{"id":28375,"date":"2025-07-10T04:07:00","date_gmt":"2025-07-09T20:07:00","guid":{"rendered":"https:\/\/csccm.org.cn\/?p=28375"},"modified":"2025-07-10T06:22:14","modified_gmt":"2025-07-09T22:22:14","slug":"jama-intern-med%e5%8f%91%e8%a1%a8%e8%ae%ba%e6%96%87%ef%bc%9a%e5%8c%bb%e7%94%9f%e6%92%b0%e5%86%99%e4%b8%8e%e5%a4%a7%e5%9e%8b%e8%af%ad%e8%a8%80%e6%a8%a1%e5%9e%8b%e7%94%9f%e6%88%90%e7%9a%84%e5%87%ba","status":"publish","type":"post","link":"https:\/\/csccm.org.cn\/?p=28375","title":{"rendered":"[JAMA Intern Med\u53d1\u8868\u8bba\u6587]\uff1a\u533b\u751f\u64b0\u5199\u4e0e\u5927\u578b\u8bed\u8a00\u6a21\u578b\u751f\u6210\u7684\u51fa\u9662\u8bb0\u5f55"},"content":{"rendered":"\n

Original Investigation <\/p>\n\n\n\n

May 5, 2025<\/p>\n\n\n\n

Physician- and Large Language Model\u2013Generated Hospital Discharge Summaries<\/h1>\n\n\n\n

Christopher Y. K. Williams, Charumathi Raghu Subramanian, Syed Salman Ali, et al<\/h3>\n\n\n\n

JAMA Intern Med. <\/em>Published online May 5, 2025. doi:10.1001\/jamainternmed.2025.0821<\/h3>\n\n\n\n

<\/a>Key Points<\/p>\n\n\n\n

Question<\/strong>  Can large language models (LLMs) draft hospital discharge summary narratives of comparable quality and safety to those written by physicians?<\/p>\n\n\n\n

Findings<\/strong>  In this cross-sectional study of 100 discharge summaries, there was no significant difference in LLM- and physician-generated narratives when rated by blinded reviewers on overall quality and preference. LLM-generated narratives were more concise and coherent than their physician-generated counterparts, but less comprehensive; LLM-generated narratives were also more likely to contain errors, but their overall potential for harm was low.<\/p>\n\n\n\n

Meaning<\/strong>  These findings suggest the potential for LLMs to aid clinicians by drafting discharge summary narratives.<\/a><\/p>\n\n\n\n

Abstract<\/p>\n\n\n\n

Importance<\/strong>  High-quality discharge summaries are associated with improved patient outcomes, but contribute to clinical documentation burden. Large language models (LLMs) provide an opportunity to support physicians by drafting discharge summary narratives.<\/p>\n\n\n\n

Objective<\/strong>  To determine whether LLM-generated discharge summary narratives are of comparable quality and safety to those of physicians.<\/p>\n\n\n\n

Design, Setting, and Participants<\/strong>  This cross-sectional study conducted at the University of California, San Francisco included 100 randomly selected inpatient hospital medicine encounters of 3 to 6 days\u2019 duration between 2019 and 2022. The analysis took place in July 2024.<\/p>\n\n\n\n

Exposure<\/strong>  A blinded evaluation of physician- and LLM-generated narratives was performed in duplicate by 22 attending physician reviewers.<\/p>\n\n\n\n

Main Outcomes and Measures<\/strong>  Narratives were reviewed for overall quality, reviewer preference, comprehensiveness, concision, coherence, and 3 error types (inaccuracies, omissions, and hallucinations). Each error individually, and each narrative overall, were assigned potential harmfulness scores ranging from 0 to 7 on an adapted Agency for Healthcare Research and Quality scale.<\/p>\n\n\n\n

Results<\/strong>\u00a0\u00a0Across 100 encounters, LLM- and physician-generated narratives were comparable in overall quality on a Likert scale ranging from 1 to 5 (higher scores indicate higher quality; mean [SD] score, 3.67 [0.49] vs 3.77 [0.57];\u00a0P<\/em>\u2009=\u2009.21) and reviewer preference (\u03c72<\/sup>\u2009=\u20095.2;\u00a0P<\/em>\u2009=\u2009.27). LLM-generated narratives were more concise (mean [SD] score, 4.01 [0.37] vs 3.70 [0.59];\u00a0P<\/em>\u2009<\u2009.001) and more coherent (mean [SD] score, 4.16 [0.39] vs 4.01 [0.53];\u00a0P<\/em>\u2009=\u2009.02) than their physician-generated counterparts, but less comprehensive (mean [SD] score, 3.72 [0.58] vs 4.13 [0.58];\u00a0P<\/em>\u2009<\u2009.001). LLM-generated narratives contained more unique errors (mean [SD] errors per summary, 2.91 [2.54]) than physician-generated narratives (mean [SD] errors per summary, 1.82 [1.94]). There was no significant difference in the potential for harm between LLM- and physician-generated narratives across individual errors (mean [SD] of 1.35 [1.07] vs 1.34 [1.05];\u00a0P<\/em>\u2009=\u2009.99), with 6 and 5 individual errors, respectively, with scores of 4 (potential for permanent harm) or greater. Both LLM- and physician-generated narratives had low overall potential for harm (scores <1 on a scale ranging from 0-7), with LLM-generated narratives scoring higher than physician narratives (mean [SD] score of 0.84 [0.98] vs 0.36 [0.70];\u00a0P<\/em>\u2009<\u2009.001) and only 1 LLM-generated narrative (compared with 0 physician-generated narratives) scoring 4 or greater.<\/p>\n\n\n\n

\"\"\/<\/figure>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n

Conclusions and Relevance<\/strong>  In this cross-sectional study of 100 inpatient hospital medicine encounters, LLM-generated discharge summary narratives were of comparable quality, and were preferred equally, to those generated by physicians. LLM-generated narratives were more likely to contain errors but had low overall harmfulness scores. These results suggest that, in clinical practice, using such narratives after human review may provide a viable option for hospitalists.<\/p>\n","protected":false},"excerpt":{"rendered":"

Original Investigation  May 5, 2025 Phys […]<\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[32,23],"tags":[],"_links":{"self":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/28375"}],"collection":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=28375"}],"version-history":[{"count":2,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/28375\/revisions"}],"predecessor-version":[{"id":28693,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/28375\/revisions\/28693"}],"wp:attachment":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=28375"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=28375"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=28375"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}