Matching-adjusted indirect comparison via a polynomial-based non-linear optimization method

Area of Science:

• Health Economics and Outcomes Research
• Biostatistics
• Pharmacoeconomics

Background:

• Matching-adjusted indirect comparison (MAIC) is crucial for synthesizing evidence from real-world data.
• Existing MAIC methods may lack flexibility in matching complex aggregate-level data.
• Improving the precision and accuracy of MAIC is essential for robust comparative effectiveness research.

Purpose of the Study:

• To evaluate the potential of fourth-order polynomials within a non-linear optimization framework for MAIC.
• To introduce and assess the performance of the polyMAIC method.
• To compare polyMAIC against the industry-standard Signorovitch MAIC approach.

Main Methods:

• Simulated individual patient data were reweighted using fourth-order polynomials (polyMAIC).
• Matching was performed against aggregate-level data across multiple baseline characteristics.
• The polyMAIC approach utilized pre-specified matching tolerances and maximum allowable weights.

Main Results:

• The polyMAIC method successfully matched aggregate-level targets within the defined tolerances.
• Effective sample sizes generated by polyMAIC were comparable or slightly higher than the Signorovitch method.
• PolyMAIC demonstrated improved performance gains with increasing matching complexity.

Conclusions:

• PolyMAIC offers enhanced flexibility compared to the standard industry MAIC approach.
• The polyMAIC method shows significant potential for improving the matching process in indirect comparisons.
• This approach holds promise for more accurate and reliable evidence synthesis in health research.