Evaluating bias in the anchor method for the minimal clinically important difference: a simulation approach

Anchor based methods have been used in clinical studies to determine minimal clinically important differences (MCID) for clinical outcome assessments. However, the theoretical properties and robustness of the methodology are not fully understood. We conducted a simulation study to explore the performance of anchor-based methods across a range of values for outcome variance, placebo effects, anchor measurement noise, and confounding. Our results demonstrate that considerable placebo effects, anchor measurement error, and confounders may introduce a substantial bias into the estimated MCID. We also discuss strategies to identify and mitigate these biases.