Effect of a mobile health intervention optimised with artificial intelligence on blood pressure and other cardiovascular risk factors in adults with high blood pressure: rationale and design of My Intelligent Cardiac Assistant (MICArdiac) randomised controlled trial - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Open Heart . 2026 Apr 9;13(1):e003819. doi: 10.1136/openhrt-2025-003819 Search in PMC Search in PubMed View in NLM Catalog Add to search Effect of a mobile health intervention optimised with artificial intelligence on blood pressure and other cardiovascular risk factors in adults with high blood pressure: rationale and design of My Intelligent Cardiac Assistant (MICArdiac) randomised controlled trial Liliana Laranjo Liliana Laranjo 1 Westmead Applied Research Centre, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia Find articles by Liliana Laranjo 1, ✉ , Edel T O’Hagan Edel T O’Hagan 1 Westmead Applied Research Centre, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia Find articles by Edel T O’Hagan 1 , Harry Klimis Harry Klimis 1 Westmead Applied Research Centre, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia Find articles by Harry Klimis 1 , Simone Marschner Simone Marschner 1 Westmead Applied Research Centre, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia Find articles by Simone Marschner 1 , Clara K Chow Clara K Chow 1 Westmead Applied Research Centre, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia 2 Westmead Hospital, Westmead, New South Wales, Australia Find articles by Clara K Chow 1, 2 Author information Article notes Copyright and License information 1 Westmead Applied Research Centre, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia 2 Westmead Hospital, Westmead, New South Wales, Australia Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise. Additional supplemental material is published online only. To view, please visit the journal online ( https://doi.org/10.1136/openhrt-2025-003819 ). None declared. ✉ Dr Liliana Laranjo; [email protected] Received 2025 Oct 27; Accepted 2026 Mar 5; Collection date 2026. Copyright © Author(s) (or their employer(s)) 2026. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ Group. This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: https://creativecommons.org/licenses/by-nc/4.0/ . PMC Copyright notice PMCID: PMC13084819 PMID: 41956556 Abstract Introduction Control of blood pressure (BP) continues to be a challenge globally. Clinical trials have shown home BP monitoring and text-message interventions to lower BP. Integrating these to personalise supportive messaging in response to changing BP and activity through leveraging artificial intelligence (AI) could improve BP control. The aim of this trial is to examine the impact on BP, compared with a text-message only programme, of the My Intelligent Cardiac Assistant (MICArdiac) programme. MICArdiac is a 6-month programme comprising tracking of BP, physical activity and heart rate informing a content stream of AI-driven personalised messages to encourage BP self-management, as-necessary medical review to up-titrate medicines and cardiovascular preventative behavioural change. Methods and analysis MICArdiac is a prospective randomised open-blinded endpoint randomised controlled trial with 1:1 (intervention:control) allocation and active control. Individuals aged 35 years old or above with high BP are randomised to the MICArdiac programme or to a text-message cardiovascular education programme. Recruitment started in primary care and hospital clinics in Western Sydney and moved to decentralised direct-to-community recruitment in the Australian Capital Territory, New South Wales and Victoria. Randomisation and allocation concealment occur via a secure web-based system. The primary outcome is mean daytime systolic BP measured by 24-hour Ambulatory Blood Pressure Monitoring at 6 months. Primary analysis will follow intention-to-treat principles; data analysts will be blinded. Process evaluation will be conducted. Ethics and dissemination Ethics approval was obtained from Western Sydney Local Health District Human Ethics Research Committee (2021/ ETH11379 ). Trial registration number ACTRN12622000091707. Keywords: Hypertension, Health Services, Telemedicine WHAT IS ALREADY KNOWN ON THIS TOPIC High blood pressure (BP) is the leading modifiable driver of cardiovascular disease burden. Mobile health technology can decrease BP through lifestyle behaviour change, but it is unknown whether leveraging artificial intelligence and self-monitoring device data to personalise messages can be more effective than text messaging-based cardiovascular education. WHAT THIS STUDY ADDS My Intelligent Cardiac Assistant (MICArdiac) integrates BP home-monitoring and activity monitoring with artificial intelligence enhanced personalised messaging to encourage BP self-management and as-needed medical review. The effect of MICArdiac on daytime average systolic BP measured using 24-hour ambulatory BP monitor at 6 months is compared with a text message only programme in this randomised controlled trial. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY MICArdiac may lead to higher decreases in BP than text-message only programmes and therefore offer a scalable, effective and engaging solution for improving BP control in the community. Introduction Cardiovascular disease (CVD) remains the largest cause of premature death and disability worldwide. 1 Much of the burden of CVD is attributable to poor blood pressure (BP) control as well as other uncontrolled CVD risk factors. 2 Optimising health behaviours, including lifestyle and medication adherence, is a key strategy to improve BP control and reduce CVD risk. 3 , 10 A range of CVD prevention interventions that provide education, health behaviour change support and thus encourage self-management of high BP have been shown to improve BP and CVD risk factors known to be associated with reduction in future cardiovascular events. 4 11 Yet common barriers to scale-up involve access and availability, heavy resource utilisation including clinician time and patient engagement. 6 11 Mobile health (mHealth) technologies such as text messaging, mobile apps and wireless monitoring devices may help overcome these barriers. A 2023 systematic review and meta-analysis found that digital interventions (smartphone apps, websites or SMS) were associated with a significant reduction in systolic (mean difference −3.62 mm Hg (95% CI −5.22 to −2.02)) and diastolic (mean difference −2.45 mm Hg (95% CI −3.83 to −1.07)) BP compared with control group. 12 Heterogeneity was high and there was wide variability between the different types of app interventions: some focused only on medication adherence 13 or on a specific lifestyle behaviour (eg, diet). 14 Meta-regression was not conducted to identify specific app components associated with higher effectiveness. More recent trials evaluating standalone app interventions (ie, without clinician involvement) for patients with high BP have found mixed results. 15 16 A trial comparing self-monitoring of BP with app-based self-monitoring (ie, via connected wireless BP monitor) did not find a significant difference in BP, 15 but the intervention did not provide support for lifestyle behaviour change or medication adherence. Another trial evaluated app-based self-monitoring and lifestyle behaviour change support for patients not medicated with antihypertensives, showing a small significant decrease (~2 mm Hg) in systolic BP compared with usual care. 16 Finally, a trial evaluating an app-based self-monitoring and lifestyle behaviour change support via conversational artificial intelligence (AI) in comparison to just app-based self-monitoring did not find a significant difference in BP. 17 However, none of the interventions in these trials provided education and support on medication adherence, nor did they support patients in seeking medication up-titration for BP control where needed. To date, no trial has evaluated a comprehensive mHealth programme to support patients with high BP self-manage. Our proposed intervention incorporates app-based monitoring of BP, heart rate and physical activity (via connected wireless devices), AI-enhanced personalised education and behaviour change support (for lifestyle behaviours and medication adherence) and prompts with letters for patients to show their General Practitioner (GP) or cardiologist for a medication review as triggered through algorithms for non-controlled BP. Novel AI methods, including machine learning and natural language processing, have been suggested as promising approaches for more engaging and personalised self-management interventions, 18 , 20 but trials assessing their effectiveness in BP management are sparse. 17 The primary objective of this trial is to assess the impact of an AI enhanced mHealth programme—the My Intelligent Cardiac Assistant (MICArdiac)—compared with standard text messaging education for CVD prevention, on daytime average ambulatory systolic BP, in participants with high BP. Secondary objectives are to determine the impact of the intervention on other BP measures and CVD risk factors, quality of life, cardiovascular knowledge and unplanned hospitalisations as well as conduct a process evaluation, assessing engagement, acceptability and implementation of the intervention. Methods and analysis Study design MICArdiac is a single-blind 1:1 (intervention:control) randomised controlled trial (RCT) with 6 months of follow-up ( figure 1 ). The trial has been registered in the Australian and New Zealand Clinical Trials Registry (Registration number: ACTRN12622000091707) and the study protocol is reported following the SPIRIT-AI checklist 21 for RCT protocols ( online supplemental 1 ) and the mHealth evidence reporting and assessment checklist 22 ( online supplemental 2 ). Figure 1. Study flow diagram. 24h ABPM, 24-hour ambulatory blood pressure monitoring; AND, and; DBP, diastolic blood pressure; MICArdiac, My Intelligent Cardiac Assistant; SBP, systolic blood pressure. Open in a new tab Study setting This study commenced recruitment in Western Sydney, NSW, Australia, with the initial sites as Westmead Hospital outpatient services and General Practice clinics in Western Sydney Primary Health Network. Recruitment was then decentralised and extended to community recruitment in New South Wales, Victoria and the Australian Capital Territory. In Australia, 9 out of 10 adults own a smartphone. 23 Participants Potential participants are eligible to participate if they are 35 years old or above, own a compatible smartphone—iPhone 5 onwards (iOS 10 and higher) or Android (6.0 and higher) and have uncontrolled hypertension or high BP, satisfying one of the following criteria: Known hypertension AND at least one recorded measure of systolic BP ≥140 mm Hg and/or diastolic BP ≥90 mm Hg in the last 12 months; OR. At least one recorded measure of systolic BP ≥160 mm Hg and/or diastolic BP ≥100 mm Hg in the last 12 months; OR. At least two recorded measures of systolic BP ≥140 mm Hg and/or diastolic BP ≥90 mm Hg in the last 12 months. Enrolment from a single reading ≥160/100 mm Hg or two readings ≥140/90 mm Hg is based on a high prevalence of elevated BP in community BP screening 24 and aligns with international guidelines stating that, although ‘a single screening office BP alone does not typically have sufficient diagnostic test performance to establish a diagnosis, especially for BP values close to diagnostic thresholds’, an ‘office BP >160/100 mm Hg is almost always consistent with a diagnosis of hypertension’. 10 Participants will be excluded from the study if they required acute care in the preceding 30 days in relation to CVD (eg, heart failure, coronary heart disease, stroke, transient ischaemic attack). Additional exclusion criteria include being unable to use the monitoring devices provided; being unable to complete the study procedures or follow-up; being unable to understand sufficient written English to provide informed consent; or having a concomitant illness, physical impairment or mental condition which in the opinion of the study team/primary physician could interfere with the conduct of the study. Participant flow through the study is presented in figure 1 . Intervention: the MICArdiac programme All participants randomised into the intervention arm will receive the MICArdiac programme for 6 months. The MICArdiac programme ( online supplemental 3 ) aims to improve BP and other cardiovascular risk factors and consists of cardiovascular education and behaviour change support through an mHealth programme—MICArdiac smartphone App ( figures2 3 )—that is connected with wireless monitoring devices (physical activity tracker and BP monitor). After randomisation, devices are provided to intervention participants (in person or mailed) with instructions to download the MICArdiac app and set up the devices; a research assistant is available to support and troubleshoot the setup process and any emerging device issues during the study. Figure 2. MICArdiac app and devices. MICArdiac, My Intelligent Cardiac Assistant. Open in a new tab Figure 3. MICArdiac features. MICArdiac, My Intelligent Cardiac Assistant. Open in a new tab The MICArdiac programme comprises tracking of BP, physical activity and heart rate, and a content stream of AI-driven personalised messages (delivered via mobile app or text messages) to encourage BP self-management and as-necessary medical review to up-titrate medicines in addition to CVD prevention programme content. By integrating data from wireless devices and goal-related messages to deliver personalised self-monitoring and lifestyle modification support, the programme incorporates several behaviour change techniques, based on the Capability Opportunity Motivation Behaviour (COM-B) model 25 26 ( online supplemental 4 ). Wireless monitoring devices Devices include a wireless, upper-arm cuff, BP monitor (Withings BPM Connect, validated) 27 and recommended by the STRIDE BP, the international scientific organisation dedicated to ensuring the accuracy of BP measurement 28 and a physical activity tracker with heart rate monitoring ability (Xiaomi Mi Smart Band or a ‘Bring Your Own device’ option, depending on participant preference). Participants are encouraged to always wear the activity tracker, removing the bracelet only when required to charge it. Participants are instructed to measure their BP using the wireless device two times a week at different times of the day. MICArdiac app The MICArdiac app graphically displays data collected from the wireless monitoring devices (step counts, exercise duration, BP, heart rate) and delivers personalised messages utilising data from the wireless monitoring devices, AI or threshold-based algorithms (see the Messages section). The MICArdiac app is downloadable via the Apple and Google app stores, and it securely collects the participant’s BP measures, heart rate and physical activity measures from the smartphone’s Google Fit app or Apple Healthkit app. Messages The frequency of messages will vary between 4 and 7 messages per week. Messages can be delivered as in-app or text message (ie, Short Message Service), depending on patient preference. The MICArdiac intervention is designed to be interactive and supportive with participants encouraged to engage with the health counsellor if they have any questions or concerns. Messages will be of three main types: Educational messages to encourage behaviour change and self-management, drawn from a message bank of at least five topic streams: BP (including medication adherence), cholesterol, diet, physical activity, support and smoking (if applicable). Messages are customised using one baseline variable (smoking status) and their content and sequence are optimised by a machine learning algorithm focusing on message intent, lexical similarity and message readability (based on participant’s health literacy, measured at baseline). The machine learning algorithm considers these three factors (intent, lexical similarity and readability) to produce a composite score for each candidate message (message scheduled to be delivered next) by comparing its text to the most recent messages that have been sent to the participant, using an adaptation of the antirepetition scoring system developed by Foster and White, and the SoftMax function to assign the highest selection probability to the best-scoring candidate message in the database. 29 , 34 The algorithm was developed based on natural language processing analysis of messages and responses from our previous RCTs (further information on the AI-enhanced message delivery protocol available in online supplemental 3 ). 35 36 The content of the educational messages was adapted from previous trials and developed by a team of clinicians (GP, cardiologists, physiotherapists, dieticians) based on national guidelines and information from reputable sources (eg, National Heart Foundation) and refined with consumer input ( online supplemental 5 ). Adaptive algorithm messages containing weekly adaptive step goals based on data from the physical activity tracker. The adaptive algorithm first monitors the participant’s step count for 1 week at baseline and then it generates a goal for next week based on this number. The algorithm then generates a new goal each week based on the participant’s step count and their ability to achieve the previous week’s goal. The aim is to gradually increase participants’ step goals in achievable increments. Threshold-based messages based on data from the wireless monitoring devices. These messages will be rule-based and one of three types: (1) messages indicating that the values of health parameters (BP, heart rate) are within the normal/recommended range and encouraging maintenance of health behaviours and management; (2) messages indicating that the values are out of the normal/recommended range and encouraging improvement of health behaviours. When BP thresholds are met, these will trigger messages to prompt patients to go to their usual doctors for medical review and the message prompting this will embed a letter of referral to their doctor containing high BP information and why they need medical review. Extreme values of BP and heart rate will trigger a phone call from a health counsellor; and (3) messages reminding participants to use wireless monitoring devices, triggered by a lack of data received in the previous week. The MICArdiac programme development Before commencing recruitment, the MICArdiac programme was evaluated in a single-arm pilot study to assess user experience and refine the intervention. Nine participants were purposively recruited to ensure diversity in age, sex, cultural background, lifestyle and baseline cardiovascular risk. Each participant received a wireless BP monitor, a physical activity tracker and instructions to download the MICArdiac app. Over a 2-week period, participants were asked to measure their BP two times a week, wear the activity tracker continuously and engage with 4 weekly text messages delivered via the app. Participants completed an online daily diary for 14 days and were invited to a semistructured interview within 7 days of the intervention period. Independent researchers with no prior relationship to participants conducted interviews, lasting 40–60 min, using a guide developed by the study team. Participants discussed each component of the intervention for its standalone merit and its synergistic value within the broader MICArdiac intervention. The pilot study highlighted acceptability and perceived utility, while identifying three areas requiring further refinement: (1) syncing and connectivity, (2) message personalisation and (3) device compatibility. We addressed these concerns by: (1) developing user manuals and dedicated support staff to streamline syncing, (2) refining message content through iterative consultation with consumers, clinicians and researchers, and supplementing messages with credible information links, and (3) expanding inclusion criteria and enhancing app compatibility with diverse wearable devices. The full details of the pilot user experience study are available in online supplemental 5 . Control group The control arm will receive a standard text message programme, similar to published trials TEXTME 35 and TextMe2, 36 which demonstrated efficacy in improving modifiable CVD risk factors. Messages will be randomly selected from five topic streams (general cardiovascular health, diet, physical activity, smoking and medication adherence) and will be customised using two baseline variables (smoking status and vegetarian diet). Four messages will be sent per week and delivered as a text message. Messages will be unidirectional, that is, participants will not be encouraged to respond to the messages and any responses provided (although monitored) will not be answered with a reply. Where there are return messages which are of a clinical nature, the message will be escalated to a medical doctor from the research team for review. Outcomes and data collection The primary outcome is the difference in mean daytime systolic BP measured by 24-hour ambulatory BP monitoring (ABPM) between intervention and control group at 6 months. Secondary outcomes include 24-hour mean systolic and diastolic BP; first systolic and diastolic BP measured by ABPM (equating to office BP); mean daytime diastolic BP; night time mean systolic and diastolic BP (all measured by ABPM); controlled SBP (daytime SBP below 135 mm Hg daytime average ABPM); Low-Density Lipoprotein (LDL) cholesterol, total cholesterol, High-Density Lipoprotein (HDL) cholesterol and triglycerides; controlled LDL cholesterol (under 1.8 mmol/L); BMI; diet (fruit and vegetable intake); physical activity; smoking cessation; alcohol intake; total number of risk factors controlled (systolic BP, LDL, BMI, diet, physical activity, smoking); hypertension knowledge; GP visits; BP medication up-titration; cholesterol-lowering medication initiation. There will be two laboratory data collection timepoints: one at baseline and another postdelivery of the 6-month programme. At both time points, ABPM and lipid profile blood test assessments can be completed at community pathology sites using a pathology referral form provided to the participant; participants will also have the option of completing the ABPM with a study-provided machine, either fitted by a research assistant on site or mailed to the participant with instructions and available phone support. All ABPM machines used in the study are validated (community pathology laboratories use the IEM Mobile-O-Graph Self-reported and the Norav NBP-24 NG; study-owned machines are the Suntech Oscar 2 and Somnomedics ABPM Pro). Participant surveys will be completed remotely via a web link. At baseline and 6-month assessments, a research assistant will collect information including medical history and medication data from participants, in person or via the phone. Eligibility criteria, medical history and health literacy (BRIEF health literacy survey) 37 will be collected at time of registration. In addition, participants will complete questionnaires to self-report on lifestyle outcomes and healthcare service utilisation at baseline and 6-month follow-up. Participants in the intervention group will be invited to participate in the process evaluation. Full details of the outcome data collected are presented in table 1 . Table 1. Study outcomes and collection time points. Outcome Measure SBP, DBP Average SBP and DBP measured using 24 hours ABPM at baseline and 6 months postrandomisation (first, diurnal, nocturnal, overall). Fasting LDL-C, TC, HDL-C, triglycerides (mmol/L) Fasting blood sample at baseline (or done in last 30 days) and 6 months post-randomisation. BMI Self-reported weight and height to calculate BMI at baseline and 6 months postrandomisation. Diet (vegetables and fruit consumption) Self-reported: validated single-question for fruit intake and for vegetable intake 52 at baseline and 6 months postrandomisation Physical activity Self-reported at baseline and 6 months postrandomisation Smoking cessation Self-reported smoking status at baseline and 6 months postrandomisation Alcohol use Self-reported alcohol use at baseline and 6 months postrandomisation Total number of risk factors controlled Guideline-controlled SBP, LDL, BMI, diet, physical activity and smoking at baseline and 6 months postrandomisation Hypertension knowledge Self-reported—modified version of Hypertension Evaluation of Lifestyle and Management Knowledge Scale 53 at baseline and 6 months postrandomisation GP visits Self-reported number of primary care visits, in the past 6 months, collected at baseline and 6 months postrandomisation. Medication up-titration/initiation Self-reported BP medication usage, medication class, medication dosage at baseline and 6 months postrandomisation; self-reported cholesterol lowering medication initiation at 6 months Acceptability and experience Questionnaire administered at 6 months Open in a new tab ABPM, ambulatory blood pressure monitoring; BMI, Body Mass Index; BP, blood pressure; DBP, diastolic blood pressure; GP, General Practitioner; HDL-C, High-Density Lipoprotein Cholesterol; LDL-C, Low-Density Lipoprotein Cholesterol; SBP, Systolic Blood Pressure; TC, total cholesterol. A process evaluation will be conducted following the Medical Research Council framework 38 to assess feasibility, engagement and implementation, through the use of intervention usage metrics , online questionnaire and semistructured interviews with participants. Surveys and interviews will explore acceptability, usability, barriers and enablers to engagement; user preferences regarding personalised messages; and patient perspectives on the use of AI for optimisation of educational content delivery. Intervention usage metrics will be automatically collected via a time-stamped electronic log recording app usage (eg, frequency and duration of use of app features and overall app use; messages received and liked; data collected from wireless monitoring devices). Additionally, a log will be kept of responses received from participants and when participants contact the study team, including the reason for contact and the method used. To examine acceptability and feasibility of MICArdiac, all participants in the intervention group at the 6-month follow-up period will be asked to complete a structured online questionnaire about their experiences of being sent health text messages customised using wireless device data and machine learning. Questions will explore participants’ acceptability, preferences and understanding regarding the messages (messages they remember, liked or disliked, whether messages were shown to family/friends, whether the messages prompted some sort of action or behaviour change, whether messages were well understood or confusing). Questions will also explore aspects related to intrusiveness, timing and content suitability of text messages as well as any potential concerns regarding privacy. Furthermore, a subset of participants in the intervention group will be invited to participate in semistructured interviews. During the interviews, participants will be asked to elaborate on their previous responses to the online questionnaire, particularly exploring what they liked or disliked about the programme, perceived utility and impact on their health behaviours and overall health, barriers and enablers to engagement; user preferences regarding personalised messages; and perspectives on the use of AI for lowering BP and improving cardiovascular health. These participants will be purposefully selected to diversify the opinions and views by choosing participants with ethnically, culturally and socioeconomically diverse backgrounds as well as with different levels of engagement with the intervention. We aim to conduct a minimum of twenty interviews, with data analysis occurring in parallel and additional individuals being recruited until there is the perception of information redundancy relative to our research objective. 39 , 46 Depending on participant preference, interviews will be conducted online, via telephone or at Westmead hospital or University of Sydney affiliated meeting space and participants will be reimbursed for their travel and parking expenses. Interviews will follow a pilot-tested interview guide and will be conducted by a trained interviewer. Interviews will be digitally recorded and transcribed. Sample size Informed by our previous TEXTME RCT, 35 we estimated that a total of 500 participants in a 1:1 ratio (250:250) will have 96% power to detect a between-group difference in mean change in ambulatory daytime average SBP of 4 mm Hg (SD 11 mm Hg) using a 5% level of significance and accounting for a drop-out rate of ~20%. Recruitment Participants are identified through cardiometabolic clinics at Westmead hospital, primary care and the community (eg, pharmacies, sports and religious organisations, libraries, community centres), starting in Western Sydney. Recruitment expansion to New South Wales, Australian Capital Territory, and Victoria is managed through social media advertising (eg, Facebook, Instagram) using study-developed materials targeting a diverse audience ( figure 4 ) and leveraging remote trial procedures. Interested participants can register their interest online by following a link or scanning a QR code. All potentially eligible participants referred to or who contact the study team directly undergo screening to determine eligibility to the study. If the member of the study team determines that a potential participant may be suitable for the study, a copy of the patient information sheet is provided to the potential participant and any questions are answered before the interested participant signs the consent form. Ineligible participants are informed and the reasons for ineligibility are documented in the screening log. Figure 4. Study advertising materials for social media. AI, artificial intelligence; MICArdiac, My Intelligent Cardiac Assistant. Open in a new tab Randomisation and blinding Randomisation will be stratified by two variables: whether or not patients are taking BP-lowering medication, and by recruitment site at primary or secondary care facility. Baseline data (including objective measures and surveys) will be collected prior to randomisation. Randomisation and allocation concealment will be conducted electronically via the Research Electronic Data Capture (REDCap) software. 47 The software will automatically allocate participants to the intervention or control group, according to the randomisation sequence generated in R (using the randomiseR package) and uploaded to REDCap, ensuring allocation concealment. Follow-up surveys will be self-filled by participants. Due to the nature of the study, participants and research study staff involved in recruitment, data collection and follow-up will not be blinded to the randomisation outcome. However, prior to final follow-up, participants will receive a text message asking them not to reveal their allocation to study personnel until after follow-up is complete. Researchers performing data analysis will be blinded. Statistical considerations and data analysis All randomised participants will be analysed according to the intention-to-treat principle, that is, subjects will be analysed in the group (intervention or control) they are assigned to during randomisation. We will follow a published framework for intention-to-treat analysis that accounts for missing outcome data 48 : (1) attempt to follow-up all randomised participants, even if they choose to stop receiving the allocated treatment (ie, discontinued treatment), as long as they have not withdrawn consent; (2) perform a main analysis of all observed data that are valid under a plausible assumption about the missing data; (3) perform sensitivity analyses to explore the effect of departures from the assumption made in the main analysis; (4) account for all randomised participants in the sensitivity analyses. Interaction analyses with predefined baseline covariates will be prespecified in the Statistical Analysis Plan prior to data lock and will include age, gender, education, health literacy level, ethnicity, use of antihypertensive therapy, BP eligibility criteria (known hypertension vs recent elevated BP readings (single reading ≥160/100 mm Hg or two readings ≥140/90 mm Hg)), and history of CVD. Continuous data will be checked for normality before performing parametric tests. Appropriate non-parametric tests will be used where data are not normally distributed. All variables will be analysed by descriptive statistical methods. The number of data available and missing data, mean, median, SD, IQR, minimum and maximum will be calculated for continuous data. Frequency tables will be generated for categorical data. For categorical variables, the number and percentage of patients with a specific level of the variable will be presented. Outcomes at 6 months will be compared between the two groups in a regression analysis adjusting for the baseline measure of the outcome. For dichotomous outcomes, we will use log-binomial regression, and for continuous outcomes, we will use linear regression. Balance in baseline variables will be considered between intervention and control groups, and if significant differences exist, additional regression analysis of outcome variables adjusting for this difference will be performed. We will report effect sizes and CIs without formal adjustment for multiple comparisons, focusing on estimation rather than dichotomous hypothesis testing. All statistical analyses will be performed using R (V.4.3.2) (R Core Team 2020). All statistical tests will be two-tailed. P values <0.05 will be considered statistically significant unless stated otherwise. A statistical analysis plan will be finalised before data lock and unblinding. Data from interviews will be analysed using reflexive thematic analysis 49 of transcribed audio-recordings, using NVivo V.12 (QRS International Pty, Melbourne, Australia). Themes will be identified using an inductive data-driven approach (ie, inductive thematic analysis). 44 The codebook will be developed and revised iteratively through discussions between researchers in the team. Identification of themes will occur by sorting the different codes into potential themes and grouping all the relevant coded data extracts within the identified themes. Themes will be identified at a semantic level, first by organising data to show patterns in semantic content, and then by interpreting the patterns and their broader meanings and implications. 44 After a candidate thematic ‘map’ is reached, the dataset will be reread to ensure the quality of the themes and refine them as needed. Reporting will follow the Consolidated Criteria for Reporting Qualitative Research checklist for reporting qualitative research. 50 Ethics and dissemination Ethics approval was obtained from Western Sydney Local Health District Human Ethics Research Committee (2021/ ETH11379 ). Informed consent will be obtained from all study participants before commencing any study procedures. Participation in the current study is entirely voluntary, and participants have the ability to withdraw at any point. Between 6 June 2023 and 1 May 2025, 451 participants have been recruited to the MICA trial. The follow-up of the last enrolled patient will be finished in November 2025. The primary results of this trial are anticipated to be available in early 2026. We aim to publish our findings in a high-reach journal to begin the process of dissemination and uptake, accompanied by a social media dissemination strategy. In addition, the findings will be presented at relevant national and international conferences. If successful, widespread implementation of the MICArdiac programme may have profound public health implications. Decreases in systolic BP of 5 to 10 mm Hg can translate into an additional 15%–20% reduction in cardiovascular events. 51 The MICArdiac programme can be conducted at relatively little expense, but if CVD can be prevented or delayed, it may result in significant reductions in morbidity and cost savings to the health system. Supplementary material online supplemental file 1 openhrt-13-1-s001.pdf (1.3MB, pdf) DOI: 10.1136/openhrt-2025-003819 Acknowledgements The authors would like to thank: Jim Cook and Viji Venkataramani from TechLab (The University of Sydney) and Joel Nothman, Marius Mather and the team at Sydney Informatics Hub (The University of Sydney) for their contributions in developing the AI-enhanced message optimisation algorithms. We thank Jason Chiang, Nicola Barrie, Arthur Shariev, Rahul Sathiaraj and the team at the Westmead Applied Research Centre for their support with various trial operation and administration tasks. We are grateful to Dr Peter Hay, Dr William Poh, Dr Yvette Castellino, Dr Sana Albasari and Dr Bharat Jain for their support with participant recruitment. Footnotes Funding: This work is supported by multiple funding sources including, NSW CV Research grant 932019, Tides Foundation (Google) and Australian Stroke and Heart Research Accelerator (ASHRA). Study funders do not have authority over study design; collection, management, analysis and interpretation of data; writing of the report; and the decision to submit the report for publication. LL is supported by a NHMRC Investigator Grant (2017642) and Sydney Horizon Fellowship. CKC is supported by a NHMRC Investigator Grant (1195326). Provenance and peer review: Not commissioned; externally peer-reviewed. Patient consent for publication: Consent obtained directly from patients. References 1. Roth GA, Mensah GA, Johnson CO, et al. Global Burden of Cardiovascular Diseases and Risk Factors, 1990-2019: Update From the GBD 2019 Study. J Am Coll Cardiol. 2020;76:2982–3021. doi: 10.1016/j.jacc.2020.11.010. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Murray CJL, Aravkin AY, Zheng P, et al. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396:1223–49. doi: 10.1016/S0140-6736(20)30752-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. O’Connor EA, Evans CV, Rushkin MC, et al. Rockville (MD): Agency for Healthcare Research and Quality; 2020. Behavioral counseling interventions to promote a healthy diet and physical activity for cardiovascular disease prevention in adults with cardiovascular risk factors: updated systematic review for the U.S. preventive services task force. Evidence synthesis, no.195. [ PubMed ] [ Google Scholar ] 4. Krist AH, Davidson KW, Mangione CM, et al. Behavioral Counseling Interventions to Promote a Healthy Diet and Physical Activity for Cardiovascular Disease Prevention in Adults With Cardiovascular Risk Factors. JAMA. 2020;324:2069. doi: 10.1001/jama.2020.21749. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Mancia G, Kreutz R, Brunström M, et al. 2023 ESH Guidelines for the management of arterial hypertension The Task Force for the management of arterial hypertension of the European Society of Hypertension: Endorsed by the International Society of Hypertension (ISH) and the European Renal Association (ERA) J Hypertens (Los Angel) 2023;41:1874–2071. doi: 10.1097/HJH.0000000000003480. [ DOI ] [ PubMed ] [ Google Scholar ] 6. Laranjo L, Lanas F, Sun MC, et al. World Heart Federation Roadmap for Secondary Prevention of Cardiovascular Disease: 2023 Update. Glob Heart. 2024;19:8. doi: 10.5334/gh.1278. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2018;71:e127–248. doi: 10.1016/j.jacc.2017.11.006. [ DOI ] [ PubMed ] [ Google Scholar ] 8. Williams B, Mancia G, Spiering W, et al. 2018 ESC/ESH Guidelines for the management of arterial hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology (ESC) and the European Society of Hypertension (ESH) Eur Heart J. 2018;39:3021–104. doi: 10.1097/HJH.0000000000001961. [ DOI ] [ PubMed ] [ Google Scholar ] 9. Virani SS, Alonso A, Aparicio HJ, et al. Heart Disease and Stroke Statistics-2021 Update: A Report From the American Heart Association. Circulation. 2021;143:e254–743. doi: 10.1161/CIR.0000000000000950. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. McEvoy JW, McCarthy CP, Bruno RM, et al. 2024 ESC Guidelines for the management of elevated blood pressure and hypertension. Eur Heart J. 2024;45:3912–4018. doi: 10.1093/eurheartj/ehae178. [ DOI ] [ PubMed ] [ Google Scholar ] 11. Shahaj O, Denneny D, Schwappach A, et al. Supporting self-management for people with hypertension: a meta-review of quantitative and qualitative systematic reviews. J Hypertens (Los Angel) 2019;37:264–79. doi: 10.1097/HJH.0000000000001867. [ DOI ] [ PubMed ] [ Google Scholar ] 12. Siopis G, Moschonis G, Eweka E, et al. Effectiveness, reach, uptake, and feasibility of digital health interventions for adults with hypertension: a systematic review and meta-analysis of randomised controlled trials. Lancet Digit Health. 2023;5:e144–59. doi: 10.1016/S2589-7500(23)00002-X. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Morawski K, Ghazinouri R, Krumme A, et al. Association of a Smartphone Application With Medication Adherence and Blood Pressure Control: The MedISAFE-BP Randomized Clinical Trial. JAMA Intern Med. 2018;178:802–9. doi: 10.1001/jamainternmed.2018.0447. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Dorsch MP, Cornellier ML, Poggi AD, et al. Effects of a Novel Contextual Just-In-Time Mobile App Intervention (LowSalt4Life) on Sodium Intake in Adults With Hypertension: Pilot Randomized Controlled Trial. JMIR Mhealth Uhealth. 2020;8:e16696. doi: 10.2196/16696. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Pletcher MJ, Fontil V, Modrow MF, et al. Effectiveness of Standard vs Enhanced Self-measurement of Blood Pressure Paired With a Connected Smartphone Application: A Randomized Clinical Trial. JAMA Intern Med. 2022;182:1025–34. doi: 10.1001/jamainternmed.2022.3355. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Kario K, Nomura A, Harada N, et al. Efficacy of a digital therapeutics system in the management of essential hypertension: the HERB-DH1 pivotal trial. Eur Heart J. 2021;42:4111–22. doi: 10.1093/eurheartj/ehab559. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Persell SD, Peprah YA, Lipiszko D, et al. Effect of Home Blood Pressure Monitoring via a Smartphone Hypertension Coaching Application or Tracking Application on Adults With Uncontrolled Hypertension: A Randomized Clinical Trial. JAMA Netw Open. 2020;3:e200255. doi: 10.1001/jamanetworkopen.2020.0255. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Lee P, Bubeck S, Petro J. Benefits, Limits, and Risks of GPT-4 as an AI Chatbot for Medicine. N Engl J Med. 2023;388:1233–9. doi: 10.1056/NEJMsr2214184. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25:44–56. doi: 10.1038/s41591-018-0300-7. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Muse ED, Topol EJ. Transforming the cardiometabolic disease landscape: Multimodal AI-powered approaches in prevention and management. Cell Metab. 2024;36:670–83. doi: 10.1016/j.cmet.2024.02.002. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Cruz Rivera S, Liu X, Chan A-W, et al. Guidelines for clinical trial protocols for interventions involving artificial intelligence: the SPIRIT-AI extension. The Lancet Digital Health . 2020;2:e549–60. doi: 10.1016/S2589-7500(20)30219-3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Agarwal S, LeFevre AE, Lee J, et al. Guidelines for reporting of health interventions using mobile phones: mobile health (mHealth) evidence reporting and assessment (mERA) checklist. BMJ. 2016;352:i1174. doi: 10.1136/bmj.i1174. [ DOI ] [ PubMed ] [ Google Scholar ] 23. Deloitte Access Economics; 2019. Mobile nation 2019: the 5G future. [ Google Scholar ] 24. O’Hagan ET, Marschner SL, Mishra S, et al. Self-Guided Blood Pressure Screening in the Community: Opportunities, and Challenges. Hypertension. 2024;81:2559–68. doi: 10.1161/HYPERTENSIONAHA.124.23283. [ DOI ] [ PubMed ] [ Google Scholar ] 25. Michie S, Richardson M, Johnston M, et al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med. 2013;46:81–95. doi: 10.1007/s12160-013-9486-6. [ DOI ] [ PubMed ] [ Google Scholar ] 26. Michie S, van Stralen MM, West R. The behaviour change wheel: a new method for characterising and designing behaviour change interventions. Implement Sci. 2011;6:42. doi: 10.1186/1748-5908-6-42. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Topouchian J, Zelveian P, Hakobyan Z, et al. Accuracy of the Withings BPM Connect Device for Self-Blood Pressure Measurements in General Population - Validation According to the Association for the Advancement of Medical Instrumentation/European Society of Hypertension/International Organization for Standardization Universal Standard. Vasc Health Risk Manag. 2022;18:191–200. doi: 10.2147/VHRM.S350006. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Stergiou GS, O’Brien E, Myers M, et al. STRIDE BP: an international initiative for accurate blood pressure measurement. J Hypertens (Los Angel) 2020;38:395–9. doi: 10.1097/HJH.0000000000002289. [ DOI ] [ PubMed ] [ Google Scholar ] 29. Foster ME, White M. Avoiding repetition in generated text. Proceedings of the Eleventh European Workshop on Natural Language Generation; Germany: Association for Computational Linguistics. 2007. pp. 33–40. [ Google Scholar ] 30. Coleman M, Liau TL. A computer readability formula designed for machine scoring. J Appl Psychol. 1975;60:283–4. doi: 10.1037/h0076540. [ DOI ] [ Google Scholar ] 31. Gao B, Pavel L. On the properties of the softmax function with application in game theory and reinforcement learning. arXiv. 2017 [ Google Scholar ] 32. Klimis H, Nothman J, Lu D, et al. Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content Analysis. JMIR Mhealth Uhealth. 2021;9:e27779. doi: 10.2196/27779. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Levenshtein VI. Binary Codes Capable of Correcting Deletions, Insertions and Reversals. Sov Phys Dokl. 1966;10:707. [ Google Scholar ] 34. Gomaa WH, Fahmy AA. A Survey of Text Similarity Approaches. IJCA. 2013;68:13–8. doi: 10.5120/11638-7118. [ DOI ] [ Google Scholar ] 35. Chow CK, Redfern J, Hillis GS, et al. Effect of Lifestyle-Focused Text Messaging on Risk Factor Modification in Patients With Coronary Heart Disease: A Randomized Clinical Trial. JAMA. 2015;314:1255–63. doi: 10.1001/jama.2015.10945. [ DOI ] [ PubMed ] [ Google Scholar ] 36. Klimis H, Thiagalingam A, McIntyre D, et al. Text messages for primary prevention of cardiovascular disease: The TextMe2 randomized clinical trial. Am Heart J. 2021;242:33–44. doi: 10.1016/j.ahj.2021.08.009. [ DOI ] [ PubMed ] [ Google Scholar ] 37. Haun J, Luther S, Dodd V, et al. Measurement variation across health literacy assessments: implications for assessment selection in research and practice. J Health Commun. 2012;17 Suppl 3:141–59. doi: 10.1080/10810730.2012.712615. [ DOI ] [ PubMed ] [ Google Scholar ] 38. Skivington K, Matthews L, Simpson SA, et al. A new framework for developing and evaluating complex interventions: update of Medical Research Council guidance. BMJ. 2021;374:n2061. doi: 10.1136/bmj.n2061. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Guest G, Bunce A, Johnson L. How many interviews are enough? An experiment with data saturation and variability. Field methods. 2006;18:59–82. doi: 10.1177/1525822X05279903. [ DOI ] [ Google Scholar ] 40. Creswell JW, Poth CN. Sage Publications; 2016. Qualitative inquiry and research design: choosing among five approaches. [ Google Scholar ] 41. Kuzel A. In: Doing qualitative research. 2nd. Crabtree BF, Miller WL, editors. Thousand Oaks, CA: Sage Publications; 1999. Sampling in qualitative inquiry. edn. [ Google Scholar ] 42. Morse J. In: Handbook for qualitative research. Denzin N, Lincoln Y, editors. Thousand Oaks, CA: Sage; 1994. Designing funded qualitative research; pp. 220–35. [ Google Scholar ] 43. Ancker JS, Benda NC, Reddy M, et al. Guidance for publishing qualitative research in informatics. J Am Med Inform Assoc. 2021;28:2743–8. doi: 10.1093/jamia/ocab195. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. 2006;3:77–101. doi: 10.1191/1478088706qp063oa. [ DOI ] [ Google Scholar ] 45. Saldaña J. The coding manual for qualitative researchers. Sage; 2021. [ Google Scholar ] 46. Braun V, Clarke V. To saturate or not to saturate? Questioning data saturation as a useful concept for thematic analysis and sample-size rationales. Qualitative Research in Sport, Exercise and Health. 2021;13:201–16. doi: 10.1080/2159676X.2019.1704846. [ DOI ] [ Google Scholar ] 47. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377–81. doi: 10.1016/j.jbi.2008.08.010. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. White IR, Horton NJ, Carpenter J, et al. Strategy for intention to treat analysis in randomised trials with missing outcome data. BMJ. 2011;342:d40. doi: 10.1136/bmj.d40. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 49. Willig C, Stainton Rogers W. The SAGE handbook of qualitative research in psychology. London: SAGE Publications; 2017. [ Google Scholar ] 50. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. Int J Qual Health Care. 2007;19:349–57. doi: 10.1093/intqhc/mzm042. [ DOI ] [ PubMed ] [ Google Scholar ] 51. Rahimi K, Bidel Z, Nazarzadeh M, et al. Pharmacological blood pressure lowering for primary and secondary prevention of cardiovascular disease across different levels of blood pressure: an individual participant-level data meta-analysis. Lancet. 2021;397:1625–36. doi: 10.1016/S0140-6736(21)00590-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 52. Cook A, Roberts K, O’Leary F, et al. Comparison of single questions and brief questionnaire with longer validated food frequency questionnaire to assess adequate fruit and vegetable intake. Nutrition. 2015;31:941–7. doi: 10.1016/j.nut.2015.01.006. [ DOI ] [ PubMed ] [ Google Scholar ] 53. Schapira MM, Fletcher KE, Hayes A, et al. The Development and Validation of the Hypertension Evaluation of Lifestyle and Management Knowledge Scale. J Clin Hypertens (Greenwich) 2012;14:461–6. doi: 10.1111/j.1751-7176.2012.00619.x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials online supplemental file 1 openhrt-13-1-s001.pdf (1.3MB, pdf) DOI: 10.1136/openhrt-2025-003819 Articles from Open Heart are provided here courtesy of BMJ Publishing Group ACTIONS View on publisher site PDF (1.8 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top