Evaluating Counterfactual Sensitivity to Patient Information in Medication-Safety Reasoning

2026-08-03

AI Summary

This study addresses the limitation of existing medical large language models in dynamically assessing the applicability of drug safety rules based on individual patient information, often relying on static associations rather than conditional reasoning. To tackle this issue, the authors propose MedPIC-Bench, a novel benchmark that introduces a verifiable counterfactual perturbation mechanism to systematically evaluate model sensitivity and reliability in applying clinical rules under varying patient conditions. Constructed from clinical guidelines, the benchmark comprises 467 expert-annotated questions spanning six clinical and reasoning dimensions. Experiments across 28 models reveal a significant performance drop—from 63.6% to 45.1% accuracy—on counterfactual instances, exposing a widespread deficiency in conditional reasoning among current domain-specific medical models.

Abstract

Applying a valid medication-safety rule when its patient-specific conditions are not met can produce an incorrect decision. Existing medical evaluations largely use isolated and fixed scenarios. A model may therefore answer correctly by recalling a drug-risk association without showing that it used patient information to decide whether the rule applies. To address this gap, we introduce MedPIC-Bench, a benchmark of source-verifiable recommendations and expert-validated questions for patient-specific medication-safety reasoning. It combines guideline-following questions with paired counterfactual questions in which a controlled change in patient information changes whether a rule applies. The benchmark contains 467 questions annotated along six clinical and reasoning dimensions. Across 28 medical-specific, general, and proprietary LLMs, every model performs worse on counterfactual questions, with mean accuracy falling from 63.6\% to 45.1\%. Models perform well when an explicit patient attribute directly signals a familiar contraindication, but struggle when patient information must narrow or withdraw a safety warning. Model rationales often acknowledge the changed patient information, yet the final answers retain the previous safety judgment. This vulnerability persists among medical-specific LLMs, whose average CF performance trails that of general LLMs. MedPIC-Bench therefore makes conditional rule application measurable and highlights the limitations of static medication-safety accuracy for assessing patient-specific reliability.

Problem

• counterfactual reasoning
• medication safety
• patient-specific information
• clinical decision-making
• rule applicability

Innovation

• counterfactual reasoning
• patient-specific reasoning
• MedPIC-Bench
• medication-safety evaluation
• conditional rule application