Generalized Pairwise Comparisons to Support Shared Decision-Making in the CODA Trial.

Area of Science:

• Comparative Effectiveness Research
• Health Services Research
• Biostatistics

Background:

• Shared decision-making (SDM) is challenged by assessing multiple treatment outcomes.
• Generalized pairwise comparisons (GPC) offer a rigorous method for analyzing multi-outcome data.
• GPC has the potential to enhance SDM by systematically evaluating diverse outcomes.

Purpose of the Study:

• To evaluate the utility of generalized pairwise comparisons (GPC) in supporting shared decision-making (SDM).
• To apply GPC to individual patient data from the Comparison of Outcomes of Antibiotic Drugs and Appendectomy (CODA) trial.

Main Methods:

• A comparative effectiveness study utilizing data from the multicenter US CODA trial (May 2016-March 2020).
• Formation of all possible patient pairs across treatment arms (antibiotics vs. appendectomy) for sequential analysis of seven outcomes.
• Analysis of data between February 2020 and early 2024, considering three distinct prioritization scenarios.

Main Results:

• The study included 1552 patients (776 per arm).
• Net treatment benefit (NTB) varied significantly across scenarios: 12.8% favoring antibiotics, 3.2% favoring neither, and -14.5% favoring appendectomy.
• Results demonstrated that the perceived benefit-risk balance is highly dependent on the prioritization of outcomes.

Conclusions:

• The GPC method provides a flexible and rigorous quantitative analysis for comparing treatment benefit-risk.
• GPC offers a nuanced assessment of differences between treatment modalities based on prioritized outcomes.
• This approach can significantly support SDM by incorporating individual patient priorities for multiple outcomes.