EUVIMEDScopri EUVIMED
Mixed-Methods-StudieOpen access

Evaluating the Clinical Effectiveness and Patient Experience of a Large Language Model–Based Digital Tool for Home-Based Blood Pressure Management: Mixed Methods Study

Alan Jelic iD, Igor Sesto iD, Luka Rotkvic iD, Luka Pavlovic iD, Nikola Erceg iD, Nina Sesto iD, Zeljko Kraljevic iD, Joshua Au Yeung iD, Amos Folarin iD, Richard Dobson iD, Petroula Laiou iD

JMIR mHealth and uHealth · 2025-11-03 · 13 · e68361-e68361

DOI: 10.2196/68361 ↗

At a glance

Mixed-Methods-Studie in JMIR mHealth and uHealth, published online on 2025-11-03. Written by Alan Jelic and 10 co-authors. Institution: Magdalena Clinic for Cardiovascular Diseases, Zagreb, Croatia. Major MEDLINE topics: Hypertension, Blood Pressure Determination, Blood Pressure Monitoring, Ambulatory. Cited 5 times so far (Crossref); 15 references. The full text is freely available in PubMed Central. Funded by: Wellcome Trust.

AbstractSubjectsDati bibliograficiCiteRelated studies

Abstract (original language)

Background

Hypertension, one of the most common cardiovascular conditions worldwide, necessitates comprehensive management due to its association with multiple health risks. Effective control often involves lifestyle changes and continuous monitoring, yet many individuals struggle to adhere to traditional management approaches. Digital health tools are emerging as promising alternatives, offering remote monitoring and real-time support. This study focuses on evaluating a digital tool specifically designed for hypertension management, analyzing its effectiveness, and gathering user perspectives on its functionality and impact.

Objective

The primary objective of this study is to assess the effectiveness of a digital health tool in managing hypertension. Additionally, the study aims to understand user experiences and satisfaction levels to gauge the tool’s acceptance and potential for long-term use. By analyzing data from a large cohort, we seek to determine whether the tool can contribute to meaningful reductions in blood pressure and support sustained engagement over time.

Methods

The study includes a cohort of 5136 participants who used the digital hypertension management tool. This tool provides continuous blood pressure monitoring, real-time feedback, and personalized health recommendations, which are crucial for tailored intervention. Participants recorded their blood pressure values over time, and we tracked retention rates to measure adherence. An online survey was administered to gather user feedback, focusing on ease of use, satisfaction levels, and perceived health benefits.

Results

Our analysis indicates a significant reduction in blood pressure values among users, with a positive correlation observed between the duration of use and the extent of blood pressure reduction. We performed a 1-sided Wilcoxon Rank Sum test to compare systolic blood pressure values in the first and last biweekly use intervals, and descriptive statistics were used to assess survey responses. High retention rates were observed, with 2583 (50.3%) participants using the tool after 1 year. Survey responses revealed high satisfaction, with users highlighting the tool’s ease of use and noting reduced anxiety related to blood pressure management. These results suggest that users found the digital tool both effective and convenient.

Conclusions

This study demonstrates the potential benefits of digital health tools in managing hypertension, emphasizing their ability to engage users over long periods and support blood pressure reduction. The high satisfaction rates and positive user feedback underscore the importance of user-centered design in creating effective health interventions. Overall, the findings suggest that digital tools, when designed with a focus on user experience, could be a valuable component in hypertension management strategies, complementing traditional health care approaches.

Subjects

MeSH terms (MEDLINE)

HypertensionBlood Pressure DeterminationBlood Pressure Monitoring, AmbulatoryHumansMaleFemaleMiddle AgedAgedSurveys and QuestionnairesAdultPatient SatisfactionCohort StudiesTelemedicineBlood PressureLarge Language Models

Author keywords

AILLMsartificial intelligencedigital toolshypertensionlarge language models

Authors and institutions

  • Magdalena Clinic for Cardiovascular Diseases, Zagreb, Croatia.
  • Megi Health Ltd, London, United Kingdom.
  • Department of Psychology, Faculty of Humanities and Social Sciences, University of Zagreb, Zagreb, Croatia.
  • King's College London, 16 De Crespigny Park, London, SE5 8AB, United Kingdom, 44 020 7848 0002.
  • King's College Hospital NHS Foundation Trust, London, United Kingdom.

ORCID profiles: Alan Jelic, Igor Sesto, Luka Rotkvic, Luka Pavlovic, Nikola Erceg, Nina Sesto, Zeljko Kraljevic, Joshua Au Yeung, Amos Folarin, Richard Dobson, Petroula Laiou

Funding

  • Wellcome Trust

Cite

APA

Jelic, A., Sesto, I., Rotkvic, L., Pavlovic, L., Erceg, N., Sesto, N., Kraljevic, Z., Au Yeung, J., Folarin, A., Dobson, R., & Laiou, P. (2025). Evaluating the Clinical Effectiveness and Patient Experience of a Large Language Model–Based Digital Tool for Home-Based Blood Pressure Management: Mixed Methods Study. JMIR mHealth and uHealth, 13, e68361-e68361. https://doi.org/10.2196/68361

Vancouver

Jelic A, Sesto I, Rotkvic L, Pavlovic L, Erceg N, Sesto N, et al. Evaluating the Clinical Effectiveness and Patient Experience of a Large Language Model–Based Digital Tool for Home-Based Blood Pressure Management: Mixed Methods Study. JMIR mHealth and uHealth. 2025;13:e68361-e68361. doi:10.2196/68361

BibTeX
@article{Jelic2025,
  author = {Jelic, Alan and Sesto, Igor and Rotkvic, Luka and Pavlovic, Luka and Erceg, Nikola and Sesto, Nina and Kraljevic, Zeljko and Au Yeung, Joshua and Folarin, Amos and Dobson, Richard and Laiou, Petroula},
  title = {Evaluating the Clinical Effectiveness and Patient Experience of a Large Language Model–Based Digital Tool for Home-Based Blood Pressure Management: Mixed Methods Study},
  journal = {JMIR mHealth and uHealth},
  year = {2025},
  volume = {13},
  pages = {e68361--e68361},
  doi = {10.2196/68361}
}
Start for free →