# Evaluation and LLM-Guided Learning of ICD Coding Rationales

arXiv: https://arxiv.org/abs/2508.16777 · HTML: https://mlsanity.com/paper/2508.16777/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 38 references that are in the map.
- First posted on arXiv 22 Aug 2025.

## Builds on (most cited first, 16 of 16)

- [ERASER: A Benchmark to Evaluate Rationalized NLP Models](https://mlsanity.com/paper/1911.03429/) (2019)
- [Explainable Prediction of Medical Codes from Clinical Text](https://mlsanity.com/paper/1802.05695/) (2018)
- [Code Synonyms Do Matter: Multiple Synonyms Matching Network for Automatic ICD Coding](https://mlsanity.com/paper/2203.01515/) (2022)
- [Automated Medical Coding on MIMIC-III and MIMIC-IV: A Critical Review and Replicability Study](https://mlsanity.com/paper/2304.10909/) (2023)
- [A Label Attention Model for ICD Coding from Clinical Text](https://mlsanity.com/paper/2007.06351/) (2020)
- [Knowledge Injected Prompt Based Fine-tuning for Multi-label Few-shot ICD Coding](https://mlsanity.com/paper/2210.03304/) (2022)
- [Explainable Automated Coding of Clinical Notes using Hierarchical Label-wise Attention Networks and Label Embedding Initialisation](https://mlsanity.com/paper/2010.15728/) (2020)
- [MDACE: MIMIC Documents Annotated with Code Evidence](https://mlsanity.com/paper/2307.03859/) (2023)
- [CoRelation: Boosting Automatic ICD Coding Through Contextualized Code Relation Learning](https://mlsanity.com/paper/2402.15700/) (2024)
- [An Unsupervised Approach to Achieve Supervised-Level Explainability in Healthcare Records](https://mlsanity.com/paper/2406.08958/) (2024)
- [ICDBigBird: A Contextual Embedding Model for ICD Code Classification](https://mlsanity.com/paper/2204.10408/) (2022)
- [Accurate and Well-Calibrated ICD Code Assignment Through Attention Over Diverse Label Embeddings](https://mlsanity.com/paper/2402.03172/) (2024)
- [Can Current Explainability Help Provide References in Clinical Notes to Support Humans Annotate Medical Codes?](https://mlsanity.com/paper/2210.15882/) (2022)
- [This Patient Looks Like That Patient: Prototypical Networks for Interpretable Diagnosis Prediction from Clinical Text](https://mlsanity.com/paper/2210.08500/) (2022)
- [Dynamically Extracting Outcome-Specific Problem Lists from Clinical Notes with Guided Multi-Headed Attention](https://mlsanity.com/paper/2008.01197/) (2020)
- [uMedSum: A Unified Framework for Advancing Medical Abstractive Summarization](https://mlsanity.com/paper/2408.12095/) (2024)

## Built on by (most cited first, 1 of 1)

- [Can Post-Training Turn LLMs into Good Medical Coders? An Empirical Study of Generative ICD Coding](https://mlsanity.com/paper/2606.13940/) (2026)

## Abstract (arXiv)

ICD coding is the process of mapping unstructured text from Electronic Health Records (EHRs) to standardised codes defined by the International Classification of Diseases (ICD) system. In order to promote trust and transparency, existing explorations on the explainability of ICD coding models primarily rely on attention-based rationales and qualitative assessments conducted by physicians, yet lack a systematic evaluation across diverse types of rationales using consistent criteria and high-quality rationale-annotated datasets specifically designed for the ICD coding task. Moreover, dedicated methods explicitly trained to generate plausible rationales remain scarce. In this work, we present evaluations of the explainability of rationales in ICD coding, focusing on two fundamental dimensions: faithfulness and plausibility -- in short how rationales influence model decisions and how convincing humans find them. For plausibility, we construct a novel, multi-granular rationale-annotated ICD coding dataset, based on the MIMIC-IV database and the updated ICD-10 coding system. We conduct a comprehensive evaluation across three types of ICD coding rationales: entity-level mentions automatically constructed via entity linking, LLM-generated rationales, and rationales based on attention scores of ICD coding models. Building upon the strong plausibility exhibited by LLM-generated rationales, we further leverage them as distant supervision signals to develop rationale learning methods. Additionally, by prompting the LLM with few-shot human-annotated examples from our dataset, we achieve notable improvements in the plausibility of rationale generation in both the teacher LLM and the student rationale learning models.