Jamie J Coleman, Jolene Atia, Felicity Evison, Lydia Wilson, Suzy Gallier, Richard Sames, Andrew Capewell, Richard Copley, Helen Gyves, Simon Ball, Tanya Pankhurst
{"title":"Adoption by clinicians of electronic order communications in NHS secondary care: a descriptive account.","authors":"Jamie J Coleman, Jolene Atia, Felicity Evison, Lydia Wilson, Suzy Gallier, Richard Sames, Andrew Capewell, Richard Copley, Helen Gyves, Simon Ball, Tanya Pankhurst","doi":"10.1136/bmjhci-2023-100850","DOIUrl":"10.1136/bmjhci-2023-100850","url":null,"abstract":"<p><strong>Background: </strong>Due to the rapid advancement in information technology, changes to communication modalities are increasingly implemented in healthcare. One such modality is Computerised Provider Order Entry (CPOE) systems which replace paper, verbal or telephone orders with electronic booking of requests. We aimed to understand the uptake, and user acceptability, of CPOE in a large National Health Service hospital system.</p><p><strong>Methods: </strong>This retrospective single-centre study investigates the longitudinal uptake of communications through the Prescribing, Information and Communication System (PICS). The development and configuration of PICS are led by the doctors, nurses and allied health professionals that use it and requests for CPOE driven by clinical need have been described.Records of every request (imaging, specialty review, procedure, laboratory) made through PICS were collected between October 2008 and July 2019 and resulting counts were presented. An estimate of the proportion of completed requests made through the system has been provided for three example requests. User surveys were completed.</p><p><strong>Results: </strong>In the first 6 months of implementation, a total of 832 new request types (imaging types and specialty referrals) were added to the system. Subsequently, an average of 6.6 new request types were added monthly. In total, 8 035 132 orders were requested through PICS. In three example request types (imaging, endoscopy and full blood count), increases in the proportion of requests being made via PICS were seen. User feedback at 6 months reported improved communications using the electronic system.</p><p><strong>Conclusion: </strong>CPOE was popular, rapidly adopted and diversified across specialties encompassing wide-ranging requests.</p>","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11097811/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140903965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Phung-Anh Nguyen, Min-Huei Hsu, Tzu-Hao Chang, Hsuan-Chia Yang, Chih-Wei Huang, Chia-Te Liao, Christine Y. Lu, Jason C. Hsu
{"title":"Taipei Medical University Clinical Research Database: a collaborative hospital EHR database aligned with international common data standards","authors":"Phung-Anh Nguyen, Min-Huei Hsu, Tzu-Hao Chang, Hsuan-Chia Yang, Chih-Wei Huang, Chia-Te Liao, Christine Y. Lu, Jason C. Hsu","doi":"10.1136/bmjhci-2023-100890","DOIUrl":"https://doi.org/10.1136/bmjhci-2023-100890","url":null,"abstract":"Objective The objective of this paper is to provide a comprehensive overview of the development and features of the Taipei Medical University Clinical Research Database (TMUCRD), a repository of real-world data (RWD) derived from electronic health records (EHRs) and other sources. Methods TMUCRD was developed by integrating EHRs from three affiliated hospitals, including Taipei Medical University Hospital, Wan-Fang Hospital and Shuang-Ho Hospital. The data cover over 15 years and include diverse patient care information. The database was converted to the Observational Medical Outcomes Partnership Common Data Model (OMOP CDM) for standardisation. Results TMUCRD comprises 89 tables (eg, 29 tables for each hospital and 2 linked tables), including demographics, diagnoses, medications, procedures and measurements, among others. It encompasses data from more than 4.15 million patients with various medical records, spanning from the year 2004 to 2021. The dataset offers insights into disease prevalence, medication usage, laboratory tests and patient characteristics. Discussion TMUCRD stands out due to its unique advantages, including diverse data types, comprehensive patient information, linked mortality and cancer registry data, regular updates and a swift application process. Its compatibility with the OMOP CDM enhances its usability and interoperability. Conclusion TMUCRD serves as a valuable resource for researchers and scholars interested in leveraging RWD for clinical research. Its availability and integration of diverse healthcare data contribute to a collaborative and data-driven approach to advancing medical knowledge and practice. All data relevant to the study are included in the article or uploaded as online supplemental information.","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140925316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ana Karen Macias Alonso, Julian Hirt, Tim Woelfle, Perrine Janiaud, Lars G Hemkens
{"title":"Definitions of digital biomarkers: a systematic mapping of the biomedical literature","authors":"Ana Karen Macias Alonso, Julian Hirt, Tim Woelfle, Perrine Janiaud, Lars G Hemkens","doi":"10.1136/bmjhci-2023-100914","DOIUrl":"https://doi.org/10.1136/bmjhci-2023-100914","url":null,"abstract":"Background Technological devices such as smartphones, wearables and virtual assistants enable health data collection, serving as digital alternatives to conventional biomarkers. We aimed to provide a systematic overview of emerging literature on ‘digital biomarkers,’ covering definitions, features and citations in biomedical research. Methods We analysed all articles in PubMed that used ‘digital biomarker(s)’ in title or abstract, considering any study involving humans and any review, editorial, perspective or opinion-based articles up to 8 March 2023. We systematically extracted characteristics of publications and research studies, and any definitions and features of ‘digital biomarkers’ mentioned. We described the most influential literature on digital biomarkers and their definitions using thematic categorisations of definitions considering the Food and Drug Administration Biomarkers, EndpointS and other Tools framework (ie, data type, data collection method, purpose of biomarker), analysing structural similarity of definitions by performing text and citation analyses. Results We identified 415 articles using ‘digital biomarker’ between 2014 and 2023 (median 2021). The majority (283 articles; 68%) were primary research. Notably, 287 articles (69%) did not provide a definition of digital biomarkers. Among the 128 articles with definitions, there were 127 different ones. Of these, 78 considered data collection, 56 data type, 50 purpose and 23 included all three components. Those 128 articles with a definition had a median of 6 citations, with the top 10 each presenting distinct definitions. Conclusions The definitions of digital biomarkers vary significantly, indicating a lack of consensus in this emerging field. Our overview highlights key defining characteristics, which could guide the development of a more harmonised accepted definition. All data relevant to the study are included in the article or uploaded as online supplemental information.","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140571872","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Chao Guo, Dianqi Yuan, Huameng Tang, Xiyuan Hu, Yiyang Lei
{"title":"Impact of a pandemic shock on unmet medical needs of middle-aged and older adults in 10 countries","authors":"Chao Guo, Dianqi Yuan, Huameng Tang, Xiyuan Hu, Yiyang Lei","doi":"10.1136/bmjhci-2023-100865","DOIUrl":"https://doi.org/10.1136/bmjhci-2023-100865","url":null,"abstract":"Objective The objective is to explore the impact of the pandemic shock on the unmet medical needs of middle-aged and older adults worldwide. Methods The COVID-19 pandemic starting in 2020 was used as a quasiexperiment. Exposure to the pandemic was defined based on an individual’s context within the global pandemic. Data were obtained from the Integrated Values Surveys. A total of 11 932 middle-aged and older adults aged 45 years and above from 10 countries where the surveys conducted two times during 2011 and 2022 were analysed. We used logistic regression models with the difference-in-difference method to estimate the impact of pandemic exposure on unmet medical needs by comparing differences before and after the pandemic across areas with varying degrees of severity. Results Among the 11 932 middle-aged and older adults, 3647 reported unmet medical needs, with a pooled unmet rate of 30.56% (95% CI: 29.74% to 31.40%). The pandemic significantly increased the risk of unmet medical needs among middle-aged and older adults (OR: 2.33, 95% CI: 1.94 to 2.79). The deleterious effect of the pandemic on unmet medical needs was prevalent among middle-aged adults (2.53, 2.00 to 3.20) and older adults (2.00, 1.48 to 2.69), as well as among men (2.24, 1.74 to 2.90) and women (2.34, 1.82 to 3.03). The results remained robust in a series of sensitivity analyses. Conclusion These findings suggest that efforts should be made by policymakers and healthcare professionals to balance healthcare resources to adequately address the comprehensive healthcare demands of individuals regarding multiple health issues, taking into account the challenges posed by pandemics. Data are available in a public, open access repository. This study is based on publicly available datasets, and the data were released to the researchers without access to any personal information from the website: <https://www.worldvaluessurvey.org/WVSEVStrend.jsp>.","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140571873","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Building a house without foundations? A 24-country qualitative interview study on artificial intelligence in intensive care medicine","authors":"Stuart McLennan, Amelia Fiske, Leo Anthony Celi","doi":"10.1136/bmjhci-2024-101052","DOIUrl":"https://doi.org/10.1136/bmjhci-2024-101052","url":null,"abstract":"Objectives To explore the views of intensive care professionals in high-income countries (HICs) and lower-to-middle-income countries (LMICs) regarding the use and implementation of artificial intelligence (AI) technologies in intensive care units (ICUs). Methods Individual semi-structured qualitative interviews were conducted between December 2021 and August 2022 with 59 intensive care professionals from 24 countries. Transcripts were analysed using conventional content analysis. Results Participants had generally positive views about the potential use of AI in ICUs but also reported some well-known concerns about the use of AI in clinical practice and important technical and non-technical barriers to the implementation of AI. Important differences existed between ICUs regarding their current readiness to implement AI. However, these differences were not primarily between HICs and LMICs, but between a small number of ICUs in large tertiary hospitals in HICs, which were reported to have the necessary digital infrastructure for AI, and nearly all other ICUs in both HICs and LMICs, which were reported to neither have the technical capability to capture the necessary data or use AI, nor the staff with the right knowledge and skills to use the technology. Conclusion Pouring massive amounts of resources into developing AI without first building the necessary digital infrastructure foundation needed for AI is unethical. Real-world implementation and routine use of AI in the vast majority of ICUs in both HICs and LMICs included in our study is unlikely to occur any time soon. ICUs should not be using AI until certain preconditions are met. Data are available upon reasonable request. Our data include pseudonymised transcripts of interviews, which cannot be made publicly available in their entirety because of (1) the terms of our ethics approval; and (2) because participants could be identifiable if placed in the context of the entire transcript. This is in line with current ethical expectations for qualitative interview research. We provide anonymised quotes within the paper to illustrate our findings (corresponding to transcript excerpts), and the complete interview guide used in the study has been included as a Supplementary Information.","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140627006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Anh Trung Hoang, Phung-Anh Nguyen, Thanh Phuc Phan, Gia Tuyen Do, Huu Dung Nguyen, I-Jen Chiu, Chu-Lin Chou, Yu-Chen Ko, Tzu-Hao Chang, Chih-Wei Huang, Usman Iqbal, Yung-Ho Hsu, Mai-Szu Wu, Chia-Te Liao
{"title":"Personalised prediction of maintenance dialysis initiation in patients with chronic kidney disease stages 3–5: a multicentre study using the machine learning approach","authors":"Anh Trung Hoang, Phung-Anh Nguyen, Thanh Phuc Phan, Gia Tuyen Do, Huu Dung Nguyen, I-Jen Chiu, Chu-Lin Chou, Yu-Chen Ko, Tzu-Hao Chang, Chih-Wei Huang, Usman Iqbal, Yung-Ho Hsu, Mai-Szu Wu, Chia-Te Liao","doi":"10.1136/bmjhci-2023-100893","DOIUrl":"https://doi.org/10.1136/bmjhci-2023-100893","url":null,"abstract":"Background Optimal timing for initiating maintenance dialysis in patients with chronic kidney disease (CKD) stages 3–5 is challenging. This study aimed to develop and validate a machine learning (ML) model for early personalised prediction of maintenance dialysis initiation within 1-year and 3-year timeframes among patients with CKD stages 3–5. Methods Retrospective electronic health record data from the Taipei Medical University clinical research database were used. Newly diagnosed patients with CKD stages 3–5 between 2008 and 2017 were identified. The observation period spanned from the diagnosis of CKD stages 3–5 until the maintenance dialysis initiation or a maximum follow-up of 3 years. Predictive models were developed using patient demographics, comorbidities, laboratory data and medications. The dataset was divided into training and testing sets to ensure robust model performance. Model evaluation metrics, including area under the curve (AUC), sensitivity, specificity, positive predictive value, negative predictive value and F1 score, were employed. Results A total of 6123 and 5279 patients were included for 1 year and 3 years of the model development. The artificial neural network demonstrated better performance in predicting maintenance dialysis initiation within 1 year and 3 years, with AUC values of 0.96 and 0.92, respectively. Important features such as baseline estimated glomerular filtration rate and albuminuria significantly contributed to the predictive model. Conclusion This study demonstrates the efficacy of an ML approach in developing a highly predictive model for estimating the timing of maintenance dialysis initiation in patients with CKD stages 3–5. These findings have important implications for personalised treatment strategies, enabling improved clinical decision-making and potentially enhancing patient outcomes. Data may be obtained from a third party and are not publicly available.","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140804000","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Development of a scoring system to quantify errors from semantic characteristics in incident reports","authors":"Haruhiro Uematsu, Masakazu Uemura, Masaru Kurihara, Hiroo Yamamoto, Tomomi Umemura, Fumimasa Kitano, Mariko Hiramatsu, Yoshimasa Nagao","doi":"10.1136/bmjhci-2023-100935","DOIUrl":"https://doi.org/10.1136/bmjhci-2023-100935","url":null,"abstract":"Objectives Incident reporting systems are widely used to identify risks and enable organisational learning. Free-text descriptions contain important information about factors associated with incidents. This study aimed to develop error scores by extracting information about the presence of error factors in incidents using an original decision-making model that partly relies on natural language processing techniques. Methods We retrospectively analysed free-text data from reports of incidents between January 2012 and December 2022 from Nagoya University Hospital, Japan. The sample data were randomly allocated to equal-sized training and validation datasets. We conducted morphological analysis on free text to segment terms from sentences in the training dataset. We calculated error scores for terms, individual reports and reports from staff groups according to report volume size and compared these with conventional classifications by patient safety experts. We also calculated accuracy, recall, precision and F-score values from the proposed ‘report error score’. Results Overall, 114 013 reports were included. We calculated 36 131 ‘term error scores’ from the 57 006 reports in the training dataset. There was a significant difference in error scores between reports of incidents categorised by experts as arising from errors (p<0.001, d =0.73 (large)) and other incidents. The accuracy, recall, precision and F-score values were 0.8, 0.82, 0.85 and 0.84, respectively. Group error scores were positively associated with expert ratings (correlation coefficient, 0.66; 95% CI 0.54 to 0.75, p<0.001) for all departments. Conclusion Our error scoring system could provide insights to improve patient safety using aggregated incident report data. Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author upon reasonable request.","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140626798","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Akiko Hanai, Tetsuo Ishikawa, Shoichiro Kawauchi, Yuta Iida, Eiryo Kawakami
{"title":"Generative artificial intelligence and non-pharmacological bias: an experimental study on cancer patient sexual health communications","authors":"Akiko Hanai, Tetsuo Ishikawa, Shoichiro Kawauchi, Yuta Iida, Eiryo Kawakami","doi":"10.1136/bmjhci-2023-100924","DOIUrl":"https://doi.org/10.1136/bmjhci-2023-100924","url":null,"abstract":"Objectives The objective of this study was to explore the feature of generative artificial intelligence (AI) in asking sexual health among cancer survivors, which are often challenging for patients to discuss. Methods We employed the Generative Pre-trained Transformer-3.5 (GPT) as the generative AI platform and used DocsBot for citation retrieval (June 2023). A structured prompt was devised to generate 100 questions from the AI, based on epidemiological survey data regarding sexual difficulties among cancer survivors. These questions were submitted to Bot1 (standard GPT) and Bot2 (sourced from two clinical guidelines). Results No censorship of sexual expressions or medical terms occurred. Despite the lack of reflection on guideline recommendations, ‘consultation’ was significantly more prevalent in both bots’ responses compared with pharmacological interventions, with ORs of 47.3 (p<0.001) in Bot1 and 97.2 (p<0.001) in Bot2. Discussion Generative AI can serve to provide health information on sensitive topics such as sexual health, despite the potential for policy-restricted content. Responses were biased towards non-pharmacological interventions, which is probably due to a GPT model designed with the ’s prohibition policy on replying to medical topics. This shift warrants attention as it could potentially trigger patients’ expectations for non-pharmacological interventions.","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140571883","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Association between daily step counts and healthy life years: a national cross-sectional study in Japan","authors":"Masahiro Nishi, Reo Nagamitsu, Satoaki Matoba","doi":"10.1136/bmjhci-2024-101051","DOIUrl":"https://doi.org/10.1136/bmjhci-2024-101051","url":null,"abstract":"Background Despite accumulating evidence concerning the association between daily step counts and mortality or disease risks, it is unclear whether daily step counts are associated with healthy life years. Methods We used the combined dataset of the Comprehensive Survey of Living Conditions and the National Health and Nutrition Survey conducted for a randomly sampled general population in Japan, 2019. Daily step counts were measured for 4957 adult participants. The associations of daily step counts with activity limitations in daily living and self-assessed health were evaluated using a multivariable logistic regression model. The bootstrap method was employed to mitigate uncertainties in estimating the threshold of daily step counts. Results The median age was 60 (44–71) years, and 2592 (52.3%) were female. The median daily step counts were 5650 (3332–8452). The adjusted OR of activity limitations in daily living for the adjacent daily step counts was 0.27 (95% CI 0.26 to 0.27) for all ages and 0.25 (95% CI 0.25 to 0.26) for older adults at the lowest, with the thresholds of significant association at 9000 step counts. The OR of self-assessed unhealthy status was 0.45 (95% CI 0.44 to 0.46) for all ages and 0.42 (95% CI 0.41 to 0.43) for older adults at the lowest, with the thresholds at 11 000 step counts. Conclusion Daily step counts were significantly associated with activity limitations in daily living and self-assessed health as determinants of healthy life years, up to 9000 and 11 000 step counts, respectively. These results suggest a target of daily step counts to prolong healthy life years within health initiatives. Data may be obtained from a third party and are not publicly available. We are prohibited from publicly opening the data. Data can be accessed through the Household Statistics Office of the Japanese Ministry of Health, Labour and Welfare (<https://www.mhlw.go.jp/toukei/itiran/eiyaku.html>).","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140833524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Prediction of high-risk emergency department revisits from a machine-learning algorithm: a proof-of-concept study","authors":"Chih-Wei Sung, Joshua Ho, Cheng-Yi Fan, Ching-Yu Chen, Chi-Hsin Chen, Shao-Yung Lin, Jia-How Chang, Jiun-Wei Chen, Edward Pei-Chuan Huang","doi":"10.1136/bmjhci-2023-100859","DOIUrl":"https://doi.org/10.1136/bmjhci-2023-100859","url":null,"abstract":"Background High-risk emergency department (ED) revisit is considered an important quality indicator that may reflect an increase in complications and medical burden. However, because of its multidimensional and highly complex nature, this factor has not been comprehensively investigated. This study aimed to predict high-risk ED revisit with a machine-learning (ML) approach. Methods This 3-year retrospective cohort study assessed adult patients between January 2019 and December 2021 from National Taiwan University Hospital Hsin-Chu Branch with high-risk ED revisit, defined as hospital or intensive care unit admission after ED return within 72 hours. A total of 150 features were preliminarily screened, and 79 were used in the prediction model. Deep learning, random forest, extreme gradient boosting (XGBoost) and stacked ensemble algorithm were used. The stacked ensemble model combined multiple ML models and performed model stacking as a meta-level algorithm. Confusion matrix, accuracy, sensitivity, specificity and area under the receiver operating characteristic curve (AUROC) were used to evaluate performance. Results Analysis was performed for 6282 eligible adult patients: 5025 (80.0%) in the training set and 1257 (20.0%) in the testing set. High-risk ED revisit occurred for 971 (19.3%) of training set patients vs 252 (20.1%) in the testing set. Leading predictors of high-risk ED revisit were age, systolic blood pressure and heart rate. The stacked ensemble model showed more favourable prediction performance (AUROC 0.82) than the other models: deep learning (0.69), random forest (0.78) and XGBoost (0.79). Also, the stacked ensemble model achieved favourable accuracy and specificity. Conclusion The stacked ensemble algorithm exhibited better prediction performance in which the predictions were generated from different ML algorithms to optimally maximise the final set of results. Patients with older age and abnormal systolic blood pressure and heart rate at the index ED visit were vulnerable to high-risk ED revisit. Further studies should be conducted to externally validate the model. Data are available on reasonable request.","PeriodicalId":9050,"journal":{"name":"BMJ Health & Care Informatics","volume":null,"pages":null},"PeriodicalIF":4.1,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140636847","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}