PMID- 38062850 OWN - NLM STAT- MEDLINE DCOM- 20240122 LR - 20240131 IS - 1527-974X (Electronic) IS - 1067-5027 (Print) IS - 1067-5027 (Linking) VI - 31 IP - 2 DP - 2024 Jan 18 TI - Applying contrastive pre-training for depression and anxiety risk prediction in type 2 diabetes patients based on heterogeneous electronic health records: a primary healthcare case study. PG - 445-455 LID - 10.1093/jamia/ocad228 [doi] AB - OBJECTIVE: Due to heterogeneity and limited medical data in primary healthcare services (PHS), assessing the psychological risk of type 2 diabetes mellitus (T2DM) patients in PHS is difficult. Using unsupervised contrastive pre-training, we proposed a deep learning framework named depression and anxiety prediction (DAP) to predict depression and anxiety in T2DM patients. MATERIALS AND METHODS: The DAP model consists of two sub-models. Firstly, the pre-trained model of DAP used unlabeled discharge records of 85 085 T2DM patients from the First Affiliated Hospital of Nanjing Medical University for unsupervised contrastive learning on heterogeneous electronic health records (EHRs). Secondly, the fine-tuned model of DAP used case-control cohorts (17 491 patients) selected from 149 596 T2DM patients' EHRs in the Nanjing Health Information Platform (NHIP). The DAP model was validated in 1028 patients from PHS in NHIP. Evaluation included receiver operating characteristic area under the curve (ROC-AUC) and precision-recall area under the curve (PR-AUC), and decision curve analysis (DCA). RESULTS: The pre-training step allowed the DAP model to converge at a faster rate. The fine-tuned DAP model significantly outperformed the baseline models (logistic regression, extreme gradient boosting, and random forest) with ROC-AUC of 0.91+/-0.028 and PR-AUC of 0.80+/-0.067 in 10-fold internal validation, and with ROC-AUC of 0.75 +/- 0.045 and PR-AUC of 0.47 +/- 0.081 in external validation. The DCA indicate the clinical potential of the DAP model. CONCLUSION: The DAP model effectively predicted post-discharge depression and anxiety in T2DM patients from PHS, reducing data fragmentation and limitations. This study highlights the DAP model's potential for early detection and intervention in depression and anxiety, improving outcomes for diabetes patients. CI - (c) The Author(s) 2023. Published by Oxford University Press on behalf of the American Medical Informatics Association. FAU - Feng, Wei AU - Feng W AUID- ORCID: 0000-0002-6843-2067 AD - Department of Medical Informatics, School of Biomedical Engineering and Informatics, Nanjing Medical University, Nanjing, Jiangsu, 210009, China. FAU - Wu, Honghan AU - Wu H AUID- ORCID: 0000-0002-0213-5668 AD - Institute of Health Informatics, University College London, London, WC1E 6BT, United Kingdom. AD - The Alan Turing Institute, London, NW1 2DB, United Kingdom. FAU - Ma, Hui AU - Ma H AD - Department of Medical Psychology, Nanjing Brain Hospital affiliated with Nanjing Medical University, Nanjing, Jiangsu, 210024, China. FAU - Tao, Zhenhuan AU - Tao Z AD - Department of Planning, Nanjing Health Information Center, Nanjing, Jiangsu, 210003, China. FAU - Xu, Mengdie AU - Xu M AD - Department of Medical Informatics, School of Biomedical Engineering and Informatics, Nanjing Medical University, Nanjing, Jiangsu, 210009, China. FAU - Zhang, Xin AU - Zhang X AD - Department of Medical Informatics, School of Biomedical Engineering and Informatics, Nanjing Medical University, Nanjing, Jiangsu, 210009, China. AD - Department of Information, The First Affiliated Hospital, Nanjing Medical University, Nanjing, Jiangsu, 210029, China. FAU - Lu, Shan AU - Lu S AD - Department of Medical Informatics, School of Biomedical Engineering and Informatics, Nanjing Medical University, Nanjing, Jiangsu, 210009, China. AD - Department of Information, The First Affiliated Hospital, Nanjing Medical University, Nanjing, Jiangsu, 210029, China. FAU - Wan, Cheng AU - Wan C AD - Department of Medical Informatics, School of Biomedical Engineering and Informatics, Nanjing Medical University, Nanjing, Jiangsu, 210009, China. FAU - Liu, Yun AU - Liu Y AD - Department of Medical Informatics, School of Biomedical Engineering and Informatics, Nanjing Medical University, Nanjing, Jiangsu, 210009, China. AD - Department of Information, The First Affiliated Hospital, Nanjing Medical University, Nanjing, Jiangsu, 210029, China. LA - eng GR - Nanjing Life and Health Technology/ GR - Nanjing City Health Science and Technology/ GR - Jiangsu Provincial Health Commission/ GR - NIHR202639/National Institute for Health Research/ GR - British Council/ GR - 202205053/Nanjing Life and Health Technology Special Project "Cooperative/ GR - YKK23197/Nanjing City Health Science and Technology Development Special Fund in 2023/ GR - Jiangsu Provincial Health Commission's medical/ GR - MR/S004149/1/UK's Medical Research Council/ GR - Legal & General Group/ GR - Care Research Centre at University of Edinburgh/ PT - Journal Article PL - England TA - J Am Med Inform Assoc JT - Journal of the American Medical Informatics Association : JAMIA JID - 9430800 SB - IM MH - Humans MH - *Diabetes Mellitus, Type 2/complications/diagnosis MH - Electronic Health Records MH - Aftercare MH - Depression MH - Machine Learning MH - Patient Discharge MH - Anxiety PMC - PMC10797279 OTO - NOTNLM OT - EHR pre-trained model OT - deep learning OT - depression and anxiety OT - regional EHRs OT - type 2 diabetes mellitus COIS- No competing interest is declared. EDAT- 2023/12/08 06:42 MHDA- 2024/01/22 06:43 PMCR- 2023/12/07 CRDT- 2023/12/08 03:43 PHST- 2023/07/31 00:00 [received] PHST- 2023/11/13 00:00 [revised] PHST- 2023/11/21 00:00 [accepted] PHST- 2024/01/22 06:43 [medline] PHST- 2023/12/08 06:42 [pubmed] PHST- 2023/12/08 03:43 [entrez] PHST- 2023/12/07 00:00 [pmc-release] AID - 7462201 [pii] AID - ocad228 [pii] AID - 10.1093/jamia/ocad228 [doi] PST - ppublish SO - J Am Med Inform Assoc. 2024 Jan 18;31(2):445-455. doi: 10.1093/jamia/ocad228.