CARE: A Conformal Safety Layer for Medical Summarization
2026-06-08 • Computation and Language
Computation and LanguageArtificial Intelligence
AI summaryⓘ
The authors developed CARE, a safety system that checks medical summaries made by large language models for missing important information (omissions) and made-up details (hallucinations). CARE works with any model without needing retraining and gives guaranteed limits on how many errors slip through. By carefully balancing detection of omissions and hallucinations, CARE flags fewer sentences for review while still keeping summaries safe. Tests on medical tasks showed CARE meets risk targets reliably, and a small clinician study found it helped spot missing information more effectively. This shows it’s possible to make medical AI summaries safer with a manageable amount of extra review.
large language modelsmedical summarizationomission detectionhallucination detectionconformal risk controlpost-hoc calibrationfinite-sample guaranteessafety layerclinician reviewerror detection
Authors
Suhana Bedi, Bridget Lin, Anson Y. Zhou, Chloe O. Stanwyck, Jenelle A. Jindal, Sanmi Koyejo, David Stutz, Nigam H. Shah
Abstract
Large language models (LLMs) are increasingly used for medical summarization, but their outputs can omit medically important information and introduce unsupported claims. Existing error-detection methods produce heuristic or uncalibrated scores, providing no formal control over missed errors and no principled way to trade off safety against clinician review burden. We introduce Conformal Assessment for Risk Evaluation (CARE), a post-hoc, model-agnostic safety layer that uses conformal risk control to overlay calibrated omission and hallucination flags onto summaries from any LLM without retraining. CARE provides finite-sample, distribution-free guarantees through two controllers: a hallucination controller that bounds the probability of a document containing any unflagged hallucinated sentence, and an omission controller that bounds the expected fraction of important omissions not surfaced for review. Unlike hallucination detection, omissions depend jointly on whether a source sentence is important and whether it is covered by the summary. We show that calibrating only one dimension can violate the target risk bound, while marginal decompositions remain valid but overly conservative. By jointly calibrating over the full $(τ,γ)$ threshold space, CARE preserves formal guarantees while surfacing up to 5$\times$ fewer sentences than alternative calibrated baselines. Across five medical summarization tasks, CARE satisfies the target risk bound at $α= 0.15$ with 95% confidence across 100 calibration/test resplits, using only ~100 labeled documents per domain. In a preliminary clinician study (75 document reviews), calibrated flags improved omission detection by 28.6 percentage points on average. These results show that sentence-level safety guarantees are feasible for LLM-assisted medical summarization and offer a tunable mechanism for balancing residual risk and review effort.