Dynamic updating of psychosis prediction models in individuals at ultra high-risk of psychosis.

Simon Hartmann, Dominic Dwyer, Isabelle Scott, Cassandra M J Wannan, Josh Nguyen, Ashleigh Lin, Christel M Middeldorp, Stephen J Wood, Alison R Yung, Patrick D McGorry, Barnaby Nelson, Scott R Clark
{"title":"Dynamic updating of psychosis prediction models in individuals at ultra high-risk of psychosis.","authors":"Simon Hartmann, Dominic Dwyer, Isabelle Scott, Cassandra M J Wannan, Josh Nguyen, Ashleigh Lin, Christel M Middeldorp, Stephen J Wood, Alison R Yung, Patrick D McGorry, Barnaby Nelson, Scott R Clark","doi":"10.1016/j.bpsc.2025.03.006","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>The performance of psychiatric risk calculators can deteriorate over time due to changes in patient population, referral pathways, and medical advances. Such temporal biases in existing models may lead to suboptimal decisions when translated into clinical practice. Methods are available to correct this bias, yet no research has been conducted to investigate their utility in psychiatry.</p><p><strong>Methods: </strong>We aimed to analyse the performance of model updating methods for predicting psychosis onset by one year in 784 individuals at ultra high-risk (UHR) of psychosis from the UHR 1000+ cohort - a longitudinal cohort of UHR individuals recruited to research studies at Orygen, Melbourne, Australia, between 1995 and 2020. Model updating was performed using a yearly adjusted model (recalibration), a continuously updated model (refitting), and a continuous Bayesian updating model (dynamic updating) and compared to a static logistic regression prediction model (original) regarding calibration, discrimination, and clinical net benefit.</p><p><strong>Results: </strong>The original model was poorly calibrated over the entire validation period. All three updating methods improved the predictive performance compared to the original model (recalibration: P= 0.014, refitting: P= 0.028, dynamic updating: P= 0.002). The dynamic updating method demonstrated the best predictive performance (Harrel's C-index = 0.70, 95% CI: [0.58, 0.81]), calibration slope (slope = 1.03, 95% CI: [0.38, 1.74]) and clinical net benefit over the entire validation period.</p><p><strong>Conclusions: </strong>Dynamic updating of psychosis prediction models may help to mitigate decreases in performance over time. Hence, existing psychosis prediction models need to be monitored for temporal biases to mitigate potentially harmful decisions.</p>","PeriodicalId":93900,"journal":{"name":"Biological psychiatry. Cognitive neuroscience and neuroimaging","volume":" ","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Biological psychiatry. Cognitive neuroscience and neuroimaging","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1016/j.bpsc.2025.03.006","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Background: The performance of psychiatric risk calculators can deteriorate over time due to changes in patient population, referral pathways, and medical advances. Such temporal biases in existing models may lead to suboptimal decisions when translated into clinical practice. Methods are available to correct this bias, yet no research has been conducted to investigate their utility in psychiatry.

Methods: We aimed to analyse the performance of model updating methods for predicting psychosis onset by one year in 784 individuals at ultra high-risk (UHR) of psychosis from the UHR 1000+ cohort - a longitudinal cohort of UHR individuals recruited to research studies at Orygen, Melbourne, Australia, between 1995 and 2020. Model updating was performed using a yearly adjusted model (recalibration), a continuously updated model (refitting), and a continuous Bayesian updating model (dynamic updating) and compared to a static logistic regression prediction model (original) regarding calibration, discrimination, and clinical net benefit.

Results: The original model was poorly calibrated over the entire validation period. All three updating methods improved the predictive performance compared to the original model (recalibration: P= 0.014, refitting: P= 0.028, dynamic updating: P= 0.002). The dynamic updating method demonstrated the best predictive performance (Harrel's C-index = 0.70, 95% CI: [0.58, 0.81]), calibration slope (slope = 1.03, 95% CI: [0.38, 1.74]) and clinical net benefit over the entire validation period.

Conclusions: Dynamic updating of psychosis prediction models may help to mitigate decreases in performance over time. Hence, existing psychosis prediction models need to be monitored for temporal biases to mitigate potentially harmful decisions.

精神病超高危人群精神病预测模型的动态更新。
背景:由于患者群体、转诊途径和医学进步的变化,精神病风险计算器的性能会随着时间的推移而恶化。现有模型中的这种时间偏差在转化为临床实践时可能导致次优决策。纠正这种偏见的方法是可用的,但还没有研究调查它们在精神病学中的效用。方法:我们的目的是分析模型更新方法在预测一年内精神病发病的表现,这些方法来自UHR 1000+队列(UHR 1000+队列是1995年至2020年在澳大利亚墨尔本Orygen进行研究的UHR个体的纵向队列)。使用年度调整模型(重新校准)、连续更新模型(重新校正)和连续贝叶斯更新模型(动态更新)进行模型更新,并与静态逻辑回归预测模型(原始)在校准、鉴别和临床净效益方面进行比较。结果:原始模型在整个验证期内校准得很差。与原始模型相比,三种更新方法均提高了预测性能(重新校准:P= 0.014,改装:P= 0.028,动态更新:P= 0.002)。动态更新方法在整个验证期内表现出最佳的预测性能(Harrel's C-index = 0.70, 95% CI:[0.58, 0.81])、校准斜率(斜率= 1.03,95% CI:[0.38, 1.74])和临床净效益。结论:随着时间的推移,精神病预测模型的动态更新可能有助于减轻表现的下降。因此,现有的精神病预测模型需要监测时间偏差,以减轻潜在的有害决策。
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
自引率
0.00%
发文量
0
×
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