Pub Date : 2025-11-15DOI: 10.1016/j.ijmedinf.2025.106187
Mingbo Chen , Fan Wang
{"title":"Artificial intelligence in colon cancer: A commentary on advances and challenges","authors":"Mingbo Chen , Fan Wang","doi":"10.1016/j.ijmedinf.2025.106187","DOIUrl":"10.1016/j.ijmedinf.2025.106187","url":null,"abstract":"","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"207 ","pages":"Article 106187"},"PeriodicalIF":4.1,"publicationDate":"2025-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145532483","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}
Pub Date : 2025-11-13DOI: 10.1016/j.ijmedinf.2025.106178
Julie Li , Judith Thomas , Melissa Baysari , Andrew Georgiou , Mirela Prgomet
Objective
Diagnostic stewardship refers to coordinated efforts to ensure timely review and appropriate follow-up of test results. This scoping review aimed to identify the mechanisms, or triggers that create conditions that facilitate the diagnostic stewardship process in the electronic management of test results by clinicians.
Methods
Searches for original studies published between January 2013 and 30 April 2024 reporting electronic diagnostic test result management interventions, and the outcomes of these, were conducted in MEDLINE, EMBASE, CINAHL, ProQuest and Scopus. ProQuest and Scopus were searched for grey literature. Identified stewardship mechanisms were categorised using an eight-dimensional sociotechnical model.
Results
Electronic stewardship efforts relied heavily on mechanisms delivered through technical components in order to increase provider awareness of returned results, advise appropriate follow-up actions, and assist with the prioritisation of abnormal test results. Non-technical mechanisms included protected time to accommodate workload, or the extension of responsibility for follow-up tasks to other staff groups. Studies reporting performance feedback in the form of self-audit, education and clarification of expectations for test result follow-up also reported significant improvements in the actioning and follow-up of test results.
Discussion
Review findings highlight an increasing integration of non-technical mechanisms in stewardship efforts that facilitate the optimal use of digital health.
Conclusion
Success of stewardship interventions depended on integrating social and technical mechanisms. Findings underscore the importance of designing digital health interventions that are not only functional but also contextually aligned with the broader clinical environment.
{"title":"Diagnostic stewardship mechanisms in electronic test results management – a scoping review","authors":"Julie Li , Judith Thomas , Melissa Baysari , Andrew Georgiou , Mirela Prgomet","doi":"10.1016/j.ijmedinf.2025.106178","DOIUrl":"10.1016/j.ijmedinf.2025.106178","url":null,"abstract":"<div><h3>Objective</h3><div>Diagnostic stewardship refers to coordinated efforts to ensure timely review and appropriate follow-up of test results. This scoping review aimed to identify the mechanisms, or triggers that create conditions that facilitate the diagnostic stewardship process in the electronic management of test results by clinicians.</div></div><div><h3>Methods</h3><div>Searches for original studies published between January 2013 and 30 April 2024 reporting electronic diagnostic test result management interventions, and the outcomes of these, were conducted in MEDLINE, EMBASE, CINAHL, ProQuest and Scopus. ProQuest and Scopus were searched for grey literature. Identified stewardship mechanisms were categorised using an eight-dimensional sociotechnical model.</div></div><div><h3>Results</h3><div>Electronic stewardship efforts relied heavily on mechanisms delivered through technical components in order to increase provider awareness of returned results, advise appropriate follow-up actions, and assist with the prioritisation of abnormal test results. Non-technical mechanisms included protected time to accommodate workload, or the extension of responsibility for follow-up tasks to other staff groups. Studies reporting performance feedback in the form of self-audit, education and clarification of expectations for test result follow-up also reported significant improvements in the actioning and follow-up of test results.</div></div><div><h3>Discussion</h3><div>Review findings highlight an increasing integration of non-technical mechanisms in stewardship efforts that facilitate the optimal use of digital health.</div></div><div><h3>Conclusion</h3><div>Success of stewardship interventions depended on integrating social and technical mechanisms. Findings underscore the importance of designing digital health interventions that are not only functional but also contextually aligned with the broader clinical environment.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"207 ","pages":"Article 106178"},"PeriodicalIF":4.1,"publicationDate":"2025-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145558239","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}
Pub Date : 2025-11-12DOI: 10.1016/j.ijmedinf.2025.106172
Gang Wang , Ru Yang , Yujie Zhang, Xiya Wen, Chang Liu, Enzhuo Liu, Mei Tang, Leixi Xue, Zhichun Liu
<div><h3>Background</h3><div>Large language models (LLMs) demonstrate significant potential in medical information provision and may serve as valuable tools for patients seeking health information. Existing research primarily focuses on individual models or general medical inquiries, with no systematic evaluation of mainstream LLMs’ performance. Particularly noteworthy is the absence of cross-comparison studies involving Chinese AI model DeepSeek-R1. This research gap may hinder the effective translation of artificial intelligence technology into clinical practice for rheumatic diseases.</div></div><div><h3>Objective</h3><div>This study aims to assess the accuracy, completeness, readability, and level of detail in the responses provided by LLMs to common questions related to connective tissue disease (CTD).</div></div><div><h3>Methods</h3><div>This cross-sectional study analyzed the responses to 250 common questions related to CTD, covering topics such as etiology and pathogenesis, risk factors, clinical manifestations, diagnostic criteria and differential diagnosis, treatment, prevention, and prognosis. These questions were collaboratively developed by three experienced clinicians and piloted by two rheumatology residents. Between February 18 and February 20, 2025, the questions were input as prompts into DeepSeek-R1, ChatGPT-4.0 (OpenAI), Copilot (Microsoft), and Gemini-2.0 (Google). The accuracy, completeness, readability, level of detail, and inclusion of health advice disclaimers in the responses were evaluated. Two experienced clinicians conducted a double-blind evaluation using four standardized scoring tools, with the average score serving as the final result. In cases of conflict or significant discrepancies in scores for the same question, the final score for each answer was determined by majority consensus.</div></div><div><h3>Results</h3><div>A total of 1000 responses (4000 scores) were generated, with an average accuracy score of 5.12 (0.78), and an average completeness score of 1.98 (0.56). The answers provided by the LLMs were “easy” to read, with an average FRES score of 80.46 (7.19). The average level of detail score was 79.38 (8.14). Overall, DeepSeek-R1 and ChatGPT-4.0 performed the best, with similar scores in accuracy, completeness, readability, and level of detail. Health advice disclaimers were included in 83%-94% of the responses.</div></div><div><h3>Conclusion</h3><div>Using LLMs as tools for education and consultation in rheumatic diseases, particularly CTD shows promising potential, but the results are varied, indicating room for further improvement. DeepSeek-R1 and ChatGPT-4.0 scored similarly, performing the best in terms of accuracy, completeness, readability, and level of detail. The study results provide a basis for decision-making regarding the integration of the Chinese AI model DeepSeek-R1 into global patient education and support systems.</div></div><div><h3>Limitation & future direction</h3><div>This study
{"title":"Evaluating the performance of Large language models in rheumatology for connective tissue Diseases: DeepSeek-R1, ChatGPT-4.0, Copilot, and Gemini-2.0","authors":"Gang Wang , Ru Yang , Yujie Zhang, Xiya Wen, Chang Liu, Enzhuo Liu, Mei Tang, Leixi Xue, Zhichun Liu","doi":"10.1016/j.ijmedinf.2025.106172","DOIUrl":"10.1016/j.ijmedinf.2025.106172","url":null,"abstract":"<div><h3>Background</h3><div>Large language models (LLMs) demonstrate significant potential in medical information provision and may serve as valuable tools for patients seeking health information. Existing research primarily focuses on individual models or general medical inquiries, with no systematic evaluation of mainstream LLMs’ performance. Particularly noteworthy is the absence of cross-comparison studies involving Chinese AI model DeepSeek-R1. This research gap may hinder the effective translation of artificial intelligence technology into clinical practice for rheumatic diseases.</div></div><div><h3>Objective</h3><div>This study aims to assess the accuracy, completeness, readability, and level of detail in the responses provided by LLMs to common questions related to connective tissue disease (CTD).</div></div><div><h3>Methods</h3><div>This cross-sectional study analyzed the responses to 250 common questions related to CTD, covering topics such as etiology and pathogenesis, risk factors, clinical manifestations, diagnostic criteria and differential diagnosis, treatment, prevention, and prognosis. These questions were collaboratively developed by three experienced clinicians and piloted by two rheumatology residents. Between February 18 and February 20, 2025, the questions were input as prompts into DeepSeek-R1, ChatGPT-4.0 (OpenAI), Copilot (Microsoft), and Gemini-2.0 (Google). The accuracy, completeness, readability, level of detail, and inclusion of health advice disclaimers in the responses were evaluated. Two experienced clinicians conducted a double-blind evaluation using four standardized scoring tools, with the average score serving as the final result. In cases of conflict or significant discrepancies in scores for the same question, the final score for each answer was determined by majority consensus.</div></div><div><h3>Results</h3><div>A total of 1000 responses (4000 scores) were generated, with an average accuracy score of 5.12 (0.78), and an average completeness score of 1.98 (0.56). The answers provided by the LLMs were “easy” to read, with an average FRES score of 80.46 (7.19). The average level of detail score was 79.38 (8.14). Overall, DeepSeek-R1 and ChatGPT-4.0 performed the best, with similar scores in accuracy, completeness, readability, and level of detail. Health advice disclaimers were included in 83%-94% of the responses.</div></div><div><h3>Conclusion</h3><div>Using LLMs as tools for education and consultation in rheumatic diseases, particularly CTD shows promising potential, but the results are varied, indicating room for further improvement. DeepSeek-R1 and ChatGPT-4.0 scored similarly, performing the best in terms of accuracy, completeness, readability, and level of detail. The study results provide a basis for decision-making regarding the integration of the Chinese AI model DeepSeek-R1 into global patient education and support systems.</div></div><div><h3>Limitation & future direction</h3><div>This study","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"207 ","pages":"Article 106172"},"PeriodicalIF":4.1,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145532491","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}
Pub Date : 2025-11-11DOI: 10.1016/j.ijmedinf.2025.106184
Simon Lewerenz , Catherine Chronaki , Anderson Carmo , Henrique Martins
Purpose
The pursuit of eHealth interoperability across Europe has seen substantial public investment in EU-funds (over €200 million) and in experts’ time over the last two decades. In an era of preparation towards the European Health Data Space (EHDS) recovering this knowledge is essential and this article aims to identify such EU-funded projects and assess the long-term accessibility of their outputs. It derives consequences for impact and knowledge transfer and highlights implications for health policy worldwide and the developing EHDS.
Methods
We conducted a narrative review, informed by methodical searching and reported following PRISMA-ScR guidance, of the outputs of 31 EU-funded digital health interoperability projects spanning two decades (2005–2024). Project websites, access to deliverables and peer-reviewed publications were methodically assessed. The findings were critically analysed regarding current EU policy frameworks and discussed considering global relevance.
Findings
Despite the ambitious goals and funding allocated to these projects, a substantial portion of their outputs is no longer accessible, published as academic or even as grey literature, hindering the development of EU interoperable eHealth systems. 8 projects generated no peer-reviewed publications and nearly half lacks functional project websites post-project terminus. Existing EU repository CORDIS appears insufficient or inappropriately used as just half of projects have comprehensive outputs available there. These shortcomings limit cumulative learning, knowledge transfer and the assessment of the sustainability, value and impact of the public funds invested.
Implications
Our findings underscore the importance of strong knowledge governance for publicly funded international digital health initiatives. We recommend funders to implement stronger instruments for mandating open-access publications that are peer-reviewed and the systematic archiving of all project deliverables with citable identifiers. Implementing these measures within publicly funded programmes can limit loss of knowledge and foster sustainable open innovation, promoting more effective and efficient public investment backing digital health systems worldwide.
{"title":"Sustainable value generation from digital health investments: lessons from EU-funded projects preceding the European health data space","authors":"Simon Lewerenz , Catherine Chronaki , Anderson Carmo , Henrique Martins","doi":"10.1016/j.ijmedinf.2025.106184","DOIUrl":"10.1016/j.ijmedinf.2025.106184","url":null,"abstract":"<div><h3>Purpose</h3><div>The pursuit of eHealth interoperability across Europe has seen substantial public investment in EU-funds (over €200 million) and in experts’ time over the last two decades. In an era of preparation towards the European Health Data Space (EHDS) recovering this knowledge is essential and this article aims to identify such EU-funded projects and assess the long-term accessibility of their outputs. It derives consequences for impact and knowledge transfer and highlights implications for health policy worldwide and the developing EHDS.</div></div><div><h3>Methods</h3><div>We conducted a narrative review, informed by methodical searching and reported following PRISMA-ScR guidance, of the outputs of 31 EU-funded digital health interoperability projects spanning two decades (2005–2024). Project websites, access to deliverables and peer-reviewed publications were methodically assessed. The findings were critically analysed regarding current EU policy frameworks and discussed considering global relevance.</div></div><div><h3>Findings</h3><div>Despite the ambitious goals and funding allocated to these projects, a substantial portion of their outputs is no longer accessible, published as academic or even as grey literature, hindering the development of EU interoperable eHealth systems. 8 projects generated no peer-reviewed publications and nearly half lacks functional project websites post-project terminus. Existing EU repository <em>CORDIS</em> appears insufficient or inappropriately used as just half of projects have comprehensive outputs available there. These shortcomings limit cumulative learning, knowledge transfer and the assessment of the sustainability, value and impact of the public funds invested.</div></div><div><h3>Implications</h3><div>Our findings underscore the importance of strong knowledge governance for publicly funded international digital health initiatives. We recommend funders to implement stronger instruments for mandating open-access publications that are peer-reviewed and the systematic archiving of all project deliverables with citable identifiers. Implementing these measures within publicly funded programmes can limit loss of knowledge and foster sustainable open innovation, promoting more effective and efficient public investment backing digital health systems worldwide.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"206 ","pages":"Article 106184"},"PeriodicalIF":4.1,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145528379","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}
Pub Date : 2025-11-10DOI: 10.1016/j.ijmedinf.2025.106183
Daniela Fantin Carro , Leda Tomiko Yamada da Silveira , Edmund Chada Baracat , Jorge Milhem Haddad , Adriana C. Lunardi , Elizabeth Alves Ferreira
Background
Paper-based bladder diaries are widely used to assess urinary symptoms, including urinary incontinence (UI). Patient adherence to completing them is a challenge. Mobile applications (apps) have become accessible and low-cost alternative tools to improve clinical data collection; however, their accuracy still lacks evidence.
Objective
To validate a mobile app to assess voiding habits.
Methods
Clinimetric property study of the bladder diary app “Minha Bexiga.” After the app’s usability has been evaluated by the System Usability Score, women with and without UI, aged over 18 and who own smartphones, used the “Minha Bexiga” app and a paper bladder diary for three consecutive days to test criterion validity via correlation between scores obtained from the paper and electronic diaries. After 7 days, the women used only the “Minha Bexiga” app again to test reliability via intraclass correlation coefficient (ICC) test.
Results
A total of 70 women were included (38.5 years (28–48.5), 30 % had up to 11 years of education, 39 with UI and 31 without UI). The “Minha Bexiga” app showed substantial reliability (ICC = 0.85 (95 % CI 0.77–0.90)) and strong validity (Rho = 0.67; p < 0.001). In the sub-analysis, grouping all women by education level and age, the validity of the app has been observed for women with more than 11 years of education (Rho = 0.90; p < 0.001), with no relation to age. Reliability was excellent for women aging over 40 years, with no relation to education level (ICC = 0.90 (95 %CI 0.80–0.94)). The usability of the app was considered acceptable by women.
Conclusions
The “Minha Bexiga” app may be a useful tool for assessing voiding habits. However, its use seems to be more accurate for women over 40 years of age and with 11 years of education.
基于纸张的膀胱日记被广泛用于评估泌尿系统症状，包括尿失禁（UI）。耐心地坚持完成它们是一个挑战。移动应用程序（app）已成为可访问和低成本的替代工具，以改善临床数据收集；然而，它们的准确性仍然缺乏证据。目的验证一款评估排尿习惯的手机应用程序。方法对膀胱日记应用程序“敏哈贝西加”进行计量学性质研究。在应用程序的可用性通过系统可用性评分进行评估后，有UI和没有UI的女性，年龄在18岁以上，拥有智能手机，连续三天使用“Minha Bexiga”应用程序和纸质膀胱日记，通过从纸质日记和电子日记中获得的分数之间的相关性来测试标准的有效性。7天后，女性再次仅使用“Minha Bexiga”应用程序，通过类内相关系数（ICC）检验信度。结果共纳入70例女性，年龄38.5岁（28 ~ 48.5岁），其中受教育程度达11年者占30%，有尿失禁者39例，无尿失禁者31例。“Minha Bexiga”应用程序具有较高的信度（ICC = 0.85 (95% CI 0.77-0.90)）和强效度（Rho = 0.67; p < 0.001）。在子分析中，根据受教育程度和年龄对所有女性进行分组，观察到该应用程序对受教育程度超过11年的女性的有效性（Rho = 0.90; p < 0.001），与年龄无关。40岁以上女性的信度极好，与教育水平无关（ICC = 0.90 (95% CI 0.80-0.94)）。这款应用的可用性被女性认为是可以接受的。结论“明哈贝西加”应用程序可作为评估患者排尿习惯的有效工具。然而，对于40岁以上、受过11年教育的女性来说，它的使用似乎更准确。
{"title":"Validity and reliability study of a free mobile app for assessing voiding habits","authors":"Daniela Fantin Carro , Leda Tomiko Yamada da Silveira , Edmund Chada Baracat , Jorge Milhem Haddad , Adriana C. Lunardi , Elizabeth Alves Ferreira","doi":"10.1016/j.ijmedinf.2025.106183","DOIUrl":"10.1016/j.ijmedinf.2025.106183","url":null,"abstract":"<div><h3>Background</h3><div>Paper-based bladder diaries are widely used to assess urinary symptoms, including urinary incontinence (UI). Patient adherence to completing them is a challenge. Mobile applications (apps) have become accessible and low-cost alternative tools to improve clinical data collection; however, their accuracy still lacks evidence.</div></div><div><h3>Objective</h3><div>To validate a mobile app to assess voiding habits.</div></div><div><h3>Methods</h3><div>Clinimetric property study of the bladder diary app “Minha Bexiga.” After the app’s usability has been evaluated by the System Usability Score, women with and without UI, aged over 18 and who own smartphones, used the “Minha Bexiga” app and a paper bladder diary for three consecutive days to test criterion validity via correlation between scores obtained from the paper and electronic diaries. After 7 days, the women used only the “Minha Bexiga” app again to test reliability via intraclass correlation coefficient (ICC) test.</div></div><div><h3>Results</h3><div>A total of 70 women were included (38.5 years (28–48.5), 30 % had up to 11 years of education, 39 with UI and 31 without UI). The “Minha Bexiga” app showed substantial reliability (ICC = 0.85 (95 % CI 0.77–0.90)) and strong validity (Rho = 0.67; p < 0.001). In the sub-analysis, grouping all women by education level and age, the validity of the app has been observed for women with more than 11 years of education (Rho = 0.90; p < 0.001), with no relation to age. Reliability was excellent for women aging over 40 years, with no relation to education level (ICC = 0.90 (95 %CI 0.80–0.94)). The usability of the app was considered acceptable by women.</div></div><div><h3>Conclusions</h3><div>The “Minha Bexiga” app may be a useful tool for assessing voiding habits. However, its use seems to be more accurate for women over 40 years of age and with 11 years of education.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"207 ","pages":"Article 106183"},"PeriodicalIF":4.1,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145532490","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}
Pub Date : 2025-11-10DOI: 10.1016/j.ijmedinf.2025.106180
Jordi de la Torre
Objective: To develop and evaluate a scalable methodology for harmonizing inconsistent units in large-scale clinical datasets, addressing a key barrier to data interoperability.
Materials and Methods: We designed a novel unit harmonization system combining BM25, sentence embeddings, Bayesian optimization, and a bidirectional transformer based binary classifier for retrieving and matching laboratory test entries. The system was evaluated using the Optum Clinformatics Datamart dataset (7.5 billion entries). We implemented a multi-stage pipeline: filtering, identification, harmonization proposal generation, automated re-ranking, and manual validation. Performance was assessed using Mean Reciprocal Rank (MRR) and other standard information retrieval metrics.
Results: Our hybrid retrieval approach combining BM25 and sentence embeddings (MRR: 0.8833) significantly outperformed both lexical-only (MRR: 0.7985) and embedding-only (MRR: 0.5277) approaches. The transformer-based reranker further improved performance (absolute MRR improvement: 0.10), bringing the final system MRR to 0.9833. The system achieved 83.39 % precision at rank 1 and 94.66 % recall at rank 5.
Discussion: The hybrid architecture effectively leverages the complementary strengths of lexical and semantic approaches. The reranker addresses cases where initial retrieval components make errors due to complex semantic relationships in medical terminology.
Conclusion: Our framework provides an efficient, scalable solution for unit harmonization in clinical datasets, reducing manual effort while improving accuracy. Once harmonized, data can be reused seamlessly in different analyses, ensuring consistency across healthcare systems and enabling more reliable multi-institutional studies and meta-analyses.
{"title":"Scalable unit harmonization in medical informatics via Bayesian-optimized retrieval and transformer-based re-ranking","authors":"Jordi de la Torre","doi":"10.1016/j.ijmedinf.2025.106180","DOIUrl":"10.1016/j.ijmedinf.2025.106180","url":null,"abstract":"<div><div><em>Objective:</em> To develop and evaluate a scalable methodology for harmonizing inconsistent units in large-scale clinical datasets, addressing a key barrier to data interoperability.</div><div><em>Materials and Methods:</em> We designed a novel unit harmonization system combining BM25, sentence embeddings, Bayesian optimization, and a bidirectional transformer based binary classifier for retrieving and matching laboratory test entries. The system was evaluated using the Optum Clinformatics Datamart dataset (7.5 billion entries). We implemented a multi-stage pipeline: filtering, identification, harmonization proposal generation, automated re-ranking, and manual validation. Performance was assessed using Mean Reciprocal Rank (MRR) and other standard information retrieval metrics.</div><div><em>Results:</em> Our hybrid retrieval approach combining BM25 and sentence embeddings (MRR: 0.8833) significantly outperformed both lexical-only (MRR: 0.7985) and embedding-only (MRR: 0.5277) approaches. The transformer-based reranker further improved performance (absolute MRR improvement: 0.10), bringing the final system MRR to 0.9833. The system achieved 83.39 % precision at rank 1 and 94.66 % recall at rank 5.</div><div><em>Discussion:</em> The hybrid architecture effectively leverages the complementary strengths of lexical and semantic approaches. The reranker addresses cases where initial retrieval components make errors due to complex semantic relationships in medical terminology.</div><div><em>Conclusion:</em> Our framework provides an efficient, scalable solution for unit harmonization in clinical datasets, reducing manual effort while improving accuracy. Once harmonized, data can be reused seamlessly in different analyses, ensuring consistency across healthcare systems and enabling more reliable multi-institutional studies and meta-analyses.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"206 ","pages":"Article 106180"},"PeriodicalIF":4.1,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145528378","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}
Pub Date : 2025-11-08DOI: 10.1016/j.ijmedinf.2025.106179
Jasmine Pani , Laura Fiorini , Erika Rovini , Francesco Giuliani , Letizia Lorusso , Sergio Russo , Adriano César do Nascimento Teixeira Fernandes , Ana Goreti de Oliveira , Elisabete Pitarma , Ângela Rodrigues , María-Victoria Bueno-Delgado , Mateja Erce Paoli , Filippo Cavallo
Introduction
The global increase in the older adult population has amplified interest in supporting aging in place. Age-related physical and cognitive limitations pose significant challenges, for example managing daily tasks such as medication or household activities can become increasingly difficult. This also burdens informal caregivers with emotional stress, time demands and care coordination. Assistive technologies can enhance autonomy, social connection and health management, but their long-term adoption and use remain limited.
Methods
Within the Pharaon project, the study explored factors influencing technology use among older adults after 12 months of use. The research focused on the Italian pilot while integrating insights from other European sites. Qualitative data from semi-structured interviews were analyzed thematically and validated through reflection meetings and iterative online questionnaires. Recurring factors were consolidated, prioritized and mapped to explore interconnections affecting adoption.
Results
Initial findings revealed 30 factors, including personal characteristics, motivational aspects, technical design, social dynamics, and environmental context. These were rated by participating teams and 25 were retained for deeper analysis. An interconnection analysis explored how these factors influenced one another. Highly interconnected factors, such as social context, personalized training, ease of use, and cognitive changes, were central to understanding and improving technology adoption. In contrast, aspects like education level, internet access, and technical requirements appeared less interconnected, indicating more isolated effects.
Discussion
This multi-site, empirical investigation emphasizes the complex and interconnected nature of technology adoption among older adults. Our findings underscore the need for user-centered design, tailored training, and sensitivity to social context. Furthermore, the interactive mapping of interrelated factors provides a practical framework for developers and policymakers to target impactful interventions. To our knowledge, this is the first cross-pilot study to empirically define and score factors affecting older adults’ technology use based on multi-pilot data across Europe, offering valuable insights for long-term, meaningful engagement.
{"title":"Identifying and evaluating factors influencing technology adoption: A multi-pilot study within the Pharaon project","authors":"Jasmine Pani , Laura Fiorini , Erika Rovini , Francesco Giuliani , Letizia Lorusso , Sergio Russo , Adriano César do Nascimento Teixeira Fernandes , Ana Goreti de Oliveira , Elisabete Pitarma , Ângela Rodrigues , María-Victoria Bueno-Delgado , Mateja Erce Paoli , Filippo Cavallo","doi":"10.1016/j.ijmedinf.2025.106179","DOIUrl":"10.1016/j.ijmedinf.2025.106179","url":null,"abstract":"<div><h3>Introduction</h3><div>The global increase in the older adult population has amplified interest in supporting aging in place. Age-related physical and cognitive limitations pose significant challenges, for example managing daily tasks such as medication or household activities can become increasingly difficult. This also burdens informal caregivers with emotional stress, time demands and care coordination. Assistive technologies can enhance autonomy, social connection and health management, but their long-term adoption and use remain limited.</div></div><div><h3>Methods</h3><div>Within the Pharaon project, the study explored factors influencing technology use among older adults after 12 months of use. The research focused on the Italian pilot while integrating insights from other European sites. Qualitative data from semi-structured interviews were analyzed thematically and validated through reflection meetings and iterative online questionnaires. Recurring factors were consolidated, prioritized and mapped to explore interconnections affecting adoption.</div></div><div><h3>Results</h3><div>Initial findings revealed 30 factors, including personal characteristics, motivational aspects, technical design, social dynamics, and environmental context. These were rated by participating teams and 25 were retained for deeper analysis. An interconnection analysis explored how these factors influenced one another. Highly interconnected factors, such as social context, personalized training, ease of use, and cognitive changes, were central to understanding and improving technology adoption. In contrast, aspects like education level, internet access, and technical requirements appeared less interconnected, indicating more isolated effects.</div></div><div><h3>Discussion</h3><div>This multi-site, empirical investigation emphasizes the complex and interconnected nature of technology adoption among older adults. Our findings underscore the need for user-centered design, tailored training, and sensitivity to social context. Furthermore, the interactive mapping of interrelated factors provides a practical framework for developers and policymakers to target impactful interventions. To our knowledge, this is the first cross-pilot study to empirically define and score factors affecting older adults’ technology use based on multi-pilot data across Europe, offering valuable insights for long-term, meaningful engagement.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"206 ","pages":"Article 106179"},"PeriodicalIF":4.1,"publicationDate":"2025-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145524787","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}
Pub Date : 2025-11-08DOI: 10.1016/j.ijmedinf.2025.106182
Jianhua Wan , Yaoyu Zou , Maobin Kuang , Shixuan Xiong , Wenhua He, Yin Zhu, Nonghua Lu, Liang Xia
Aims
This study aims to evaluate the predictive value of cumulative creatinine exposure (CumCr) and dynamic creatinine trajectories for severe acute pancreatitis (SAP) prognosis and construct a machine learning-based risk stratification model.
Methods
An international multicenter retrospective cohort study included SAP patients from the Nanchang cohort (n = 1,545) and the MIMIC-IV database (n = 530). CumCr during the first 7 days of hospitalization was calculated. Latent class growth modeling (LCGM) identified creatinine trajectory patterns, and restricted cubic spline (RCS) analysis explored non-linear relationships between CumCr and mortality. The Boruta algorithm screened variables, and LASSO regression construct a nomogram model.
Results
A non-linear association between CumCr and mortality was observed, with a distinct threshold (K = 964.029 in the Nanchang cohort). Below this threshold, each standard deviation increases in CumCr elevated mortality risk 12-fold (OR = 12.135). LCGM classified four creatinine trajectory patterns. The persistently high-level group (PHL-T4) exhibited the highest mortality (43 % vs. 6.8 % in persistently low-level group (PLL-T1), P < 0.001). The nomogram integrated age, heart rate, calcium, and creatinine trajectories, achieving an AUC of 0.79, outperforming APACHE II (0.68) and SIRS (0.58). External validation yielded an AUC of 0.71.
Conclusions
This study combining CumCr and trajectory modeling demonstrates that dynamic creatinine monitoring enhances early identification of high-risk SAP patients.
{"title":"Predictive value of dynamic creatinine cumulative exposure and trajectory classification on mortality in severe acute pancreatitis: a multicenter retrospective cohort study","authors":"Jianhua Wan , Yaoyu Zou , Maobin Kuang , Shixuan Xiong , Wenhua He, Yin Zhu, Nonghua Lu, Liang Xia","doi":"10.1016/j.ijmedinf.2025.106182","DOIUrl":"10.1016/j.ijmedinf.2025.106182","url":null,"abstract":"<div><h3>Aims</h3><div>This study aims to evaluate the predictive value of cumulative creatinine exposure (CumCr) and dynamic creatinine trajectories for severe acute pancreatitis (SAP) prognosis and construct a machine learning-based risk stratification model.</div></div><div><h3>Methods</h3><div>An international multicenter retrospective cohort study included SAP patients from the Nanchang cohort (n = 1,545) and the MIMIC-IV database (n = 530). CumCr during the first 7 days of hospitalization was calculated. Latent class growth modeling (LCGM) identified creatinine trajectory patterns, and restricted cubic spline (RCS) analysis explored non-linear relationships between CumCr and mortality. The Boruta algorithm screened variables, and LASSO regression construct a nomogram model.</div></div><div><h3>Results</h3><div>A non-linear association between CumCr and mortality was observed, with a distinct threshold (K = 964.029 in the Nanchang cohort). Below this threshold, each standard deviation increases in CumCr elevated mortality risk 12-fold (OR = 12.135). LCGM classified four creatinine trajectory patterns. The persistently high-level group (PHL-T4) exhibited the highest mortality (43 % vs. 6.8 % in persistently low-level group (PLL-T1), P < 0.001). The nomogram integrated age, heart rate, calcium, and creatinine trajectories, achieving an AUC of 0.79, outperforming APACHE II (0.68) and SIRS (0.58). External validation yielded an AUC of 0.71.</div></div><div><h3>Conclusions</h3><div>This study combining CumCr and trajectory modeling demonstrates that dynamic creatinine monitoring enhances early identification of high-risk SAP patients.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"206 ","pages":"Article 106182"},"PeriodicalIF":4.1,"publicationDate":"2025-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145497541","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}
Pub Date : 2025-11-08DOI: 10.1016/j.ijmedinf.2025.106175
Melissa L. Kramer , Aileen K. Ho , Emil Lamprecht , Kayleigh Maxwell , Abigail F. Newlands , Jessica L. Price , Lindsey Roberts , Katherine A. Finlay
Background
Clinical trial recruitment faces significant challenges, with 55% of trials terminated due to low enrolment and more than 80% failing to reach targets on time. While digital recruitment strategies show promise, standardised implementation frameworks using digital health informatics approaches remain underdeveloped. Referral partnerships combined with multi-platform analytics offer potential solutions but lack systematic implementation methodologies.
Objective
To demonstrate a structured methodology for implementing and measuring multi-channel digital recruitment campaigns for clinical trials using integrated analytics platforms and referral partnerships.
Methods
A six-month multi-channel digital recruitment campaign was implemented across seven channels to support two ongoing Phase III clinical trials (EAGLE studies, NCT04020341, NCT04187144), from May to October 2022. The campaign was integrated with an analytics platform to track performance across mass emails, website announcements, browser notifications, Instagram posts and three email automations. The implementation utilised both direct and indirect funnel architectures, with real-time performance optimisation.
Results
The integrated analytics framework successfully tracked 4829 clicks across seven channels, achieving an overall click-through rate (CTR) of 2.79%, substantially exceeding clinical trial banner advertisement benchmarks (0.1–0.3%) and healthcare industry Facebook advertisement standards (0.83%). Website announcements generated the highest volume (52.54% of total clicks), followed by mass emails (28.00%).
Conclusions
This study provides a replicable informatics framework for implementing analytics-driven digital recruitment campaigns for clinical trials. The methodology demonstrates how clinical trial recruiters can integrate analytics platforms and referral partners to optimise outreach and achieve performance substantially above industry benchmarks.
{"title":"Multi-Platform analytics integration for clinical trial Recruitment: A digital health informatics implementation framework","authors":"Melissa L. Kramer , Aileen K. Ho , Emil Lamprecht , Kayleigh Maxwell , Abigail F. Newlands , Jessica L. Price , Lindsey Roberts , Katherine A. Finlay","doi":"10.1016/j.ijmedinf.2025.106175","DOIUrl":"10.1016/j.ijmedinf.2025.106175","url":null,"abstract":"<div><h3>Background</h3><div>Clinical trial recruitment faces significant challenges, with 55% of trials terminated due to low enrolment and more than 80% failing to reach targets on time. While digital recruitment strategies show promise, standardised implementation frameworks using digital health informatics approaches remain underdeveloped. Referral partnerships combined with multi-platform analytics offer potential solutions but lack systematic implementation methodologies.</div></div><div><h3>Objective</h3><div>To demonstrate a structured methodology for implementing and measuring multi-channel digital recruitment campaigns for clinical trials using integrated analytics platforms and referral partnerships.</div></div><div><h3>Methods</h3><div>A six-month multi-channel digital recruitment campaign was implemented across seven channels to support two ongoing Phase III clinical trials (EAGLE studies, NCT04020341, NCT04187144), from May to October 2022. The campaign was integrated with an analytics platform to track performance across mass emails, website announcements, browser notifications, Instagram posts and three email automations. The implementation utilised both direct and indirect funnel architectures, with real-time performance optimisation.</div></div><div><h3>Results</h3><div>The integrated analytics framework successfully tracked 4829 clicks across seven channels, achieving an overall click-through rate (CTR) of 2.79%, substantially exceeding clinical trial banner advertisement benchmarks (0.1–0.3%) and healthcare industry Facebook advertisement standards (0.83%). Website announcements generated the highest volume (52.54% of total clicks), followed by mass emails (28.00%).</div></div><div><h3>Conclusions</h3><div>This study provides a replicable informatics framework for implementing analytics-driven digital recruitment campaigns for clinical trials. The methodology demonstrates how clinical trial recruiters can integrate analytics platforms and referral partners to optimise outreach and achieve performance substantially above industry benchmarks.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"206 ","pages":"Article 106175"},"PeriodicalIF":4.1,"publicationDate":"2025-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145528377","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}
Pub Date : 2025-11-06DOI: 10.1016/j.ijmedinf.2025.106174
Fayez Alahmri, Shibili Nuhmani, Qassim Muaidi
Objective
The success of telerehabilitation in comparison with conventional rehabilitation in the management of chronic low back pain remains uncertain. This study investigates the effectiveness of telerehabilitation in reducing pain and disability in chronic low back pain, compared to conventional rehabilitation.
Methods
A search was conducted across electronic databases, including Medline/PubMed, PEDro, and the Cochrane Library. The methodological quality of the included studies was assessed using the Cochrane Risk of Bias tool (ROB 2). Data synthesis included pooled mean differences with 95% confidence intervals (CI) calculated using a random-effects model. The two groups’ mean differences in function and pain intensity were the main outcomes. Heterogeneity was evaluated using I2 and Chi2 tests, and pooled mean differences and 95% CI were calculated using a random-effects model.
Results
Eight studies were included in the review. The pain intensity findings showed the pooled mean difference was 0.12 (95 % CI: −0.23 to 0.48), slightly favoring the control group. For functional disability, the pooled mean difference was −0.43 (95 % CI: −4.12 to 3.26), slightly favoring telerehabilitation; however, both results were not statistically significant. Significant heterogeneity was observed across studies for both outcomes (I2 = 43 % for pain intensity, I2 = 70 % for functional disability), suggesting substantial variability among the included studies.
Conclusion
No significant differences were found between telerehabilitation and conventional rehabilitation for chronic low back pain, but the small number of studies and heterogeneity limit firm conclusions. Larger trials are needed to confirm these findings.
{"title":"Effectiveness of telerehabilitation on chronic low back Pain: Systematic review and Meta-Analysis","authors":"Fayez Alahmri, Shibili Nuhmani, Qassim Muaidi","doi":"10.1016/j.ijmedinf.2025.106174","DOIUrl":"10.1016/j.ijmedinf.2025.106174","url":null,"abstract":"<div><h3>Objective</h3><div>The success of telerehabilitation in comparison with conventional rehabilitation in the management of chronic low back pain remains uncertain. This study investigates the effectiveness of telerehabilitation in reducing pain and disability in chronic low back pain, compared to conventional rehabilitation.</div></div><div><h3>Methods</h3><div>A search was conducted across electronic databases, including Medline/PubMed, PEDro, and the Cochrane Library. The methodological quality of the included studies was assessed using the Cochrane Risk of Bias tool (ROB 2). Data synthesis included pooled mean differences with 95% confidence intervals (CI) calculated using a random-effects model. The two groups’ mean differences in function and pain intensity were the main outcomes. Heterogeneity was evaluated using I<sup>2</sup> and Chi<sup>2</sup> tests, and pooled mean differences and 95% CI were calculated using a random-effects model.</div></div><div><h3>Results</h3><div>Eight studies were included in the review. The pain intensity findings showed the pooled mean difference was 0.12 (95 % CI: −0.23 to 0.48), slightly favoring the control group. For functional disability, the pooled mean difference was −0.43 (95 % CI: −4.12 to 3.26), slightly favoring telerehabilitation; however, both results were not statistically significant. Significant heterogeneity was observed across studies for both outcomes (I<sup>2</sup> = 43 % for pain intensity, I<sup>2</sup> = 70 % for functional disability), suggesting substantial variability among the included studies.</div></div><div><h3>Conclusion</h3><div>No significant differences were found between telerehabilitation and conventional rehabilitation for chronic low back pain, but the small number of studies and heterogeneity limit firm conclusions. Larger trials are needed to confirm these findings.</div><div><strong>Protocol Registration</strong>: PROSPERO (CRD42022375048)</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"206 ","pages":"Article 106174"},"PeriodicalIF":4.1,"publicationDate":"2025-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145467309","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号