The FAV is a non-invasive wearable monitor designed to spot-check transthoracic bioimpedance and send the measurement data to a remote telehealth center. The FAV system consists of a measurement vest, an electronics module, a mobile phone-based app, and a remote database (Figure 1). The measurement vest is a functional textile that is fitted snugly to the patient’s chest circumference using an adjustment strap and enables at-home self-measurements [13]. The inside of the vest has 4 textile electrodes arranged pair wise on 2 supporting pads located on either side of the rib cage, at the inferior part of the lungs. The electronics module connects to the back of the vest. The module determines transthoracic bioimpedance at multiple frequencies, which enables a model-based assessment of intra- and extra-cellular fluid status within the thorax [14] and the detection of respiration [15]. The module also records a 1-lead ECG as well as the patient’s motion and posture. The electronics module communicates wirelessly with a mobile phone-based app that guides the user through the measurement steps, controls the measurement parameters, and receives the measurement data. After each measurement, the app automatically transmits the data to a remote database hosted by the sponsor in a secure data center.

Self-assessments with the FAV are performed by following a simple measurement routine that takes 8-10 minutes (Figure 2).

The aim of the SENTINEL-HF study is to evaluate whether transthoracic bioimpedance assessed daily using the FAV system following discharge from an HF hospitalization can predict recurrent HF decompensation events. The study also aims to assess patient adherence to the FAV system and delineate adherence profiles based on novel patient demographic and psychosocial characteristics.

SENTINEL-HF is a prospective, non-randomized, observational study. Patients are enrolled during a baseline hospitalization for HF and are monitored for 45-days post-discharge by means of daily self-assessments with the FAV. Patient-reported data related to healthcare utilization and quality of life are collected through interviews conducted during the baseline hospitalization and throughout the home monitoring period, at days 7, 14, and 45. Passive surveillance of participants for clinical events is performed through medical record review for 90 days following discharge. The extended passive surveillance period enables the examination of bioimpedance trends in relation with long-term outcomes.

Approximately 180 patients are enrolled at 2 teaching hospitals that comprise the University of Massachusetts Memorial Medical Center (UMMMC), a large academic medical center in Central Massachusetts. The UMMMC serves a heterogeneous patient population with a diverse ethnic/racial, socioeconomic, urban/rural, and other socio-demographic background. Medicare 30-days readmission and death rates for patients discharged from an HF-related hospitalization at UMMMC are similar to the US national rates (23% and 12%, respectively) [16].

The SENTINEL-HF study enrolls patients in New York Heart Association (NYHA) functional classes II-IV who are hospitalized with a primary diagnosis of HF or with a sign (ie, vascular congestion on chest radiograph, rales on lung exam, or peripheral edema) and a symptom (ie, dyspnea or orthopnea) consistent with ADHF. Adult patients (≥21 years) are eligible to participate if they are willing and able to comply with the clinical investigation plan, in particular, able to handle the technical devices, and are available for follow-up visits throughout a 45-days post-discharge period. Exclusion criteria include an N-terminal pro-brain natriuretic peptide (NT-proBNP) level ≤100 pg/dL, end-stage chronic kidney disease requiring hemodialysis, chronic obstructive pulmonary disease (COPD), a primary diagnosis of gastrointestinal bleed, asthma, cardiac arrest or acute coronary syndrome at baseline, primary pulmonary hypertension, psychiatric or neurological disorders of moderate to severe degree, non-English speakers, pregnant women, prisoners, patients planning to move from their residence within 2 months from baseline, a body habitus that prevents the patient from fitting into a measurement vest or does not allow for adequate electrode-skin contact while wearing the vest, non-intact skin at the location of the electrodes, or the presence of an implantable pacemaker/defibrillator.

The ability of the FAV system to predict HF decompensation events is evaluated in relation to the composite endpoint of unplanned HF-related rehospitalization and HF worsening. A rehospitalization is defined as a presentation to the emergency department or any unplanned admission into a hospital environment with at least one overnight stay. HF worsening is defined as either diuretic up-titration or HF-related worsening quality of life, as assessed by the Kansas City Cardiomyopathy Questionnaire [17]. Secondary endpoints of the study include major adverse cardiac events, emergency department visits, all-cause rehospitalization, and death.

SENTINEL-HF’s rich data collection is focused on 3 main activities (home monitoring, interviews, and event tracking), which are initiated during the baseline hospitalization and continued up to 90-days post-discharge. The study procedures are summarized in Figure 3 and are detailed in the following sections. Data collected during the study is outlined in Table 1.

Patients with possible ADHF are identified using a computerized patient tracking system in the emergency department and by monitoring hospital admission logs on a daily basis. Screening is performed based on the diagnosis, chief complaint, and medical history. The eligibility of potential participants is confirmed based on their medical record information including laboratory records, ECG reports, and physical examination findings. Eligible patients are approached during hospitalization and provided with information about the study. Patients who agree to participate sign an informed consent and an authorization for disclosure of protected health information during the follow-up period (eg, details of subsequent readmissions). If study staff have concerns about the capacity of a potential participant to provide informed consent, an assessment [20] is conducted to determine the patient’s ability to express choice, to understand relevant information, and to appreciate the situation and its likely consequences. Patients who fail the assessment instrument are ineligible for the study.

Shortly before discharge, each enrolled participant is provided with a personal FAV monitoring kit consisting of a measurement vest of appropriate size, an electronics module, a study dedicated mobile phone, a manual of operation, and a short form instructions and troubleshooting guide. Staff adjust the measurement vest to the participant’s chest circumference ensuring that the electrodes have good contact with the participant’s skin, and train participants, and their caregivers when applicable, to use the study equipment.

During a 45-day home monitoring period, participants are required to perform daily self-assessments with the FAV at a consistent time of day, preferably upon waking up in the morning. Study staff use scheduled follow-ups to ensure that the self-assessments are carried out correctly and, if necessary, to provide assistance in the form of additional instructions for using the FAV. On day 1, staff call each enrolled participant to check on their experience with the first FAV self-assessment. On day 7, study staff visit the participants at home providing a booster training for the use of the FAV system. On day 14, study staff call the participants encouraging them to continue with the daily self-assessments throughout the following month (Figure 3). Participants who require additional assistance at any time during the home monitoring period may call a study hotline maintained by the sponsor (Philips Healthcare).

Participants’ adherence to the daily self-assessment routine is monitored based on the data transmitted to the remote database. Study staff call participants with 2 consecutive missed or non-evaluable days to identify problems and to provide support in resuming the daily self-assessments. If 2 consecutive missed or non-evaluable days continue to occur after the call, study staff visit the struggling participants at home to troubleshoot technical issues and to walk the participants through the FAV self-assessment routine. Participants are removed from the study if 2 consecutive missed or non-evaluable days continue to occur after the home visit.

In order to identify characteristics associated with the acceptance and the successful use of the FAV, participant interviews are conducted in-person or over the telephone at baseline and at day 7, 14, and 45 post-discharge (Figure 3). During these follow-up interactions, key patient reported outcomes and covariates, including severity of depressive symptoms and HF self-care are assessed. An overview of collected data and instruments used in the interviews is provided in Table 1.

At each follow-up interview, hospital-to-home transition quality is assessed using the 3-item Care Transition Measure (CTM3), an assessment designed to measure the quality of care transitions from the patient’s perspective [23]. The measure has good psychometric properties and has predictive validity for rehospitalizations. We also assess for HF self-management, patient response to changes in symptoms of HF, and confidence with self-management using the validated Self-Care for Heart Failure Index [19]. We measure both general and disease-specific quality of life at baseline and during each follow-up interview. General quality of life is measured by using the Short Form Health Survey SF-12, a validated 12-item version of the SF-36, a quality of life instrument often used for patients with heart disease [18]. HF-specific quality of life is assessed using the Kansas City Cardiomyopathy Questionnaire (KCCQ) [17].

Each participant’s psychosocial profile is characterized for a better understanding of factors associated with successful FAV use. Cognitive impairment and depressive symptoms are assessed at baseline and during each follow-up interview using the telephone interview for cognitive status (TICS) [21] and the 9-item Patient Health Questionnaire (PHQ-9) [22], respectively. We also gather information about clinical endpoints, including medication changes, doctor’s visits, emergency department visits, or readmissions. These reports are confirmed and extended by reviewing each participant’s medical record.

Clinical information (medical history, physical exam data, chest x-ray findings, and serum laboratory values) and demographics (age, gender, marital status, race/ethnicity, height, weight, and blood pressure and heart rate) are abstracted at baseline from the participants’ electronic medical records. Surveillance of participants continues for 90 days post-discharge by reviewing all additional clinical data including hospitalizations or physician visits related to possible ADHF, cardiac studies, consultations, discharge summaries, emergency service records, lab reports, office and clinic notes, operative and procedure reports, problem lists, pulmonary studies, and radiology results from the electronic medical records. Because the majority (80-90%) of patients hospitalized for ADHF at UMMMC follow-up with a UMMMC cardiologist and are rehospitalized at UMMMC, the vast majority of participant data is available in one centralized database of electronic medical records. Information related to patient-reported clinical events occurring outside the UMMMC health system is obtained from the respective healthcare provider based on the participant’s consent for solicitation of outside records. Endpoints are adjudicated on a rolling basis by 3 expert physicians (authors TEM, DDM, CD) blinded to FAV measures and clinical outcomes.

The study database is used to evaluate the performance of a previously developed ADHF detection algorithm informed by bioimpedance measures.

Our prospective sample size calculations were guided by the a priori assumption that the FAV would have a sensitivity of 0.75 to detect a primary endpoint of ADHF hospitalization. This assumption was based on a previous dataset collected for the development of our ADHF detection algorithm [12]. In SENTINEL-HF, we plan to have 107 patients complete the study period and adhere to the FAV daily over 45-days to exclude a real sensitivity lower than 0.60. To account for non-evaluable data and a 30% dropout rate, we require a minimum of 160 participants to provide a 0.1 95% confidence interval around an observed sensitivity of 0.75. We are employing an adaptive study design with interim sample size re-assessments based on the observed drop-out rate, non-evaluable data, and the number of ADHF events.

The SENTINEL-HF study is conducted in compliance with Good Clinical Practice and the principles outlined by the Declaration of Helsinki. The study has been approved by the Committee for the Protection of Human Subjects at University of Massachusetts Medical School with Institutional Review Board number H00001760. All study participants provide written informed consent at enrollment. The study is sponsored by Philips Healthcare and is monitored by the sponsor in compliance with the Code of Federal Regulations. Unanticipated adverse device effects are reported to the sponsor and to the Massachusetts Medical School Institutional Review Board. The study trial registration number is NCT01877369.