIET) programs. Data was collected between January 1, 2021 and March 31, 2022.

Instrument: EQ-5D-5L.

Outcome measures: Health-related quality of life.

Results: Among patients in the KIN program (n=301), 72% were female, with an overall mean age of 57 (±16) years, and 90.2% reported problems (levels 2-5) in pain/discomfort. The mean index and VAS scores were 0.74 (±0.18) and 66.5 (±18.9), respectively. Among those with repeated measurement (n=112) the greatest improvement was in usual activities (17%); the greatest deterioration was in pain/discomfort (25%). In the DIET program (n=573), 71% of patients were female, with an overall mean age of 50 (±17) years, and 75.9% reported problems in pain/discomfort. The mean index and VAS scores were 0.77 (±0.19) and 70.1 (±15.6), respectively. From those with repeated measurement (n=212), the greatest improvement was in anxiety/depression (16%); the greatest deterioration was in pain/discomfort (26%). In both programs, HRQL was lower than the general Alberta population norms at baseline and improved slightly by repeated measurement. No significant predictors of change in index or VAS scores were found.

Conclusions: These results inform EOPCN programming by describing the health status of patients accessing these programs.