# CARE: A Conformal Safety Layer for Medical Summarization

arXiv: https://arxiv.org/abs/2606.08969 · HTML: https://mlsanity.com/paper/2606.08969/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 39 references that are in the map.
- First posted on arXiv 8 Jun 2026.

## Builds on (most cited first, 20 of 20)

- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation](https://mlsanity.com/paper/2305.14251/) (2023)
- [SummaC: Re-Visiting NLI-based Models for Inconsistency Detection in Summarization](https://mlsanity.com/paper/2111.09525/) (2021)
- [AlignScore: Evaluating Factual Consistency with a Unified Alignment Function](https://mlsanity.com/paper/2305.16739/) (2023)
- [Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Control](https://mlsanity.com/paper/2110.01052/) (2021)
- [QAFactEval: Improved QA-Based Factual Consistency Evaluation for Summarization](https://mlsanity.com/paper/2112.08542/) (2021)
- [Language Models with Conformal Factuality Guarantees](https://mlsanity.com/paper/2402.10978/) (2024)
- [ACI-BENCH: a Novel Ambient Clinical Intelligence Dataset for Benchmarking Automatic Visit Note Generation](https://mlsanity.com/paper/2306.02022/) (2023)
- [ConU: Conformal Uncertainty in Large Language Models with Correctness Coverage Guarantees](https://mlsanity.com/paper/2407.00499/) (2024)
- [Mitigating LLM Hallucinations via Conformal Abstention](https://mlsanity.com/paper/2405.01563/) (2024)
- [Towards Interpretable and Efficient Automatic Reference-Based Summarization Evaluation](https://mlsanity.com/paper/2303.03608/) (2023)
- [Improving the Faithfulness of Abstractive Summarization via Entity Coverage Control](https://mlsanity.com/paper/2207.02263/) (2022)
- [SCOPE: Selective Conformal Optimized Pairwise LLM Judging](https://mlsanity.com/paper/2602.13110/) (2026)
- [From Rubrics to Reliable Scores: Evidence-Grounded Text Evaluation with LLM Judges](https://mlsanity.com/paper/2601.08654/) (2026)
- [Document Summarization with Conformal Importance Guarantees](https://mlsanity.com/paper/2509.20461/) (2025)
- [Attribute Structuring Improves LLM-Based Evaluation of Clinical Text Summaries](https://mlsanity.com/paper/2403.01002/) (2024)
- [MedFactEval and MedAgentBrief: A Framework and Workflow for Generating and Evaluating Factual Clinical Summaries](https://mlsanity.com/paper/2509.05878/) (2025)
- [Unsupervised Conformal Inference: Bootstrapping and Alignment to Control LLM Uncertainty](https://mlsanity.com/paper/2509.23002/) (2025)
- [Conformal Prediction for Risk-Controlled Medical Entity Extraction Across Clinical Domains](https://mlsanity.com/paper/2603.00924/) (2026)

## Built on by (most cited first, 2 of 2)

- [LLM Judges Verify Presence, Not Absence: Omission Blindness in AI Clinical Notes and What Recovers It](https://mlsanity.com/paper/2608.31016/) (2026)
- [Deployment-Centered Evaluation: Predicting Query-Level Rejection Risk in a Clinical LLM System](https://mlsanity.com/paper/2606.12702/) (2026)

## Abstract (arXiv)

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× 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.