2114-P: Artificial Intelligence–Assisted Identification of Severe Hypoglycemia Risk Factors in Pediatric Type 1 Diabetes

IF 9.3 1区医学 Q1 ENDOCRINOLOGY & METABOLISM

Diabetes Pub Date : 2026-06-11 DOI:10.2337/db26-2114-p

NATALI BORREGO, DANIEL PALACIOS, HU CHEN, ZHANDONG LIU, DANIEL J. DESALVO, REBECCA S. AGUIRRE, MARIA J. REDONDO, MUSTAFA TOSUR, MICHAEL L. FERM

{"title":"2114-P: Artificial Intelligence–Assisted Identification of Severe Hypoglycemia Risk Factors in Pediatric Type 1 Diabetes","authors":"NATALI BORREGO, DANIEL PALACIOS, HU CHEN, ZHANDONG LIU, DANIEL J. DESALVO, REBECCA S. AGUIRRE, MARIA J. REDONDO, MUSTAFA TOSUR, MICHAEL L. FERM","doi":"10.2337/db26-2114-p","DOIUrl":null,"url":null,"abstract":"Introduction and Objective: Severe hypoglycemia (SH) is a life-threatening complication of type 1 diabetes (T1D), making accurate risk identification essential. We used artificial intelligence (AI)-based chart review to improve identification of factors associated with SH in children with T1D. Methods: We studied children (&lt;18y) with T1D seen in endocrinology clinics at a large tertiary health center between April 2020 and April 2025 (n = 4,018). Clinicians were required to screen for SH at each visit using questionnaires built into the medical chart. Data scientists employed a hybrid AI-assisted and traditional approach to extract clinical, socioeconomic, and CGM data using PySpark transformations, ChatGPT 5.1 within Palantir Foundry, and custom Python scripts. Groups with and without SH were compared using Fisher's exact, χ² and Mann-Whitney tests. Select multivariate regressions were performed. Results: Clinician-documented SH occurred in 5% of patients (206/4018). Table 1 presents group comparisons. Controlling for insurance type, Black race was no longer significantly associated with SH (p=0.07). Conclusion: AI-assisted analyses of SH reflect established disparities in glycemic control and diabetes technology use, possibly associated with insurance coverage. Future AI-generated predictive models may improve risk assessment and guide focused interventions. Disclosure N. Borrego: None. D. Palacios: None. H. Chen: None. Z. Liu: None. D.J. DeSalvo: Consultant; Current; Dexcom, Inc., Insulet Corporation. Advisory Panel; Current; Sanofi. Research Support; Current; Medtronic. R.S. Aguirre: Advisory Panel; Ended; Sanofi. Research Support; Current; Sanofi. M.J. Redondo: Advisory Panel; Current; Sanofi. Other - Data Safety Monitoring committee; Current; Lilly. M. Tosur: None. M.L. Ferm: Advisory Panel; Ended; Rezolute Inc.","PeriodicalId":11376,"journal":{"name":"Diabetes","volume":"11 1","pages":""},"PeriodicalIF":9.3000,"publicationDate":"2026-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Diabetes","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.2337/db26-2114-p","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENDOCRINOLOGY & METABOLISM","Score":null,"Total":0}

引用次数: 0

Abstract

Introduction and Objective: Severe hypoglycemia (SH) is a life-threatening complication of type 1 diabetes (T1D), making accurate risk identification essential. We used artificial intelligence (AI)-based chart review to improve identification of factors associated with SH in children with T1D. Methods: We studied children (<18y) with T1D seen in endocrinology clinics at a large tertiary health center between April 2020 and April 2025 (n = 4,018). Clinicians were required to screen for SH at each visit using questionnaires built into the medical chart. Data scientists employed a hybrid AI-assisted and traditional approach to extract clinical, socioeconomic, and CGM data using PySpark transformations, ChatGPT 5.1 within Palantir Foundry, and custom Python scripts. Groups with and without SH were compared using Fisher's exact, χ² and Mann-Whitney tests. Select multivariate regressions were performed. Results: Clinician-documented SH occurred in 5% of patients (206/4018). Table 1 presents group comparisons. Controlling for insurance type, Black race was no longer significantly associated with SH (p=0.07). Conclusion: AI-assisted analyses of SH reflect established disparities in glycemic control and diabetes technology use, possibly associated with insurance coverage. Future AI-generated predictive models may improve risk assessment and guide focused interventions. Disclosure N. Borrego: None. D. Palacios: None. H. Chen: None. Z. Liu: None. D.J. DeSalvo: Consultant; Current; Dexcom, Inc., Insulet Corporation. Advisory Panel; Current; Sanofi. Research Support; Current; Medtronic. R.S. Aguirre: Advisory Panel; Ended; Sanofi. Research Support; Current; Sanofi. M.J. Redondo: Advisory Panel; Current; Sanofi. Other - Data Safety Monitoring committee; Current; Lilly. M. Tosur: None. M.L. Ferm: Advisory Panel; Ended; Rezolute Inc.

查看原文本刊更多论文

2114-P：儿童1型糖尿病严重低血糖危险因素的人工智能辅助识别

简介与目的：严重低血糖（SH）是危及生命的1型糖尿病（T1D）并发症，准确的风险识别至关重要。我们使用基于人工智能（AI）的图表审查来提高对T1D儿童SH相关因素的识别。方法：我们研究了2020年4月至2025年4月期间在一家大型三级卫生中心内分泌科诊所就诊的T1D儿童（<18y）（n = 4,018）。临床医生被要求在每次就诊时使用内置在医疗图表中的问卷对SH进行筛查。数据科学家采用人工智能辅助和传统的混合方法，使用PySpark转换、Palantir Foundry中的ChatGPT 5.1和自定义Python脚本提取临床、社会经济和CGM数据。采用Fisher精确检验、χ 2检验和Mann-Whitney检验比较有无SH组。选择多元回归进行分析。结果：5%的患者发生临床记录的SH（206/4018）。表1给出了组间比较。控制保险类型后，黑人种族与SH不再显著相关（p=0.07）。结论：人工智能辅助的SH分析反映了血糖控制和糖尿病技术使用的既定差异，可能与保险覆盖范围有关。未来人工智能生成的预测模型可能会改进风险评估并指导有针对性的干预措施。博雷戈：没有。帕拉西奥斯：没有。H. Chen：没有。刘：没有。D.J.德萨佛：顾问；当前的;Dexcom, Inc., Insulet Corporation。顾问小组;当前的;赛诺菲。研究支持;当前的;美敦力公司。R.S. Aguirre：顾问小组；结束;赛诺菲。研究支持;当前的;赛诺菲。M.J. Redondo：顾问小组；当前的;赛诺菲。其他-数据安全监察委员会；当前的;莉莉。托苏尔：没有。M.L. Ferm：顾问小组；结束;Rezolute Inc .)

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Diabetes 医学-内分泌学与代谢

CiteScore

12.50

自引率

2.60%

发文量

1968

审稿时长

1 months

期刊介绍： Diabetes is a scientific journal that publishes original research exploring the physiological and pathophysiological aspects of diabetes mellitus. We encourage submissions of manuscripts pertaining to laboratory, animal, or human research, covering a wide range of topics. Our primary focus is on investigative reports investigating various aspects such as the development and progression of diabetes, along with its associated complications. We also welcome studies delving into normal and pathological pancreatic islet function and intermediary metabolism, as well as exploring the mechanisms of drug and hormone action from a pharmacological perspective. Additionally, we encourage submissions that delve into the biochemical and molecular aspects of both normal and abnormal biological processes. However, it is important to note that we do not publish studies relating to diabetes education or the application of accepted therapeutic and diagnostic approaches to patients with diabetes mellitus. Our aim is to provide a platform for research that contributes to advancing our understanding of the underlying mechanisms and processes of diabetes.