Problem: Aggressive titration of guideline-directed medical therapy is the cornerstone of heart failure management to reduce heart failure related hospitalization and mortality. A major hurdle to uptitration of medical therapy is the safety concerns of these medications namely hypotension and bradycardia. Early and frequent clinic visits will be required to ensure the safe uptitration of these treatments, but this is difficult to implement in daily practice due to resources constrain. Advancement in wearable and wireless technologies enables remote and continuous monitoring of patients in ambulatory setting and potentially achieving safe and effective uptitration of medical therapy while minimizing the demand on the current healthcare system. However, access to these wearable and wireless technologies is not universal and patients in low socio-economic tiers might be deprived of the opportunity to realize the benefits from these technologies.

Solution: This project therefore aims to develop a mobile platform and to provide wearable devices to patients with acute heart failure to improve and optimize the use of guideline-directed medical therapy in these patients.

Impacts: The finding from this project will inform future heart failure management and help to implement and integrate telemedicine in the vulnerable stage of heart failure care. It will help to maximize healthcare resources utilization and help to bridge the gap between those who have limited access to healthcare services. The model of heart failure care generated from this project can be propagate and replicated in general patients care.

Optimal and aggressive titration of guideline-directed medical therapy (GDMT) is the cornerstone of heart failure (HF) management to reduce HF related hospitalization and improve patient survival1. However, the percentage of patient achieving targeted dose of GDMT in real-world setting is consistently low across different regions2,3,4. Failure to achieve targeted dose of GDMT has been a main factor contributing to the high mortality of HF patients1. A major hurdle to up-titration of GDMT is the therapeutic inertia5 on the part of the treatment providers partly due to safety concerns for these medication namely symptomatic hypotension and bradycardia. The safety concern is especially great in patients with acute decompensated HF in whom fluctuation in hemodynamic can be substantial. Therefore, the major challenge in HF care is the attainment of optimal dose of GDMT without increasing serious adverse effect to the patients. Two approaches to patient care have been proposed to overcome this challenge. Patients can be either kept inpatient for medication up-titration5,6 or followed up by intensive outpatient clinic visits. Recently the STRONG-HF study has demonstrated the feasibility of rapid up-titration of GDMT through frequent outpatient clinic visits7. However, neither of these approaches is feasible in the local setting given the severe hospital bed and healthcare manpower shortage in Hong Kong public healthcare system. Remote monitoring (RM) has been used in chronic HF patients with aims to reduce clinic visits and improve HF care but has not been demonstrated uniformly beneficial in various studies9,10,11. Large part of the lack of benefit of RM in these clinical trials might be attributed to poor adherence to devices usage in the long term (50% of adherence) and a lack of a clearly defined actionable plan to the data received. Since then, significant advancement has been made in wireless and wearable technologies. Continuous monitoring and automatic transmission of patient hemodynamic data may mitigate the non-adherence issues of RM. As only a finite period of medication titration is generally required to achieve maximum tolerated dose of GDMT in HF patients, RM potentially may play a role in facilitating rapid medication up-titration and maximizing healthcare resources utilization during the initial phase of acute HF care.

Aims and Hypotheses to be Tested:

Study question: Can ambulatory telemonitoring improve and optimize guideline-directed medical therapy (GDMT) in patients with acute heart failure?

Study hypothesis: Telemonitoring can yield more efficient up-titration of GDMT to targeted dose in patients with acute heart failure than routine physical clinic visit.

Study aims:

1. to compare the percentage of targeted GDMT dose attainment in patients subjected to intensive medication up-titration guided by integrated telemonitoring against patients in routine medical care.
2. To evaluate the feasibility and safety of aggressive GDMT titration as guided by telemonitoring.
3. To evaluate the adherence and acceptance of modern commercially available wireless telemonitoring devices in HF patients.
4. To evaluate the impact of wearable telemonitoring on the quality of life of HF patients
5. To assess the role of different parameters and components from telemonitoring in predicting HF decompensation and adverse events from medications.

Plan of Investigation:

This is a prospective, open-labelled, single-centre, randomized controlled trial comparing telemonitoring (intervention) in patients with acute heart failure (HF) with routine clinical care (control). Eligible patient will be recruited primarily from Prince of Wales Hospital by both the Cardiologists and Internal Medicine physicians.

(i) Subjects (with justification on the sample size)

The target patients are those aged >18, who are hospitalized for first episode of acute decompensate heart failure, which is defined by all of the following criteria:

1. Clinical symptoms of Dyspnoea (exertional or at rest) and 1 of the following signs:

   • Congestion on chest X-ray
   • Rales on chest auscultation
   • Clinically relevant oedema (e.g. ≥1+ on a 0 to 3+ scale)
   • Elevated jugular venous pressure.

2. NT-proBNP which is measured within the same hospitalization with level >300ng/L or 600ng/L if ongoing atrial fibrillation/atrial flutter.

3. Left ventricular ejection fraction less than 40% by any form of imaging.

Exclusion criteria:

1. Documented history of HF with previous HF admission.

2. Current hospitalization for acute HF primarily triggered by pulmonary embolism, cerebrovascular accident, or acute MI.

3. Cardiac interventions (listed below) in past 30 days prior or planned during the study period.

   1. Major cardiac surgery, or Tran Aortic Valve Implantation, or Percutaneous Coronary Intervention, or MitraClip
   2. Current or expected heart transplant, LVAD, IABP, or patients with planned inotropic support in an outpatient setting.

4. Haemodynamically severe uncorrected primary cardiac valvular disease.

5. eGFR <30 mL/min/1.73m2 as measured during index hospitalization (latest measurement before randomization) or patients requiring dialysis.

6. Documented allergy or intolerance to beta-blocker, renin-angiotensin-aldosterone system inhibitor, mineralocortical receptor antagonist, or angiotensin receptor/neprilysin inhibitor.

7. Pregnant or nursing women.

8. Psychiatric or severe neurological disorder, cirrhosis, or active malignancy leading to a life expectancy < 6 months.

9. Non-ambulatory patients.

10. Patients who are unable to operate simple electronic devices; or unable to assess a mobile network service in the place of residence.

11. Patients who are unable to given written informed consent.

Sample size calculation:

From previous HF registry conducted at local sites and from previous published data from the same locality, the percentage of GDMT use in HF patients under routine care is around 20%. Assuming the percentage of targeted dose GDMT attainment to be 50% in the intervention arm7, an estimate of 56 patients with 28 in each arm is needed to show a significant difference between two groups, with a 5% significance level and 80% power. Assuming the drop-out rate to be 30%, a total of 80 patients with 40 patients in each arm will be required.

(ii) Methods The study will be conducted according to the Declaration of Helsinki and comply with ICH-GCP. Institutional Ethics approval will be obtained, and the trial will be registered at Clinicaltrials.gov before the start of this study. Subjects will be recruited from Prince of Wales Hospital medical inpatient units. Written informed consent will be taken from each recruited subject. Patients meeting eligibility criteria will be randomized via a freely accessible interactive web response system (www.randomization.com) to one of the two study arms in a 1:1 ratio, according to a central randomization scheme.

After recruitment, baseline information including patient's demographics, past medical history, discharge medication history, baseline laboratory, and baseline echocardiography will be collected.

As a background CHF treatment, one of the Sodium-glucose Cotransporter 2 (SGLT2) Inhibitor either Empagliflozin 10mg daily or Dapagliflozin 10mg daily will be started before discharge as this class of drug does not require any dose up-titration12,13. Other classes of HF drug such as beta-blocker (BB), mineralocortical receptor antagonist (MRA), or angiotensin receptor/neprilysin inhibitor (ARNI) will be initiated and titrated by physicians during the index admission according to their clinical judgement before discharge.

(iii) Study design This is a prospective, open-labelled, single-centre, randomized controlled trial consisting of an intervention group and a control group. A physical clinic visit with cardiologist consultation will be scheduled for both groups every 30 days with blood testing for NT-proBNP and renal function until 90 days. At 90 days, a follow-up echocardiogram will be performed for both groups of patients with the primary aim of left ventricular function evaluation. The follow-up period for both groups is 90 days. The proposed study duration is 24 months.

Intervention group Patients randomized to the intervention group will be taught how to operate individual devices of telemonitoring before discharge by the research team. 30 days of all three classes medication up to the targeted dose (as listed below) will be given to the patient with clear labelling and instructions on their administration. Pills of minimal dosage will be provided for easier patient communication and dosage titration. Medication up-titration will be performed over the phone by a senior research nurse according to algorithms as listed below. Patients will also be asked during each phone call regarding the occurrence of any adverse event.

The integrated telemonitoring consists of 1. a wearable wrist heart rate and activity monitor, 2.a blue-tooth enabled body weight scale with bioimpedance analysis, and 3. a blue-tooth enabled electronic blood pressure device.

1. Wearable wrist heart rate and activity monitor: A commercially available smart watch capable of continuous monitoring of heart rate and activities including daily steps count and daily activity level, detection of arrythmia using a proprietary FDA approved algorithm14, and measuring peripheral capillary oxygen saturation will be provided to the patient. Patients will be encouraged to wear the watch as often as possible during daytime and during sleep. Data gathered will be automatically transmitted wirelessly to a cloud-based platform provided by the manufacturer accessible only by the research team and the patient.
2. Bluetooth enabled body weight scale with bioimpedance analysis: An electronic scale which measures bioimpedance using bioelectrical analysis from foot-to-foot impedance and body fluid percentage (BF%) using a proprietary equation15,16 will be provided to the patients. Patient will be encouraged to measure their bodyweight (BW) and bioimpedance twice per day, one in the morning, and one before bath, with minimal clothing. Both data will be automatically transmitted wirelessly to a cloud-based platform provided by the manufacturer accessible only by the research team and the patients.
3. Bluetooth enabled electronic blood pressure device: An electronic cuffed arm blood pressure monitor will be provided to the patient with instruction on how to perform proper blood pressure measurement at home16. Patients are encouraged to take at least 2 blood pressure reading daily, one in the morning and one in the evening. Blood pressure reading will be transmitted wirelessly to an online platform provided by the manufacturer accessible only by the research team and the patients.

Uploaded hemodynamic and biological data will be reviewed by the research team on a daily basis.

Medication up-titration scheme

Up-titration of medication will be performed over the phone every 3 days by a senior research nurse until the maximum tolerated dose is reached. In patient in whom any 1 of the 3 classes of GDMT has not been started upon discharge, the starting dose of the missing class of GDMT will be started first. In patient in whom none of the GDMT has been started upon discharge, the medication will be started with the starting dose as listed in the sequence of MRA -> BB -> ARNI. In patient in whom all 3 classes of GDMT has been started upon discharge, the sequence of medication titration is as followed: MRA -> BB -> ARNI, with aim of up-titration to the maximum dose of 1 class followed by the next class. This approach has been evaluated in a simulated model to be the most effective in reducing HF related hospitalization18. The maximum dose is defined by the suggested target dose1 (as listed below) or the maximum tolerated dose without side effect, whichever is higher. The next escalating dose of any class will not be initiated if systolic blood pressure is less than 90mmHg. Additionally, the next escalating dose of ARNI or MRA will not be initiated if weekly blood test showed serum potassium greater than 5.5mmol/L, or if estimated GFR <30ml/min/1.73m2, or if serum creatinine level rise above 30% over baseline. Also, the next escalating dose of BB will not be initiated if systolic blood pressure is less than 90mmHg or if the lowest heart rate on the 24 hours monitor is less than 40 beat/min or the average heart rate is less than 55 beat/min. If any severe adverse event is encountered after up-titration of medication dose, patient will be instructed to fall back on the previous dose. If the adverse event resolved on the next up-titration schedule, patient will be instructed to up-titrate the previous drug dose again. If similar adverse event arises again, then the previous dose will be considered the maximum tolerated dose. The suggested medications starting and target dose for each individual class of GDMT are as followed1:

1. Mineralocortical receptor antagonist - Eplerenone starting dose 25mg daily with 25mg daily increment until target dose of 50mg daily; Spironolactone starting dose 25mg daily with 25mg daily increment until target dose of 50mg daily.
2. Beta-blocker: Bisoprolol starting dose 1.25mg daily with 1.25mg daily increment until target dose of 10mg daily; Carvedilol starting dose 3.125mg twice daily with 3.125mg twice daily increment until target dose of 25mg twice daily; Metoprolol succinate starting dose 25mg daily with 25mg daily increment until target dose of 200mg daily.
3. Angiotensin receptor/neprilysin inhibitor - Sacubitril/valsartan starting dose 24/26 mg twice daily with 24/26 mg twice daily increment until target dose of 97/103 mg twice daily.

Blood checking for renal function, electrolytes and Point-of-care NT-proBNP will be performed weekly during the medication up-titration period for safety reason until the maximum dose is reached.

Control Group Patient will be seen at clinic for physician consultation every 30 days after discharge as per current practice until the end of follow-up which is at 90 days. Guidance on medication up-titration similar to intervention group will be provided to the physician. Medication titration will be performed during clinic consultation at the physician's discretion.

Outcomes Primary outcomes: Percentage of patients receiving target dose of individual class of GDMT which is defined as above1 at 90 days.

Secondary outcomes:

• Wearable device and monitoring adherence in the intervention group.
• Difference in New York Heart Association (NYHA) functional class, 6 minutes' walk test, heart failure related quality of life measure (by EQ-5D questionnaire) at 90 days.
• Difference in NT-proBNP level at 90 days.
• Unscheduled medical visit.
• HF related Hospitalization.
• HF related mortality and all-cause mortality. Safety outcomes: Occurrence of any Serious Adverse Event which is defined as
• fatal or life-threatening events,
• resulting in persistent or significant disability/incapacity
• requiring inpatient hospitalization or prolongation of existing hospitalization,
• being medically significant, i.e. defined as an event that jeopardizes the patient or may require medical or surgical intervention to prevent one of the outcomes listed above.

Data processing and analysis All data were reported as mean ± SD unless otherwise specified. Following the intent-to-treat principle, patients will be analyzed according to the arm to which they were assigned. The primary outcome, difference in the percentage of patients receiving targe dose of individual class of GDMT at 90 days will be compared between the two groups using unpaired Student t-test. Other parameters and outcomes will be compared between baseline and at 90 days, as well as between the 2 groups. Student's t test and Mann-Whitney U test will be used for continuous variables while Chi-squared test and Fisher's exact test will be used to compare categorical variables, where appropriate. Two groups will be compared with respect to the change in NT-proBNP level at baseline and at 90-days using ANCOVA with the baseline value and randomization stratification factors as covariates. Central laboratory measures of NT-proBNP will be log-transformed for this analysis. All tests were two-sided, with significance set at 5%.

Additionally, a cost-effective analysis will be performed on the total cost incurred in both groups and incremental cost-effectiveness ratio expressed as incremental cost per hospitalization avoided at day 0, day 15, day 30, and day 90, based on data from the stimulated model18 will be calculated. All cost data will be expressed in Hong Kong dollar and valued on the starting date of the study from a healthcare system perspective (i.e., all the medical costs will be valued on the basis of non-subsidized cost). The average cost difference between the intervention and control groups will be used to derive the incremental total cost (or total cost saved). All statistical analyses will be performed using STATA version 15 software (College Station, TX, USA).

Inclusion Criteria:

1. Clinical symptoms of Dyspnoea (exertional or at rest) and 1 of the following signs:

   • Congestion on chest X-ray
   • Rales on chest auscultation
   • Clinically relevant oedema (e.g. ≥1+ on a 0 to 3+ scale)
   • Elevated jugular venous pressure.

2. NT-proBNP which is measured within the same hospitalization with level >300ng/L or 600ng/L if ongoing atrial fibrillation/atrial flutter.

3. Left ventricular ejection fraction less than 40% by any form of imaging.

Exclusion Criteria:

1. Documented history of HF with previous HF admission.

2. Current hospitalization for acute HF primarily triggered by pulmonary embolism, cerebrovascular accident, or acute MI.

3. Cardiac interventions (listed below) in past 30 days prior or planned during the study period.

   1. Major cardiac surgery, or Tran Aortic Valve Implantation, or Percutaneous Coronary Intervention, or MitraClip
   2. Current or expected heart transplant, LVAD, IABP, or patients with planned inotropic support in an outpatient setting.

4. Haemodynamically severe uncorrected primary cardiac valvular disease.

5. eGFR <30 mL/min/1.73m2 as measured during index hospitalization (latest measurement before randomization) or patients requiring dialysis.

6. Documented allergy or intolerance to beta-blocker, renin-angiotensin-aldosterone system inhibitor, mineralocortical receptor antagonist, or angiotensin receptor/neprilysin inhibitor.

7. Pregnant or nursing women.

8. Psychiatric or severe neurological disorder, cirrhosis, or active malignancy leading to a life expectancy < 6 months.

9. Non-ambulatory patients.

10. Patients who are unable to operate simple electronic devices; or unable to assess a mobile network service in the place of residence.

11. Patients who are unable to given written informed consent.