New Taipei City 888 Digital Medical AI Platform for Prevention and Care of the Three Highs and Cardio-Renal-Vascular Diseases

New Taipei City 888 Digital Medical AI Platform for the Prevention and Treatment of the Three Highs (Hypertension, Hyperglycemia, Hyperlipidemia) and Cardio-Renal-Vascular Diseases: Validation and Implementation Through Large-Scale Cluster Randomized Clinical Trials: T-888-DIGICARE-DREAM

Non-communicable diseases (NCDs), or chronic diseases, are a major public health burden globally and in Taiwan, and control of the "three highs" (hypertension, dyslipidemia, and hyperglycemia) is a national priority. Nearly half of the 10 leading causes of death in Taiwan are directly or indirectly related to atherosclerotic cardiovascular disease (ASCVD), for which hypertension, dyslipidemia, and abnormal blood glucose are the major risk factors. National health insurance data indicate that over 70% of middle-aged and older adults have at least one of these chronic conditions. Early stages are often asymptomatic, and inadequate control may lead to complications of cardiovascular disease, stroke, renal vascular disease, and retinopathy, causing irreversible organ damage and death.

For blood pressure, large clinical trials such as SPRINT and STEP have shown that targeting systolic blood pressure below 130 mmHg significantly reduces ASCVD events. For LDL cholesterol, "the lower, the better" applies, with guideline-recommended LDL-C targets determined by baseline ASCVD risk, with levels below 55 mg/dL for very high-risk patients. For diabetes, treatment goals generally include fasting plasma glucose below 130 mg/dL and glycated hemoglobin (HbA1c) below 7%.

Although antihypertensive therapy has been proven effective in preventing cardiovascular disease and chronic kidney disease attributable to hypertension, fewer than one-third of patients receiving antihypertensive medications achieve current guideline-recommended blood pressure targets. No randomized clinical trial to date has used home blood pressure as the primary therapeutic reference. Moreover, long-term evidence is lacking regarding the organ-protective effects of strategies targeting morning hypertension and of bedtime dosing regimens. Although the TIME study is currently the largest and longest-followed trial addressing dosing time, its bedtime-dosing arm was not specifically targeted to patients with morning hypertension. Therefore, we propose a clinical trial to investigate management strategies for morning hypertension. Aligned with the 2022 Taiwan Hypertension Guidelines, we will employ the "722 protocol" for home blood pressure monitoring and enroll patients with morning home blood pressure ≥130/80 mmHg to compare selective nocturnal administration of antihypertensive agents versus exclusive morning dosing, assessing differences in morning home blood pressure control rates, end-organ damage, and atherosclerotic cardiovascular disease (ASCVD) events.

This project will implement two large-scale cluster-randomized clinical trials within the New Taipei City healthcare network. The first trial (T-888-DIGICARE) will evaluate whether a mobile digital health platform augmented with interactive digital modules can more effectively achieve the "888" targets for prevention and treatment of the Three Highs (Hypertension, Hyperglycemia, Hyperlipidemia). The second trial (DREAM-G) will use the same mobile health delivery model to investigate whether the timing of antihypertensive medication administration (nocturnal versus morning dosing) differentially affects patients with poor morning home blood pressure control. Beyond generating rigorous evidence through a novel clinical approach, this program is expected to have substantial global clinical impact and to showcase Taiwan's healthcare capabilities internationally. Operational challenges encountered and solutions developed during the trials will also provide critical feasibility data for the concurrent real-world implementation registry (T-888-DIGICARE-Registry). The registry will run in parallel with the cluster trials and will enroll individuals who decline trial participation at baseline as well as participants after trial completion, thereby serving as a continuity and real-world evidence platform.

Participants identified through the 888 screening program with at least one abnormal value will be invited for group allocation. Home blood pressure measurement will be performed according to the "722 protocol." If morning home BP ≥130/80 mmHg (average home BP ≥130/80 mmHg or already receiving antihypertensive therapy), participants will be invited to join the randomized clinical trial DREAM-G, investigating optimal timing of antihypertensive medication for morning hypertension. All other participants will be invited to the randomized clinical trial T-888-DIGICARE, evaluating care intervention via a mobile smart healthcare platform. Those not enrolled will be invited to participate in the pragmatic digital health registry program T-888-DIGICARE-Registry.

Execution methods and procedures for these three large-scale clinical trials are as follows:

Randomized Clinical Trial of Mobile Smart Healthcare Platform Intervention (T-888-DIGICARE)

1. Randomization Unit: Participating primary care clinics. Each clinic must complete informed consent and enrollment preparation prior to participation.

2. Randomization Method: Computer-generated block randomization, stratified by clinic location. Stratification is based on the median household disposable income across the 29 administrative districts of New Taipei City, categorized into three socioeconomic tiers. Clinics are randomized within each tier to minimize confounding by socioeconomic status on health behaviors and healthcare accessibility, and to facilitate future policy expansion in lower-income regions.

   • Given the practical nature of digital platform implementation, blinding of physicians and patients is not feasible. Therefore, a single-blind design is adopted: investigators responsible for data aggregation and statistical analysis of primary and secondary endpoints remain blinded to clinic allocation, reducing analytic bias.

3. Intervention Measures:

   • Years 1-5 (Trial duration: 3 years enrollment, 2 years follow-up)
   • Intervention group: Comprehensive integration of the "888 Digital Management Platform" into participating clinics and patient mobile devices. Features include app-based interactive modules, personalized message push notifications, input of "three-high" (hypertension, hyperlipidemia, hyperglycemia) test results, and AI-assisted quantification of target organ damage.
   • Control group: Standard care provided by attending physicians according to routine practice. The digital platform is also installed but limited to home BP recording and automated analysis, without remote monitoring, interactive modules, personalized messaging, laboratory data input, AI-based organ quantification, or risk alerts. Consent forms specify that the control group will not receive enhanced digital interventions.

Pragmatic Digital Health Registry Program (T-888-DIGICARE-Registry)

1. Blood Pressure Intervention Logic (per 2022 Taiwan Hypertension Guidelines):

   • System prompts physicians to initiate/adjust antihypertensive therapy if average home BP >130/80 mmHg.
   • Untreated patients: initiate therapy.
   • Poorly controlled patients: increase dosage if average BP 130-139/80-89 mmHg, or add a new agent if >140/90 mmHg.
   • If isolated morning BP >130/80 mmHg: shift one antihypertensive agent to bedtime dosing.
   • If isolated bedtime BP >130/80 mmHg: switch short-acting agent to long-acting formulation.
   • Monthly reminders for patients to perform "722 protocol" home BP monitoring (≥4 days). Automated push notifications are sent if records are insufficient. Notifications are logged for case manager follow-up.

2. Lipid Intervention Logic (per 2022 Taiwan Dyslipidemia Guidelines):

   • LDL-C >100 mg/dL without complications: initiate moderate-intensity statin.
   • Diabetes with LDL-C >70 mg/dL: initiate high-intensity statin or maximally tolerated dose, ± non-statin therapy.
   • ASCVD with LDL-C >70 mg/dL: initiate high-intensity statin or maximally tolerated dose, ± non-statin therapy.

3. Glycemic Intervention Logic (per ADA Guidelines):

   • Abnormal HbA1c within 3 months: ≥6.5%, or insufficient improvement (e.g., >8% without ≥1% reduction, >10% without ≥2% reduction).
   • Treatment recommendations:
   • HbA1c <9% and drug-naïve: initiate metformin.
   • HbA1c ≥9%: initiate dual oral therapy (e.g., metformin + SGLT2i or DPP-4i).
   • If uncontrolled after 3 months: add 1-2 oral agents.
   • If still uncontrolled: introduce GLP-1 RA or basal insulin.
   • If already on injectables but uncontrolled: adjust dose or consider mixed insulin regimen.
   • With cardiovascular disease: prioritize SGLT2i or GLP-1 RA.

4. Lifestyle Intervention:

   • Weekly randomized delivery of short educational messages (CV disease knowledge, medication reminders, nutrition, physical activity).
   • Targeted reinforcement for unmet goals (e.g., medication adherence, exercise, diet).

5. Patient Interaction Mechanisms:

   • 24-hour AI chatbot for health consultation and emergency alerts.
   • Integrated digital health records: patients can review longitudinal data on BP, lipids, glucose, renal function, and goal attainment trends.

Randomized Clinical Trial on Optimal Timing of Antihypertensive Medication for Morning Hypertension (DREAM-G)

• Patients perform home BP monitoring per "722 protocol" (≥4 morning and ≥4 evening readings within 7 days) prior to clinic visit.
• Hypertensive patients with morning BP ≥130/80 mmHg (average home BP ≥130/80 mmHg or on antihypertensive therapy) are invited.
• Trial sites include community clinics. Participants are randomized 1:1 to intervention (bedtime dosing) or control (morning dosing).
• Physicians prescribe antihypertensive regimens (CCB, ARB/ACEI, or combinations) aiming to control morning and average BP <130/80 mmHg.
• Objective: Both morning and average blood pressure <130/80 mmHg (722 protocol: at least 4 measurements of morning and average BP within 7 consecutive days)
• Intervention group: stepwise bedtime dosing strategy: ACEI/ARB → CCB → fixed-dose ACEI(ARB)/CCB combination → diuretic → α-blocker → β-blocker → mineralocorticoid receptor antagonist (MRA).
• Control group: morning dosing adjustments.

Adjustment principles:

• If average home BP ≥130/80 mmHg and morning BP ≥130/80 mmHg: both groups escalate therapy.
• If average home BP <130/80 mmHg but morning BP ≥130/80 mmHg: control group maintains regimen; intervention group shifts one agent to bedtime dosing.

Principles of Antihypertensive Medication Adjustment:

Treated HTN Untreated HTN Average Home BP Monitoring (HBPM) HBPM < 130/80 mmHg HBPM ≥ 130/80 mmHg HBPM ≥ 130/80 mmHg Morning Home BP Monitoring (HBPM) HBPM ≥ 130/80 mmHg HBPM ≥ 130/80 mmHg HBPM ≥ 130/80 mmHg Intervention Group Shift one antihypertensive agent from daytime to bedtime dosing Initiate a new antihypertensive agent at bedtime, or adjust the dosage of existing medication Initiate a new antihypertensive agent at bedtime Control Group No dosage adjustment required Initiate a new antihypertensive agent during the daytime, or adjust the dosage of existing medication Initiate a new antihypertensive agent during the daytime

Clinical Assessments:

• Monthly home BP monitoring per "722 protocol," recorded every 2 months until study completion.
• Laboratory and imaging assessments at baseline, and months 2, 4, 6, 12, 14, 16, 20, and 24: complete blood count, ALT, lipid profile, fasting glucose/HbA1c, BUN/creatinine, electrolytes, NT-proBNP, uric acid, UACR, fundus photography, ECG, and BP measurements.
• Detailed schedule provided in trial flowchart (see p.7).

Inclusion Criteria:

I.T-888-DIGICARE：

Eligible participants must meet the definition of hypertension and at least one of the following chronic (non-hypertensive) conditions:

1. Hypertension

   • With or without treatment, and meeting one of the following:
   • Office blood pressure: two consecutive measurements >130/80 mmHg
   • Home blood pressure: weekly average (morning/evening) >130/80 mmHg, or more than half of the weekly measurements exceeding this threshold

2. Hyperlipidemia

   • With or without treatment, and
   • LDL-C >100 mg/dL (or >70 mg/dL in patients with diabetes or established ASCVD)

3. Diabetes Mellitus

   • With or without treatment, and
   • HbA1c >6.5%

4. Chronic Kidney Disease (CKD)

   • Estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m², and
   • Urine albumin-to-creatinine ratio (UACR) >30 mg/g

5. Atherosclerotic Cardiovascular Disease (ASCVD):

   • Including coronary artery disease, cerebrovascular disease, peripheral arterial disease, or aortic pathology

II. DREAM-G:

1. Hypertension

   • With or without treatment, and meeting one of the following:
   • Office blood pressure: two consecutive measurements >130/80 mmHg
   • Home blood pressure: weekly average (morning/evening) >130/80 mmHg, or more than half of weekly measurements exceeding this threshold

2. Hyperlipidemia

   • With or without treatment, and
   • LDL-C >100 mg/dL (or >70 mg/dL in patients with diabetes or established ASCVD)

3. Diabetes Mellitus

   • With or without treatment, and
   • HbA1c >6.5%

4. Chronic Kidney Disease (CKD)

   • Estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m², and
   • Measurable urinary microalbumin-to-creatinine ratio

5. Atherosclerotic Cardiovascular Disease (ASCVD):

   • Including coronary artery disease, cerebrovascular disease, peripheral arterial disease, or aortic pathology

   Participants must meet at least one of the above chronic disease conditions and:

6. Untreated Hypertension:

   • Morning home blood pressure (HBP) >130/80 mmHg: weekly average (≥4 days) >130/80 mmHg,
   • Or more than half of morning HBP measurements (≥4 days per week) >130/80 mmHg

7. Treated Hypertension:

   • Morning home blood pressure >130/80 mmHg: weekly average (≥4 days) >130/80 mmHg,
   • Or more than half of morning HBP measurements (≥4 days per week) >130/80 mmHg
   • Average home blood pressure ≤130/80 mmHg (used to determine treatment strategy)

Exclusion Criteria:

I. T-888-DIGICARE

1. Symptomatic heart failure (New York Heart Association functional class II-IV)
2. End-stage renal disease requiring long-term dialysis
3. Pregnant women or those planning pregnancy

II. DREAM-G:

1. Life expectancy <1 year
2. End-stage renal disease (ESRD) requiring regular renal replacement therapy, including dialysis, kidney transplantation, or palliative care
3. Severe liver cirrhosis
4. Malignancy under active treatment