Using EQ-5D Data to Measure Hospital Performance: Are General Population Values Distorting Patients' Choices?

Area of Science:

• Health Services Research
• Quality of Life Measurement
• Health Economics

Background:

• Hospital performance indicators for hip replacement surgery rely on EQ-5D index scores.
• These scores use general population preference weights, raising concerns about patient relevance.
• The study investigates the impact of different valuation sources on these performance estimates.

Purpose of the Study:

• To explore how the source of valuation (general population vs. patients) affects hospital performance estimates for hip replacement surgery.
• To compare the influence of different valuation techniques (VAS vs. TTO) and experience with health states.
• To assess the impact on national hospital rankings and outlier status.

Main Methods:

• Four distinct value sets were used, varying in valuation source, technique, and health state experience.
• New EQ-5D-3L value sets were estimated from data of 122,921 hip replacement patients and 3,381 UK general public members.
• Hospital rankings and outlier status changes were compared across these different valuation methods for 243 hospitals.

Main Results:

• National hospital rankings remained largely stable across different valuations (rank correlations >0.92).
• However, 9.5% of hospitals changed outlier status when using patient VAS valuations compared to the current general population TTO approach.
• Changes in outlier status were significant at the local level, primarily driven by the valuation technique.

Conclusions:

• The choice of value set can influence prospective patients' hospital selection for hip replacement surgery.
• Empirical evidence using VAS valuations does not support normative concerns regarding the use of general population preferences.
• Valuation technique appears to be a more critical factor than the source of valuation in altering performance estimates.