Applied Clinical Informatics最新文献

{"title":"Communication Challenges Experienced by Clinicians and Patients During Teleconsultation. Scoping Review.","authors":"Takashi Sota, Tim Jackson, Eleanor Yang, Annie Ys Lau","doi":"10.1055/a-2425-8626","DOIUrl":"https://doi.org/10.1055/a-2425-8626","url":null,"abstract":"<p><p>Background As teleconsultations continue to rise in popularity due to their convenience and accessibility, it's crucial to identify and address the challenges they present in order to improve the patient experience, enhance outcomes, and ensure the quality of care. To identify communication challenges that clinicians and patients experience during teleconsultation, a scoping review was conducted. Methods Studies were obtained from four databases (Ovid [MEDLINE], Ovid [Embase], CINAHL and Scopus). Gray literatures were not included. Studies focused on communication challenges between clinicians and their patients during teleconsultation in the context of COVID-19 and published from January 2000 to December 2022 were collected. Screening process was conducted by 2 independent reviewers. Data extraction was performed using a standardized form to capture study characteristics and communication challenges. Extracted data were analyzed to identify the communication challenges during teleconsultation, adherent to PRISMA-ScR. Results A total of 893 studies were collected from 4 databases and 26 studies were selected based on inclusion/exclusion criteria. Of these 26 eligible studies, 12 (46%) were from the US, 3 studies (12%) were from Australia, 2 (8%) were from the UK and Canada. These studies included 12 (46%) qualitative studies, 6 (23%) quantitative studies, 6 (23%) review articles, and 2 (8%) case reports. Eight factors contributing to communication challenges between clinicians and patients during teleconsultations were identified: technical issues, difficulties in developing rapport, lack of non-verbal communication, lack of physical examination, language barrier, spatial issues, clinician preparation, and difficulties in assessing patients' health literacy. Conclusion Eight factors were identified contributing to communication challenges during teleconsultation in the context of COVID-19. These findings highlight the need for addressing communication challenges to ensure effective teleconsultations. With the rise of teleconsultation in routine healthcare delivery, further research is warranted to confirm these findings, and to explore ways to overcome communication challenges during teleconsultation.</p>","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2024-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142330581","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and validation of the Nursing Information Security Questionnaire.","authors":"Xiaoyun Zhou, Tingting Gao, Xiujuan Jing, Hong Liu, Xuebing Jing","doi":"10.1055/a-2424-2103","DOIUrl":"https://doi.org/10.1055/a-2424-2103","url":null,"abstract":"<p><p>Background Ensuring the security of nursing information holds substantial importance. The awareness of information security among nurses in China is generally inadequate, and there is a lack of standardized evaluation tools for nurse information security in nursing practice. The nursing sector necessitates the establishment of a robust culture surrounding information security. Objective The aim of this study was to construct a self-reporting instrument for evaluating nursing information security. Methods The research team utilized literature analysis and group discussions to draft the item pool. After two rounds of Delphi consultation by 15 experts and pilot testing, the initial questionnaire was formed. Item analysis was carried out on the questionnaire, and the validity and reliability of the instrument were statistically tested by computing the Keiser-Meier-Olkin (KMO) and Bartlett tests, an exploratory factor analysis, a confirmatory factor analysis, convergent and discriminative validity, descriptive statistics, Cronbach's α and test-retest reliability. Results A total of 501 nurses participated in the study, supplemented by the inclusion of five experts who were invited to contribute to the assessment of content validity. Four factors were formed using exploratory factor analysis (n=250), and the cumulative variance contribution rate was found to be 60.10%. The confirmatory factor analysis (n=251) showed the model fit was good. The overall Cronbach's α coefficient of the questionnaire was 0.948, and the test-retest reliability was 0.837. Conclusion Finally, the NIS-Q with 38 items and three dimensions of knowledge, attitude and practice were formed. A promising assessment instrument for gauging the degree of nursing information security was introduced. Further, a foundational platform was established for implementing specific enhancement strategies aimed at advancing nursing information security.</p>","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2024-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142330582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Epidemiology of Patient Record Duplication.","authors":"Onur Sahin, Audrey Zhao, Reuben Applegate, Todd Johnson, Elmer V Bernstam","doi":"10.1055/a-2423-8499","DOIUrl":"https://doi.org/10.1055/a-2423-8499","url":null,"abstract":"<p><strong>Objective: </strong>Duplicate patient records can increase cost and medical errors. We assessed the association between demographic factors, comorbidities, healthcare usage and duplicate electronic health records.</p><p><strong>Materials and methods: </strong>We analyzed the association between duplicate patient records and multiple demographic variables (race, Hispanic ethnicity, sex and age) as well as Charlson comorbidity index (CCI), number of diagnoses, and number of healthcare encounters. The study population included 3,018,413 patients seen at a large urban academic medical center with at least one recorded diagnosis. Duplication of patient medical records was determined by using a previously validated enterprise Master Person Index.</p><p><strong>Results: </strong>Unknown or missing demographic data, Black race when compared to White Race (OR 1.35, p < 0.001), Hispanic compared to non-Hispanic ethnicity (OR 1.48, p < 0.001), older age (OR 1.01, p < 0.001), and \"Other\" sex compared to female sex (OR 4.71, p < 0.001) were associated with higher odds of having a duplicate record. Comorbidities (CCI, OR 1.10, p < 0.001) and more encounters with the health care system (OR 1.01, p < 0.001) were also associated with higher odds of having a duplicate record. In contrast, male sex compared to female sex was associated with lower odds of having a duplicate record (OR 0.88, p < 0.001).</p><p><strong>Discussion: </strong>The odds of duplications in medical records were higher in Black, Hispanic, older, non-male patients with more healthcare encounters, more comorbidities, and unknown demographic data. Understanding the epidemiology of duplicate records can help guide prevention and mitigation efforts for high-risk populations.</p><p><strong>Conclusion: </strong>Duplicate records can contribute to disparities in health care outcomes in minority populations.</p>","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2024-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142330583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Comprehensive Multi-Functional Approach for Measuring Parkinson's Disease Severity.","authors":"Morteza Rahimi, Zeina Al Masry, John Michael Templeton, Sandra Schneider, Christian Poellabauer","doi":"10.1055/a-2420-0413","DOIUrl":"https://doi.org/10.1055/a-2420-0413","url":null,"abstract":"<p><strong>Objectives: </strong>This research study aims to advance the staging of Parkinson's disease (PD) by incorporating machine learning to assess and include a broader multi-functional spectrum of neurocognitive symptoms in the staging schemes beyond motor-centric assessments. Specifically, we provide a novel framework to modernize and personalize PD staging more objectively by proposing a hybrid feature scoring approach.</p><p><strong>Methods: </strong>We recruited thirty-seven individuals diagnosed with PD, each of whom completed a series of tablet-based neurocognitive tests assessing motor, memory, speech, executive functions, and tasks ranging in complexity from single to multi-functional. Then, the collected data was used to develop a hybrid feature scoring system to calculate a weighted vector for each function. We evaluated current PD staging schemes and developed a new approach based on the features selected and extracted using Random Forest and Principal Component Analysis.</p><p><strong>Results: </strong>Our findings indicate a substantial bias in current PD staging systems toward fine-motor skills, i.e., other neurological functions (memory, speech, executive function, etc.) do not map into current PD stages as well as fine-motor skills do. The results demonstrate that a more accurate and personalized assessment of PD severity could be achieved by including a more exhaustive range of neurocognitive functions in the staging systems either by involving multiple functions in a unified staging score or by designing a function-specific staging system.</p><p><strong>Conclusions: </strong>The proposed hybrid feature score approach provides a comprehensive understanding of PD by highlighting the need for a staging system that covers various neurocognitive functions. This approach could potentially lead to more effective, objective, and personalized treatment strategies. Further, this proposed methodology could be adapted to other neurodegenerative conditions such as Alzheimer's disease or ALS.</p>","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2024-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142308859","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Patient–Clinician Diagnostic Concordance upon Hospital Admission","authors":"Alyssa Lam, Savanna Plombon, Alison Garber, Pamela Garabedian, Ronen Rozenblum, Jacqueline A. Griffin, Jeffrey L. Schnipper, Stuart R. Lipsitz, David W. Bates, Anuj K. Dalal","doi":"10.1055/s-0044-1788330","DOIUrl":"https://doi.org/10.1055/s-0044-1788330","url":null,"abstract":"<p>\u0000<b>Objectives</b> This study aimed to pilot an application-based patient diagnostic questionnaire (PDQ) and assess the concordance of the admission diagnosis reported by the patient and entered by the clinician.</p> <p>\u0000<b>Methods</b> Eligible patients completed the PDQ assessing patients' understanding of and confidence in the diagnosis 24 hours into hospitalization either independently or with assistance. Demographic data, the hospital principal problem upon admission, and International Classification of Diseases 10th Revision (ICD-10) codes were retrieved from the electronic health record (EHR). Two physicians independently rated concordance between patient-reported diagnosis and clinician-entered principal problem as full, partial, or no. Discrepancies were resolved by consensus. Descriptive statistics were used to report demographics for concordant (full) and nonconcordant (partial or no) outcome groups. Multivariable logistic regressions of PDQ questions and a priori selected EHR data as independent variables were conducted to predict nonconcordance.</p> <p>\u0000<b>Results</b> A total of 157 (77.7%) questionnaires were completed by 202 participants; 77 (49.0%), 46 (29.3%), and 34 (21.7%) were rated fully concordant, partially concordant, and not concordant, respectively. Cohen's kappa for agreement on preconsensus ratings by independent reviewers was 0.81 (0.74, 0.88). In multivariable analyses, patient-reported lack of confidence and undifferentiated symptoms (ICD-10 “R-code”) for the principal problem were significantly associated with nonconcordance (partial or no concordance ratings) after adjusting for other PDQ questions (3.43 [1.30, 10.39], <i>p</i> = 0.02) and in a model using selected variables (4.02 [1.80, 9.55], <i>p</i> < 0.01), respectively.</p> <p>\u0000<b>Conclusion</b> About one-half of patient-reported diagnoses were concordant with the clinician-entered diagnosis on admission. An ICD-10 “R-code” entered as the principal problem and patient-reported lack of confidence may predict patient–clinician nonconcordance early during hospitalization via this approach.</p> ","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":"50 1","pages":""},"PeriodicalIF":2.9,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142265943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Imaging Informatics Education in Clinical Informatics Programs: Perspective from Imaging and Clinical Informatics Professionals","authors":"Nathan A. Bumbarger, Alexander J. Towbin, Pamela Garcia-Filion, James Whitfill, Tessa Cook, Les R. Folio","doi":"10.1055/s-0044-1788327","DOIUrl":"https://doi.org/10.1055/s-0044-1788327","url":null,"abstract":"<p>\u0000<b>Background</b> Imaging and Clinical Informatics are domains of biomedical informatics. Imaging Informatics topics are often not covered in depth in most Clinical Informatics fellowships. While dedicated Imaging Informatics fellowships exist, they may not have the same rigor as ACGME (Accreditation Council for Graduate Medical Education) accredited Clinical Informatics fellowships and they do not provide a direct path toward subspecialty board certification.</p> <p>\u0000<b>Objectives</b> We compared published curricula and test content between Clinical and Imaging Informatics fellowship programs. We then highlighted differences between training programs and identified overlapping topics and opportunities for additional education for each type of trainee.</p> <p>\u0000<b>Methods</b> Published consensus curricula and topics were extracted for each specialty. Two informaticists compared topics as shared or not shared between specialties. Next, test content outlines were compared for each specialty exam, extracted, and classified as shared or not shared content. A Venn diagram was created to highlight areas unique to each specialty as well as areas of overlap.</p> <p>\u0000<b>Results</b> There were 139 Clinical Informatics topics compared with 97 Imaging Informatics topics. Of the 139 Clinical Informatics topics, 115 (83%) were covered in the Imaging Informatics curriculum. Of the 97 Imaging Informatics topics, 74 (76%) were covered in the Clinical Informatics curriculum. When using test content outline data, 170 out of 397 (43%) Imaging Informatics topics matched to 64 out of 139 (46%) Clinical Informatics topics. We describe examples of overlapping topics and those unique to each program to identify potential areas to expand.</p> <p>\u0000<b>Conclusion</b> Imaging Informatics and Clinical Informatics fellowship programs have some overlap with areas unique to each. Our review may help guide those seeking informatics education and potential certification. As enterprise imaging evolves, these differences may become more important and create knowledge gaps, if not systematically evaluated.</p> ","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":"31 1","pages":""},"PeriodicalIF":2.9,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142265804","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Special Issue on Informatics Education: Exploring the Impact of GitHub Copilot on Health Informatics Education.","authors":"Sanja Avramovic,Ivan Avramovic,Janusz Wojtusiak","doi":"10.1055/a-2414-7790","DOIUrl":"https://doi.org/10.1055/a-2414-7790","url":null,"abstract":"BACKGROUNDThe use of artificial intelligence-driven code completion tools, particularly the integration of GitHub Copilot with Visual Studio, has potential implications for Health Informatics education, particularly for students learning SQL and Python.OBJECTIVESThis study aims to evaluate the effectiveness of these tools in solving or assisting with the solution of problems found in Health Informatics coursework, ranging from simple to complex.METHODSThe study assesses the performance of GitHub Copilot in generating code for Health Informatics coding assignments from graduate classes, with a focus on the impact of detailed explanations on the tool's effectiveness.RESULTSFindings reveal that GitHub Copilot can generate correct code for straightforward problems. The correctness and effectiveness of solutions decrease with problem complexity, and the tool struggles with the most challenging problems, although performance on complex problems improves with more detailed explanations.CONCLUSIONSThe study underscores the relevance of these tools to programming in Health Informatics education but also highlights the need for critical evaluation by students. It concludes with a call for educators to adapt swiftly to this rapidly evolving technology.","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":"32 1","pages":""},"PeriodicalIF":2.9,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142265942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Developing PRISM: A Pragmatic Institutional Survey and Bench Marking Tool to Measure Digital Research Maturity of Cancer Centers","authors":"Carlos Berenguer Albiñana, Matteo Pallocca, Hayley Fenton, Will Sopwith, Charlie Van Eden, Olof Akre, Annika Auranen, François Bocquet, Marina Borges, Emiliano Calvo, John Corkett, Serena Di Cosimo, Nicola Gentili, Julien Guérin, Sissel Jor, Tomas Kazda, Alenka Kolar, Tim Kuschel, Maria Julia Lostes, Chiara Paratore, Paolo Pedrazzoli, Marko Petrovic, Jarno Raid, Miriam Roche, Christoph Schatz, Joelle Thonnard, Giovanni Tonon, Alberto Traverso, Andrea Wolf, Ahmed H. Zedan, Piers Mahon","doi":"10.1055/s-0044-1788331","DOIUrl":"https://doi.org/10.1055/s-0044-1788331","url":null,"abstract":"<p>\u0000<b>Background</b> Multicenter precision oncology real-world evidence requires a substantial long-term investment by hospitals to prepare their data and align on common Clinical Research processes and medical definitions. Our team has developed a self-assessment framework to support hospitals and hospital networks to measure their digital maturity and better plan and coordinate those investments. From that framework, we developed PRISM for Cancer Outcomes: <b>PR</b>agmatic <b>I</b>nstitutional <b>S</b>urvey and bench<b>M</b>arking.</p> <p>\u0000<b>Objectives</b> The primary objective was to develop PRISM as a tool for self-assessment of digital maturity in oncology hospitals and research networks; a secondary objective was to create an initial benchmarking cohort of >25 hospitals using the tool as input for future development.</p> <p>\u0000<b>Methods</b> PRISM is a 25-question semiquantitative self-assessment survey developed iteratively from expert knowledge in oncology real-world study delivery. It covers four digital maturity dimensions: (1) Precision oncology, (2) Clinical digital data, (3) Routine outcomes, and (4) Information governance and delivery. These reflect the four main data types and critical enablers for precision oncology research from routine electronic health records.</p> <p>\u0000<b>Results</b> During piloting with 26 hospitals from 19 European countries, PRISM was found to be easy to use and its semiquantitative questions to be understood in a wide diversity of hospitals. Results within the initial benchmarking cohort aligned well with internal perspectives. We found statistically significant differences in digital maturity, with Precision oncology being the most mature dimension, and Information governance and delivery the least mature.</p> <p>\u0000<b>Conclusion</b> PRISM is a light footprint benchmarking tool to support the planning of large-scale real-world research networks. It can be used to (i) help an individual hospital identify areas most in need of investment and improvement, (ii) help a network of hospitals identify sources of best practice and expertise, and (iii) help research networks plan research. With further testing, policymakers could use PRISM to better plan digital investments around the Cancer Mission and European Digital Health Space.</p> ","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":"1 1","pages":""},"PeriodicalIF":2.9,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142199666","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Issue on Teaching and Training Future Health Informaticians:Partnering with Students to Develop a Capstone for a Graduate Health Informatics Program.","authors":"Rita Jezrawi,Stephanie Zahorka Derka,Elizabeth Warnick,Jasmine Foley,Vritti Patel,Neethu Pavithran,Thérèse Bernier,Nicole Wagner,Neil G Barr,Vincent Maccio,Margaret Leyland,Cynthia Lokker","doi":"10.1055/a-2412-3535","DOIUrl":"https://doi.org/10.1055/a-2412-3535","url":null,"abstract":"OBJECTIVETo assess the desirability, feasibility, and sustainability of integrating a project-based capstone course with the course-based curriculum of an interdisciplinary MSc health informatics program guided with a student-partnered steering committee and student-centered approach.METHODSWe conducted an online cross-sectional survey (n=87) and three semi-structured focus groups (n=18) of health informatics students and alumni. Survey data was analyzed descriptively. Focus groups were audio-recorded and transcribed verbatim and then analyzed using a general inductive and classic analysis approach.RESULTSMost students were supportive of including a capstone project but desired an option to work independently or within a group. Students perceived several benefits to capstone courses while concerned over perceived challenges to capstone implementation, evaluation, and managing group processes. Themes identified were: 1) professional development, identity, and career advancement; 2) emulating the real world and learning beyond the classroom, 3) embracing new, full circle learning, 4) anticipated course structure, delivery, and preparation, 5) balancing student choice, interests, and priorities, and 6) concerns over group dynamics, limitations, and support.CONCLUSIONSThis study demonstrates the value of having students as partners at each stage in the process from methods conception to course curriculum design. With the steering committee and the curriculum developer, we codeveloped a student-centered course that integrates foundational digital health-related project knowledge acquisition with an inquiry-based project which can be completed independently or in small groups. This study demonstrates the potential benefits and challenges that health informatics educators may consider when (re)-designing capstone courses.","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":"44 1","pages":""},"PeriodicalIF":2.9,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142199635","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing Suicide Risk Prediction Models with Temporal Clinical Note Features.","authors":"Kevin Krause,Sharon Davis,Zhijun Yin,Katherine Schafer,Trent Rosenbloom,Colin Walsh","doi":"10.1055/a-2411-5796","DOIUrl":"https://doi.org/10.1055/a-2411-5796","url":null,"abstract":"OBJECTIVEThe objective of this study was to investigate the impact of enhancing a structured-data-based suicide attempt risk prediction model with temporal Concept Unique Identifiers (CUIs) derived from clinical notes. We aimed to examine how different temporal schemes, model types, and prediction ranges influenced the model's predictive performance. This research sought to improve our understanding of how the integration of temporal information and clinical variable transformation could enhance model predictions.MATERIALS AND METHODSWe identified modeling targets using diagnostic codes for suicide attempts within 30, 90, or 365 days following a temporally grouped visit cluster. Structured data included medications, diagnoses, procedures, and demographics, while unstructured data consisted of terms extracted with regular expressions from clinical notes. We compared models trained only on structured data (controls) to hybrid models trained on both structured and unstructured data. We used two temporalization schemes for clinical notes: fixed 90-day windows and flexible epochs. We trained and assessed random forests and hybrid LSTM neural networks using AUPRC and AUROC, with additional evaluation of sensitivity and PPV at 95% specificity.RESULTSThe training set included 2,364,183 visit clusters with 2,009 30-day suicide attempts, and the testing set contained 471,936 visit clusters with 480 suicide attempts. Models trained with temporal CUIs outperformed those trained with only structured data. The window-temporalized LSTM model achieved the highest AUPRC (0.056 ± 0.013) for the 30-day prediction range. Hybrid models generally showed better performance compared to controls across most metrics.DISCUSSION AND CONCLUSIONThis study demonstrated that incorporating EHR-derived clinical note features enhanced suicide attempt risk prediction models, particularly with window-temporalized LSTM models. Our results underscored the critical value of unstructured data in suicidality prediction, aligning with previous findings. Future research should focus on integrating more sophisticated methods to continue improving prediction accuracy, which will enhance the effectiveness of future intervention.","PeriodicalId":48956,"journal":{"name":"Applied Clinical Informatics","volume":"4 1","pages":""},"PeriodicalIF":2.9,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142199654","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}