<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-post">
    <titleInfo>
        <title>Can Post-Training Turn LLMs into Good Medical Coders? An Empirical Study of Generative ICD Coding</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Ziqing</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Weihao</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Shijie</namePart>
        <namePart type="family">Chen</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yuan</namePart>
        <namePart type="family">Luo</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kaize</namePart>
        <namePart type="family">Ding</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-462-0</identifier>
    </relatedItem>
    <abstract>Automated International Classification of Diseases (ICD) coding is a core medical-coding task for billing, epidemiology, and clinical decision support. Generative large language models (LLMs) are often reported as weak medical coders, but this finding mainly comes from inference-time settings such as prompting, retrieval, reranking, or tool use; the role of task-specific post-training has received limited systematic study. We present a controlled empirical study of post-training for generative ICD coding, comparing discriminative baselines with LLM coders across prompting, supervised fine-tuning (SFT), and reinforcement learning (RL) under a common protocol and metric set. To our knowledge, this is the first comprehensive controlled study of post-training for generative ICD prediction. We further introduce Progressive Hint Injection (PHI), a diagnostic missed-code curriculum within Group Relative Policy Optimization (GRPO) training. Our results show that prompting-only evaluation substantially underestimates the potential of LLMs for ICD coding. SFT provides the main capability jump, GRPO further improves code-set prediction beyond SFT, and PHI provides targeted gains on macro-level performance. These findings suggest that the main bottleneck is not the generative formulation alone, but how the model is adapted and optimized for full-taxonomy recall. We release our code, data splits, and checkpoints at https://anonymous.4open.science/r/LLM4ICD.</abstract>
    <identifier type="citekey">wang-etal-2026-post</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.459/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>9643</start>
            <end>9658</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Can Post-Training Turn LLMs into Good Medical Coders? An Empirical Study of Generative ICD Coding
%A Wang, Ziqing
%A Li, Weihao
%A Chen, Shijie
%A Luo, Yuan
%A Ding, Kaize
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F wang-etal-2026-post
%X Automated International Classification of Diseases (ICD) coding is a core medical-coding task for billing, epidemiology, and clinical decision support. Generative large language models (LLMs) are often reported as weak medical coders, but this finding mainly comes from inference-time settings such as prompting, retrieval, reranking, or tool use; the role of task-specific post-training has received limited systematic study. We present a controlled empirical study of post-training for generative ICD coding, comparing discriminative baselines with LLM coders across prompting, supervised fine-tuning (SFT), and reinforcement learning (RL) under a common protocol and metric set. To our knowledge, this is the first comprehensive controlled study of post-training for generative ICD prediction. We further introduce Progressive Hint Injection (PHI), a diagnostic missed-code curriculum within Group Relative Policy Optimization (GRPO) training. Our results show that prompting-only evaluation substantially underestimates the potential of LLMs for ICD coding. SFT provides the main capability jump, GRPO further improves code-set prediction beyond SFT, and PHI provides targeted gains on macro-level performance. These findings suggest that the main bottleneck is not the generative formulation alone, but how the model is adapted and optimized for full-taxonomy recall. We release our code, data splits, and checkpoints at https://anonymous.4open.science/r/LLM4ICD.
%U https://aclanthology.org/2026.findings-emnlp.459/
%P 9643-9658

Markdown (Informal)

[Can Post-Training Turn LLMs into Good Medical Coders? An Empirical Study of Generative ICD Coding](https://aclanthology.org/2026.findings-emnlp.459/) (Wang et al., Findings 2026)

• Can Post-Training Turn LLMs into Good Medical Coders? An Empirical Study of Generative ICD Coding (Wang et al., Findings 2026)

ACL

• Ziqing Wang, Weihao Li, Shijie Chen, Yuan Luo, and Kaize Ding. 2026. Can Post-Training Turn LLMs into Good Medical Coders? An Empirical Study of Generative ICD Coding. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 9643–9658, Budapest, Hungary. Association for Computational Linguistics.