Statistical Methods for Analyzing EQ-5D in Randomized Clinical Trials: A Systematic Literature Review

Area of Science:

• Health Economics
• Biostatistics
• Clinical Trial Methodology

Background:

• The EuroQol 5-dimension (EQ-5D) is a widely used patient-reported outcome measure in clinical trials.
• Analyzing treatment effects on EQ-5D data requires appropriate statistical methods to ensure valid and reliable results.

Purpose of the Study:

• To systematically review and summarize the statistical methods applied to analyze treatment effects on EQ-5D measures in randomized clinical trials (RCTs).
• To identify common practices and variations in the analysis of EQ-5D data, including utilities and visual analog scale (VAS) scores.

Main Methods:

• A systematic literature search was conducted in MEDLINE and EMBASE databases up to 2021.
• Eligible studies were RCTs analyzing postbaseline EQ-5D data by treatment group.
• Data on trial characteristics, EQ-5D data, and statistical methods were extracted and summarized using descriptive statistics.

Main Results:

• EQ-5D utilities were the most frequently analyzed outcome, followed by EQ Visual Analog Scale (VAS).
• Linear fixed-effect and linear mixed-effect models were common for analyzing utilities.
• A significant proportion of studies did not adjust for baseline EQ-5D scores or adequately address missing data using imputation methods.

Conclusions:

• Health utilities and EQ VAS are the primary EQ-5D data analyzed in clinical trials.
• There is considerable heterogeneity in statistical modeling approaches for EQ-5D.
• Improvements are needed in handling baseline data and missing data to enhance the rigor of EQ-5D analyses in RCTs.