开发和评估机器学习模型,以预测参与社区治疗计划的医疗补助登登者中阿片类药物使用障碍的急性护理。

IF 5.2 1区 医学 Q1 PSYCHIATRY
Addiction Pub Date : 2025-04-29 DOI:10.1111/add.70079
Lingshu Xue, Ruofei Yin, Evan S Cole, Wei-Hsuan Lo-Ciganic, Walid F Gellad, Julie Donohue, Lu Tang
{"title":"开发和评估机器学习模型,以预测参与社区治疗计划的医疗补助登登者中阿片类药物使用障碍的急性护理。","authors":"Lingshu Xue, Ruofei Yin, Evan S Cole, Wei-Hsuan Lo-Ciganic, Walid F Gellad, Julie Donohue, Lu Tang","doi":"10.1111/add.70079","DOIUrl":null,"url":null,"abstract":"<p><strong>Aims: </strong>To develop machine-learning algorithms for predicting the risk of a hospitalization or emergency department (ED) visit for opioid use disorder (OUD) (i.e. OUD acute events) in Pennsylvania Medicaid enrollees in the Opioid Use Disorder Centers of Excellence (COE) program and to evaluate the fairness of model performance across racial groups.</p><p><strong>Methods: </strong>We studied 20 983 United States Medicaid enrollees aged 18 years or older who had COE visits between April 2019 and March 2021. We applied multivariate logistic regression, least absolute shrinkage and selection operator models, random forests, and eXtreme Gradient Boosting (XGB), to predict OUD acute events following the initial COE visit. Our models included predictors at the system, patient, and regional levels. We assessed model performance using multiple metrics by racial groups. Individuals were divided into a low, medium and high-risk group based on predicted risk scores.</p><p><strong>Results: </strong>The training (n = 13 990) and testing (n = 6993) samples displayed similar characteristics (mean age 38.1 ± 9.3 years, 58% male, 80% White enrollees) with 4% experiencing OUD acute events at baseline. XGB demonstrated the best prediction performance (C-statistic = 76.6% [95% confidence interval = 75.6%-77.7%] vs. 72.8%-74.7% for other methods). At the balanced cutoff, XGB achieved a sensitivity of 68.2%, specificity of 70.0%, and positive predictive value of 8.3%. The XGB model classified the testing sample into high-risk (6%), medium-risk (30%), and low-risk (63%) groups. In the high-risk group, 40.7% had OUD acute events vs. 16.5% and 5.0% in the medium- and low-risk groups. The high- and medium-risk groups captured 44% and 26% of individuals with OUD events. The XGB model exhibited lower false negative rates and higher false positive rates in racial/ethnic minority groups than White enrollees.</p><p><strong>Conclusions: </strong>New machine-learning algorithms perform well to predict risks of opioid use disorder (OUD) acute care use among United States Medicaid enrollees and improve fairness of prediction across racial and ethnic groups compared with previous OUD-related models.</p>","PeriodicalId":109,"journal":{"name":"Addiction","volume":" ","pages":""},"PeriodicalIF":5.2000,"publicationDate":"2025-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Development and evaluation of a machine learning model to predict acute care for opioid use disorder among Medicaid enrollees engaged in a community-based treatment program.\",\"authors\":\"Lingshu Xue, Ruofei Yin, Evan S Cole, Wei-Hsuan Lo-Ciganic, Walid F Gellad, Julie Donohue, Lu Tang\",\"doi\":\"10.1111/add.70079\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Aims: </strong>To develop machine-learning algorithms for predicting the risk of a hospitalization or emergency department (ED) visit for opioid use disorder (OUD) (i.e. OUD acute events) in Pennsylvania Medicaid enrollees in the Opioid Use Disorder Centers of Excellence (COE) program and to evaluate the fairness of model performance across racial groups.</p><p><strong>Methods: </strong>We studied 20 983 United States Medicaid enrollees aged 18 years or older who had COE visits between April 2019 and March 2021. We applied multivariate logistic regression, least absolute shrinkage and selection operator models, random forests, and eXtreme Gradient Boosting (XGB), to predict OUD acute events following the initial COE visit. Our models included predictors at the system, patient, and regional levels. We assessed model performance using multiple metrics by racial groups. Individuals were divided into a low, medium and high-risk group based on predicted risk scores.</p><p><strong>Results: </strong>The training (n = 13 990) and testing (n = 6993) samples displayed similar characteristics (mean age 38.1 ± 9.3 years, 58% male, 80% White enrollees) with 4% experiencing OUD acute events at baseline. XGB demonstrated the best prediction performance (C-statistic = 76.6% [95% confidence interval = 75.6%-77.7%] vs. 72.8%-74.7% for other methods). At the balanced cutoff, XGB achieved a sensitivity of 68.2%, specificity of 70.0%, and positive predictive value of 8.3%. The XGB model classified the testing sample into high-risk (6%), medium-risk (30%), and low-risk (63%) groups. In the high-risk group, 40.7% had OUD acute events vs. 16.5% and 5.0% in the medium- and low-risk groups. The high- and medium-risk groups captured 44% and 26% of individuals with OUD events. The XGB model exhibited lower false negative rates and higher false positive rates in racial/ethnic minority groups than White enrollees.</p><p><strong>Conclusions: </strong>New machine-learning algorithms perform well to predict risks of opioid use disorder (OUD) acute care use among United States Medicaid enrollees and improve fairness of prediction across racial and ethnic groups compared with previous OUD-related models.</p>\",\"PeriodicalId\":109,\"journal\":{\"name\":\"Addiction\",\"volume\":\" \",\"pages\":\"\"},\"PeriodicalIF\":5.2000,\"publicationDate\":\"2025-04-29\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Addiction\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1111/add.70079\",\"RegionNum\":1,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"PSYCHIATRY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Addiction","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1111/add.70079","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"PSYCHIATRY","Score":null,"Total":0}
引用次数: 0

摘要

目的:开发机器学习算法,用于预测宾夕法尼亚州阿片类药物使用障碍卓越中心(COE)计划中阿片类药物使用障碍(OUD)(即OUD急性事件)的住院或急诊科(ED)就诊风险,并评估模型跨种族表现的公平性。方法:我们研究了20983名18岁或以上的美国医疗补助参保人,他们在2019年4月至2021年3月期间进行了COE就诊。我们应用多元逻辑回归、最小绝对收缩和选择算子模型、随机森林和极端梯度增强(XGB)来预测首次COE就诊后的OUD急性事件。我们的模型包括系统、患者和区域层面的预测因子。我们使用种族群体的多个指标来评估模型的性能。根据预测的风险评分,将个体分为低、中、高风险组。结果:训练(n = 13990)和测试(n = 6993)样本显示出相似的特征(平均年龄38.1±9.3岁,58%为男性,80%为白人),其中4%在基线时出现OUD急性事件。XGB预测效果最好(C-statistic = 76.6%[95%置信区间= 75.6% ~ 77.7%],其他方法为72.8% ~ 74.7%)。在平衡截止点,XGB的敏感性为68.2%,特异性为70.0%,阳性预测值为8.3%。XGB模型将测试样本分为高风险(6%)、中风险(30%)和低风险(63%)组。在高危组中,40.7%发生OUD急性事件,中、低危组分别为16.5%和5.0%。高风险和中等风险组分别占OUD事件患者的44%和26%。XGB模型显示,与白人受试者相比,少数种族/民族受试者的假阴性率较低,假阳性率较高。结论:与之前的阿片类药物使用相关模型相比,新的机器学习算法在预测美国医疗补助计划参选者阿片类药物使用障碍(OUD)急性护理使用风险方面表现良好,并提高了跨种族和族裔群体预测的公平性。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Development and evaluation of a machine learning model to predict acute care for opioid use disorder among Medicaid enrollees engaged in a community-based treatment program.

Aims: To develop machine-learning algorithms for predicting the risk of a hospitalization or emergency department (ED) visit for opioid use disorder (OUD) (i.e. OUD acute events) in Pennsylvania Medicaid enrollees in the Opioid Use Disorder Centers of Excellence (COE) program and to evaluate the fairness of model performance across racial groups.

Methods: We studied 20 983 United States Medicaid enrollees aged 18 years or older who had COE visits between April 2019 and March 2021. We applied multivariate logistic regression, least absolute shrinkage and selection operator models, random forests, and eXtreme Gradient Boosting (XGB), to predict OUD acute events following the initial COE visit. Our models included predictors at the system, patient, and regional levels. We assessed model performance using multiple metrics by racial groups. Individuals were divided into a low, medium and high-risk group based on predicted risk scores.

Results: The training (n = 13 990) and testing (n = 6993) samples displayed similar characteristics (mean age 38.1 ± 9.3 years, 58% male, 80% White enrollees) with 4% experiencing OUD acute events at baseline. XGB demonstrated the best prediction performance (C-statistic = 76.6% [95% confidence interval = 75.6%-77.7%] vs. 72.8%-74.7% for other methods). At the balanced cutoff, XGB achieved a sensitivity of 68.2%, specificity of 70.0%, and positive predictive value of 8.3%. The XGB model classified the testing sample into high-risk (6%), medium-risk (30%), and low-risk (63%) groups. In the high-risk group, 40.7% had OUD acute events vs. 16.5% and 5.0% in the medium- and low-risk groups. The high- and medium-risk groups captured 44% and 26% of individuals with OUD events. The XGB model exhibited lower false negative rates and higher false positive rates in racial/ethnic minority groups than White enrollees.

Conclusions: New machine-learning algorithms perform well to predict risks of opioid use disorder (OUD) acute care use among United States Medicaid enrollees and improve fairness of prediction across racial and ethnic groups compared with previous OUD-related models.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Addiction
Addiction 医学-精神病学
CiteScore
10.80
自引率
6.70%
发文量
319
审稿时长
3 months
期刊介绍: Addiction publishes peer-reviewed research reports on pharmacological and behavioural addictions, bringing together research conducted within many different disciplines. Its goal is to serve international and interdisciplinary scientific and clinical communication, to strengthen links between science and policy, and to stimulate and enhance the quality of debate. We seek submissions that are not only technically competent but are also original and contain information or ideas of fresh interest to our international readership. We seek to serve low- and middle-income (LAMI) countries as well as more economically developed countries. Addiction’s scope spans human experimental, epidemiological, social science, historical, clinical and policy research relating to addiction, primarily but not exclusively in the areas of psychoactive substance use and/or gambling. In addition to original research, the journal features editorials, commentaries, reviews, letters, and book reviews.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信