Measuring and illustrating statistical evidence in a cost-effectiveness analysis.

Area of Science:

• Health Economics
• Biostatistics
• Decision Sciences

Background:

• Cost-effectiveness acceptability curves (CEAC) are widely used to measure statistical evidence of cost-effectiveness.
• Existing CEAC interpretations are based on either frequentist or Bayesian frameworks, leading to fundamental differences.
• A unified approach is needed to interpret statistical evidence in economic evaluations.

Purpose of the Study:

• To propose an alternative method for characterizing statistical evidence of cost-effectiveness.
• To introduce a new graphical display for presenting this evidence.
• To demonstrate the utility of the proposed likelihood-based approach using real-world data.

Main Methods:

• Characterizing statistical evidence using the likelihood function, independent of sample space or prior distributions.
• Developing a novel graphical method to display likelihood-based evidence for cost-effectiveness.
• Applying the proposed method to data from an economic evaluation of a Program in Assertive Community Treatment (PACT).

Main Results:

• The likelihood function offers an interpretation of statistical evidence that is robust and not dependent on specific statistical frameworks.
• The proposed graphical method effectively displays the evidence about cost-effectiveness.
• The likelihood approach demonstrates controllable probability of misleading evidence, aligning with frequentist principles.

Conclusions:

• The likelihood function provides a justifiable and unified framework for assessing statistical evidence in cost-effectiveness analysis.
• The proposed graphical tool enhances the interpretation of economic evaluation results.
• This approach offers a valuable alternative to traditional CEAC interpretations in health economics.