The BEAT-HF approach differs from telemonitoring programs previously described in several ways. It incorporates elements of successful care transition programs by engaging patients during their hospitalization and combining centralized structured telephone support with telemonitoring. Tele-HF was classified as a structured telephone support intervention, because it used an interactive voice response system that requires patients to enter information such as weight and blood pressure on a standard telephone keypad, but the ‘interaction’ was between the patient and a recorded voice. In BEAT-HF, the telephone is used for a coaching interaction between patient and nurse. In Tele-HF, monitoring data were sent directly to a cardiologist affiliated with the recruitment site, not the patient’s regular physician. The BEAT-HF nurses first call the patient in response to an alert trigger, enabling them to screen out false readings and to assess the severity of the patient’s symptoms. They also interact by telephone with patients’ regular providers, communicating the trigger in its clinical context and reinforcing existing patient-provider relationships by providing additional information about their patients’ needs.
Conducting transition coaching by telephone enables the nurses to support significantly more patients than if they were traveling to patients’ homes. Centralizing the nurses enables them to serve patients discharged from multiple hospitals, including those where the volume of heart failure patients would not be sufficient to employ a full-time nurse efficiently. The issue of scale is important because the prevalence of patients with heart failure may be too great to accommodate them in comprehensive disease management programs that rely on specialized clinics or home visits. In addition, some patients are too frail to make regular visits to outpatient clinics [33, 34]. The initial follow-up for monitoring triggers is performed centrally, assuring a uniform response. Finally, combining nurse coaching with telemonitoring allows the nurse to address issues related to patient use of the equipment.
Data captured as part of BEAT-HF will enable us to elucidate some of the reasons for non-adherence, a problem experienced by Tele-HF and other studies. These reasons may include real or perceived equipment failure or lack of reliability, patients who travel, patients who are admitted to skilled nursing or readmitted to a hospital, as well as behavioral non-adherence. This information will be extracted from the call logs maintained by the call center nurses and Ideal Life service representatives. Some of the technical difficulties may be due to human factors issues.
It is likely there will be variation in the ‘dose’ of intervention received by patients over the 180 days of the study in terms of both their adherence to telemonitoring and acceptance of coaching telephone calls. The primary outcome will be analyzed on an intention-to-treat basis. However, it also will be important to evaluate whether there is a dose-response effect. The ‘dose’ of telemonitoring will be captured automatically by the data transmissions. The dose of nurse coaching is affected not only by patients’ willingness to accept calls, but by the number of alert triggers generated by the patients’ transmitted parameters. The quantity of triggers may be associated with a patient’s severity of illness, but it may also reflect persistently high or low blood pressure or heart rate, or technical difficulties with the equipment.
The BEAT-HF study has other strengths. It is larger than most care transition or telehealth studies for heart failure, and may be the only large U.S. study to include a substantial proportion of patients with limited or no English proficiency. The patients are treated in different systems of care, with medical centers ranging from a safety net provider for a large, mixed urban-rural area to a major urban hospital that incorporates elements of both academic medicine and a large private practice medical staff.
At the same time, the study faces several challenges. The ascertainment of patient readmissions and ED visits is key. Because approximately 20% of readmissions occur to other (non-study) hospitals, and patients may not remember hospitalizations or ED visits, complete determination of utilization will depend on the success of matching patient identification data with OSHPD data.
Rapidly changing telecommunications technology led to technical problems at the beginning of the study. The Ideal Life equipment deployed initially used a modem that required landline telephone service to transmit data. However, a growing number of households no longer have traditional telephone service because they use a subscription or prepaid mobile phone exclusively, or they have bundled television, internet, and telephone service which disrupted landline data transmission. In response, Ideal Life introduced the Pod™, which uses any available cellular signal. Technology issues have led both research staff and Ideal Life representatives to dedicate more time than planned to individual patient trouble-shooting and home visits and increased the number of contacts between the call center nurses and some patients.
This study is taking place during a time of significant change in hospitals’ financial and regulatory environment. Reducing readmissions plays a key role in each of the top three priorities identified by hospital CEOs responding to an annual survey conducted by the American College of Healthcare Executives: financial concerns, patient safety and quality, and implementing health reform . To the extent possible, it will be important to identify other disease management initiatives underway at the participating sites and to assess their effect on both control and intervention patients . We recognize that other initiatives may create an environment that is particularly responsive - or unresponsive - to telemonitoring and patient education.
Despite our best efforts to recruit a sample that is representative of the target population of older adults with heart failure, our results may have limited generalizability. In particular, our sample will inevitably under-represent some of the highest risk, most vulnerable patients in our communities: those housed long-term in skilled nursing facilities, those without a usual source of medical care, those without insurance or with limited Medicaid coverage, those who require chronic renal dialysis, those with moderate-to-severe cognitive impairment, those who are too functionally impaired to use a scale and a telephone, and immigrants who speak only Asian or other languages not used in the study. However, it may be inevitable that system interventions, such as those studied in BEAT-HF, can only be delivered in certain settings, and must be adapted before they can be delivered in other settings or to other populations.
BEAT-HF is one of the largest randomized controlled trials of a telehealth approach to improving outcomes for patients with heart failure. It incorporates elements of successful care transition programs, combining centralized structured telephone support provided by nurses with home telemonitoring using the latest remote monitoring technology. BEAT-HF is enrolling patients with a wide range of socioeconomic and demographic backgrounds and collecting extensive data on intermediate factors that potentially affect patient adherence. As a result, the study will provide a wealth of information on how different individuals use technology and respond to interventions that are not face-to-face encounters. Once completed, BEAT-HF is poised to serve as an important research resource to understand how best to use telehealth approaches to improve key healthcare processes and outcomes, including care transitions and hospital readmissions, and to set the stage for future comparative effectiveness research on chronic disease management for heart failure.