Heart failure is a common and serious disease responsible for significant healthcare costs and the need of hospitalizastions. The course of the disease is characterized by periods of progressive deterioration with repeated hospital admissions, especially in the final stages of life. Telemedical self-monitoring is a promising alternative for remote monitoring that can provide individualized treatment, smooth titration of medications and reduce hospital stays. However, the evidence for its benefits is limited, which requires further research.

Our hypotheses are that self-monitoring in heart failure can:

1. Reduce avoidable inpatient care and mortality.
2. Optimize the escalation of medications to optimal medical therapy.
3. Increase self-care and security.
4. Improve the prediction of deterioration in heart failure. Work Plan: We will compare six months of telemedical monitoring with standard care, and integrate telemedical data with electronic health records (EHR) for analysis and development of prognostic models for clinical outcomes (data collection is ongoing). Consecutive heart failure patients (target 300) will receive digital equipment for reporting vital parameters, experiences, and symptoms over six months. Medication adjustments are made remotely, and physical visits as needed. Data on mortality, healthcare needs, and health economics will be collected over two years after the monitoring period. We plan to retrieve a matched control population from the Swedish heart failure registry (SwedeHF). Telemonitoring data and EHR will be analyzed with traditional regression models and machine learning for identifying predictive factors for i) death, ii) readmission for heart failure or other cardiovascular disease.

Significance: The study can contribute to more cost-effective, patient-centered, and medically purposeful care of heart failure.

Heart failure (HF) affects 2% of the population, rising to 10% in individuals over 80 years, causing considerable healthcare costs and suffering. In newly diagnosed HF titration of medication to optimal medical therapy (OMT) usually requires multiple physical visits in specialized clinics. Recent advances have enabled telemonitoring, wireless transmission of sensor-generated health data as an alternative which need scientific evaluation.

HF is characterized by periods of stability followed by episodic deterioration, with reduced quality of life (QoL) and hospitalizations. Decline in vital signs appear long before symptoms occur, successively leading to a point when hospitalization cannot be avoided. Telemonitoring allows for consistent monitoring of vital parameters, which potentially may guide treatment and prevent hospitalizations. It also promotes patient empowerment and personalized treatment. The European Society of Cardiology's (ESC) guidelines for HF, 2021 calls for further research on real world telemedicine data to optimize patient selection, equipment, and protocols. The Swedish national HF register, SwedeHF, contains detailed data on the majority of acute HF patients, providing an excellent opportunity to generate a well characterized matched real world population, including yearly up-dates of clinical information. In addition, electronic health records (EHRs) are an increasingly important source for studies of routinely collected healthcare data, offering granular real world data for research and prognostic modeling.

Work plan: Can RPM/telemonitoring for HF provide effective, safe and cost-efficient care with high patient satisfaction? We will compare six months of telemonitoring intervention with standard care to study clinical outcomes, patient experience and health economics (data collection in progress).

Participants: Consecutive HF patients (≥18 years) at specialized HF clinics are being recruited for intervention. Inclusion criteria are new or worsening HF in patients requiring monitoring due to a) titration of medication to OMT or b) decompensation requiring regular physical monitoring.

Intervention: Patients receive equipment for report of vital signs and patient reported experience measures (PREMS), automatically transmitted via smartphone/bluetooth for six months (possible extension to 12 months in case of instability). A digital smartphone application is used for registration and transmission of data, including modes for asynchronous (chat) communication with the healthcare provider. Planned titration to OMT or diuretic adjustments may be done remotely depending on clinical assessment. The digital sensors include an electronic scale, pulse and blood pressure monitor for daily measurements by the patient. The healthcare provider is notified of abnormal values, with services provided during office hours. Physical visits will be planned on a need-to basis. A care plan for target OMT or adjustments of diuretic treatment is established.

A matched control population (5:1, pseudonymized data) will be obtained from the SwedeHF registry (national heart failure registry).

Data at baseline and outcomes is collected from medical records, questionnaires, and national registers:

At baseline from medical records (risk factors, comorbidities, laboratory data, and clinical findings). Psychometric rating scales are administered weekly and on deterioration. General health status, QoL questionnaires (EQ5D), assessments of comfort with the technology, satisfaction with the healthcare provider and method, validated protocols for evaluation of self-efficacy (Swedish version of the general self-efficacy scale), and trust in the healthcare system (Sense of security in care) will be completed according to schedule. The questionnaires are distributed via the software solution or by 1177, and SwedeHF is updated with yearly clinical data. Final data will be collected after 2 years. Health care contacts will be measured by national registers: Inpatient Register, Outpatient Register, SwedeHF, and the Cause of Death Register. Clinical data is registered in a web-based case record form (eCRF) via REDCap, a secure web-based database management application, compiling variables corresponding to SwedeHF data, which includes variables not automatically identified from medical records such as: type of HF, date of first diagnosis, LVEF and NYHA at inclusion, etiology, medications, dosages at start of intervention and OMT, graded symptoms of tiredness, dyspnea och swelling and estimated overall QoL (1-100%). Economic data will be collected from administrative systems and managerial information from interviews with staff and management using predefined forms.

Outcomes until 2 years include: mortality, healthcare needs (re-hospitalizations, days hospitalized, out-patient visits), PREMs, time to and dosages at OMT. Health economic outcomes are costs (direct, indirect), efficiency (technical and allocative) and description of management systems. Outcome measures as previously outlined will be combined with EHR data including structured and unstructured clinical information and metadata tracing the trajectory of HF over time.

Analyses: Data will be analyzed using traditional regression models and machine learning techniques aiming to identify predictive factors for i) death, ii) rehospitalization for HF or other cardiovascular disease.

Clinical data from the intervention population will be matched (5:1) for age, sex, LVEF and duration of HF, to a pseudonymized population identified from SwedeHF. Data will be summarized and presented using traditional statistical methods, linear and logistic regression, and survival analysis. Analysis models for the main research questions will be adjusted for possible confounders.

For the telemedicine study, based on a calculation with 80% power, significance level 0.05, and an intervention effect of 20% (death and unplanned hospitalization) over 2 years of follow-up, a sample size of 300 is required. The effect measure is the composite "time to hospitalization due to heart failure or death." The expected frequency of events is based on outcomes in a heart failure database from 1987-2010 of 361,000 patients.

Consecutive HF patients (≥18 years) at specialized HF clinics are being recruited for intervention.

Inclusion criteria:

• Newly diagnosed heart failure requiring titration of medication
• Heart failure with decompensation: ≥1 hospitalization during the preceding 12 months
• Heart failure with decompensation: worsening between clinical controls

Exclusion criteria:

• Insufficient understanding of spoken or written Swedish, and no access to assistance
• Lacking cognitive and physical ability to handle equipment and make measurements
• Not motivated to participate
• Does not hold a national digital identification: Mobilt BankID
• Does not have smartphone
• No internet availability in home
• Not followed at the participating cardiology clinics
• Palliative care or planned care that will interfere with patient´s participation

Consecutive patients fulfilling the criteria above, treated in specialized heart failure clinics at Sahlgrenska University Hospital, Southern Älvsborg Hospital and Skaraborg Hospital.