Presentation (CTI)

OBJECTIVES: Maximum acceptable risk (MAR) measures the largest acceptable level of a single risk that patients would accept in exchange for a specific treatment benefit, assuming all other risks are held constant. Simultaneous maximum acceptable risk thresholds (SMARTs) have been proposed to estimate tolerance for multiple risks occurring simultaneously. We demonstrate the feasibility of estimating SMARTs and compare SMARTs to individual MAR estimates within and across regions.
METHODS: Using data collected from a previously conducted DCE, we estimate SMARTs for 3 risk attributes (A, B, and C) for a specific treatment benefit. Confidence bands were computed to reflect uncertainty in the estimates. We also estimate the MAR for each risk independently for the same treatment benefit.
RESULTS: Data from N=177 UK patients and N=246 EU patients (France, Germany, Italy, and Spain) were included. In all cases, tolerance for any individual risk was lower when considered in combination with other risks (SMART) than when considered alone (MAR), as expected. Across regions, the MAR of Risk A in exchange for the treatment benefit was higher for patients in the UK than in the EU (3.2% [CI=2.8%-3.6%] vs 2.1% [CI=1.9%-2.4%]). The MARs for Risk B and Risk C were close to or above their respective maximum levels tested in the DCE (0.5% and 5%, respectively) for both regions. When Risks A and B were considered jointly, the SMART demonstrated patients in the UK were willing to accept significantly higher levels of each risk than patients in the EU if Risk C was below approximately 3%. No difference in SMART estimates for Risks A or B was observed between regions when Risk C was held at 5%.
CONCLUSIONS: Calculating SMARTs is feasible and provides a more nuanced understanding of benefit-risk trade-offs, offering insights into real-world treatment decision-making that extend beyond traditional MAR estimates.