The statistical approach in trial-based economic evaluations matters: get your statistics together!

Area of Science:

• Health Economics
• Biostatistics
• Clinical Trial Analysis

Background:

• Trial-based economic evaluations face statistical challenges including baseline imbalances, skewed costs, correlated costs and effects, and missing data.
• These challenges are often inadequately addressed in standard analyses, potentially leading to biased results.
• This study highlights the critical impact of these statistical issues on cost-effectiveness findings.

Purpose of the Study:

• To illustrate the impact of accounting for statistical challenges in trial-based economic evaluations.
• To compare different statistical approaches in analyzing cost-effectiveness data.
• To provide guidance for researchers on handling statistical complexities in economic evaluations.

Main Methods:

• Utilized data from two trial-based economic evaluations: the REALISE and HypoAware studies.
• Conducted 14 full cost-effectiveness analyses per study, systematically incorporating four key statistical challenges.
• Compared statistical approaches based on cost and effect differences, incremental cost-effectiveness ratios (ICERs), and probabilities of cost-effectiveness.

Main Results:

• In the REALISE study, the ICER shifted from 636,744€/QALY to -7,502€/QALY when accounting for all challenges.
• In the HypoAware study, the ICER changed from 90,989€/QALY to 46,592€/QALY after addressing statistical issues.
• The probability of an intervention being cost-effective decreased significantly when statistical challenges were properly handled in both studies.

Conclusions:

• Failure to account for baseline imbalances, skewed costs, correlated costs/effects, and missing data can substantially alter economic evaluation outcomes.
• Aligning statistical methods with identified challenges is crucial for reliable trial-based economic evaluations.
• Software code is provided to assist researchers in managing these statistical complexities.