RPM in Chronic Disease Management
The increasing burden of chronic diseases has been a major driver for RPM programs. 38 RPM allows providers to deliver care in nontraditional health care settings. 38 It is beginning to reach the burden of proof required to catalyze large-scale adoption for specific indications. 38 The strongest evidence comes from the experiences of early large-scale adopters, from a few larger RCTs, and from meta-analyses and systematic reviews.
Lessons from Early Adopters
Kaiser Permanente was one of the early pioneers using home telehealth for chronic disease management. The Tele-Home Health Research Project was one of the earliest RCTs in home telehealth, enrolling 102 intervention and 110 control patients in 1996–1997. Patients with congestive heart failure, chronic obstructive pulmonary disease, cerebrovascular accident, cancer, diabetes, anxiety, or chronic wounds were eligible for randomization. Intervention patients received usual care (which included home visits), plus monitoring services that included peripheral devices to assess cardiopulmonary status and video equipment providing two-way interaction with home health providers 24/7. The outcomes measured included three quality indicators (medication adherence, knowledge of disease, and ability for self-care); service utilization; and patient satisfaction. There were no differences between usual care and home telehealth in the quality indicators or patient satisfaction. However, patients receiving home telehealth had total mean costs, excluding home health services, of $1,948, compared to $2,674 for the usual care group. Home health services were higher in the intervention group, but the study group concluded that RPM could potentially substitute for some of the in-person home services provided to both groups. 41
Faced with increasing demands for chronic disease management by an aging population of veterans, the VHA was another early adopter of RPM, and did so on a large scale. Following a successful Community Care Coordination Service (CCCS) pilot from 2000–2003, the VHA launched their Care Coordination/Home Telehealth (CCHT) program in 2003 using a first-generation commercially available system called Health Buddy (incorporated videophones, messaging devices, biometric devices, digital cameras, and telemonitoring devices). 18 The goals were to coordinate care of veterans with chronic conditions and reduce unnecessary admissions to long-term institutional care through the systematic implementation of health informatics, home telehealth, and disease management technologies. 18 CCHT leveraged RPM technology to scale its home services to care for a large population while carefully preserving the components shown to be important for successful care coordination and chronic disease management. 11 The study population included patients with diabetes mellitus (48.4%), hypertension (40.3%), congestive heart failure (24.8%), chronic obstructive pulmonary disease (11.6%), depression (2.3%), posttraumatic stress disorder (1.1%), and/or a mental health condition other than depression or posttraumatic stress disorder (2.3%).
Care was coordinated by a dedicated team of trained health care professionals, usually nurses or social workers. Patients were formally assessed by their care coordinator, who selected the appropriate home health technology, provided coaching and training to patients and family members, monitored patient data, coordinated services, and communicated with the patient's provider. Care coordinators selected biometrics to be tracked based on patient-specific medical conditions.
Messaging devices installed with disease management protocols (DMPs) texted questions to patients to assess their health status and educational needs. Based on preset thresholds for biometric data and survey questions, the telemonitoring device risk stratified patients daily and provided color-coded alerts for care coordinators concerning changes in status, allowing for immediate intervention. The census for CCHT expanded from 2,000 patients in 2003 to 31,570 in 2007, for a total of 43,430 patients. Analysis of care utilization between 2003 and 2007 showed a 25% reduction in days of inpatient care, 19% reduction in hospital admissions, and a patient satisfaction score of 86%. The annual cost of CCHT was $1,600/patient compared to $13,121 for the VHA's home-based primary services.
The Whole System Demonstrator (WSD) launched by the National Health Service (NHS) in England in 2008 was an RCT of the impact of home telemonitoring on utilization and outcomes among 3,230 patients with diabetes, chronic obstructive lung disease, or heart failure spread among 179 general practices in three areas of England over a 12-month period. The cohort with telemonitoring had 18% lower rates of hospitalization and 46% lower mortality rates than did patients in the control arm. 23 Although enrollment is slower than they had hoped, based on the WSD results, England planned to expand RPM to 3 million lives between 2012 and 2017.
The Ontario Telemedicine Network's Telehomecare initiative observed a 46% reduction in emergency department visits, 53% reduction in hospital admissions, and 25% shorter length of stay among 466 patients after enrollment in RPM. 24 Canada is also expanding its remote monitoring programs based on early successes.
Excepting ICDs and glucometers, RPM for chronic diseases typically relies on physiologic information from noninvasive biosensors and from patient survey. A 2016 review of RCTs by Vegesna et al. identified 62 studies on health outcomes. 38 Seventy-one percent were RCTs, and the number of patients varied from 30 to 730. The mostly common conditions were respiratory (22.6% of studies), cardiovascular (17.7%), obesity (17.7%), and diabetes/metabolic syndrome (16.1%). RPM has also been used in psychiatric illness, substance abuse, neurologic diseases, and cancer. Positive health outcomes were reported in 74.2% of the studies, whereas 25.8% were neutral. None identified negative outcomes compared to usual care. Although much of the available literature on the impact of RPM for chronic disease management comes in the form of small RCTs focused on specific conditions, the body of evidence for congestive heart disease, chronic obstructive pulmonary disease, and stroke is more substantive.
In a 2014 review, Bashshur et al. reviewed the empirical evidence of telemedicine for chronic disease management focusing on congestive heart failure, chronic obstructive pulmonary disease, and stroke. 39 The review included studies on RPM between 2000 and 2014 with a robust research design and a sample size of at least 150 patients. For each of the conditions studied, the preponderance of evidence supported benefit from RPM. In general, the benefits were from reductions in service use, including hospitalizations and emergency department visits, but there were often reductions in mortality as well. Few of the studies were mixed or failed to show a benefit. Economic analysis supported the benefits of RPM compared to usual care.
Several meta-analyses and systematic reviews confirm the conclusions reached by Bashshur et al. A 2010 systematic review by Inglis et al. focused on the impact of telemonitoring and structured telephone support in patients with congestive heart failure, analyzing 11 RCT RPM trials with a total of 2,710 patients. Telemonitoring was associated with 34% lower all-cause mortality and 21% fewer heart failure-related hospitalizations, 42 although the studies were small and of low to moderate quality. A 2015 review by the same investigators included nine new trials, 43 for a total of 3,860 patients. RPM was associated with a 20% reduction in all-cause mortality and 15% reduction in heart failure–related hospitalizations, with additional improvements in health-related quality of life, heart failure knowledge, and self-care behaviors. A third systematic review also focused on patients with congestive heart failure, analyzing 12 RCTs and nine observational studies comprising 3,082 patients. 44 The analysis demonstrated a 36% lower all-cause mortality and fewer hospitalizations among patients using RPM. 44
A systematic review of nine studies using telemonitoring or telephone support versus usual care in 858 patients with chronic obstructive pulmonary disease demonstrated that telemonitoring was associated with lower numbers of hospitalizations and emergency department visits, fewer scheduled visits, and a lower number of home visits. 45 A systematic review and meta-analysis of 26 studies (5,069 patients) on home telehealth (telemonitoring or telephone support) for management of diabetes, 21 of which used telemonitoring, found that telemonitoring was associated with lower glycosylated hemoglobin levels and fewer hospitalizations compared to usual care. 46
The Impact of Remote Patient Monitoring in Transitional (Postacute) Care
The HRRP outlined in section 3025 of the ACA stimulated the application of several effective transitional care strategies to reduce hospital readmissions, among them Project RED, 47 BOOST, 48 CTI, 49 and Transitional Care Model (TCM). 50 Successful transition programs share the following elements:
Patient education and coaching on self-management and symptom recognition
Follow-up planning and reminders
Coordination of information between providers/settings
Short-term intensive postdischarge follow-up
In theory, RPM should lend itself nicely to the core components of effective translational care programs and could optimize the efficiency and scaling of these services across large populations. Several health care systems are incorporating RPM into their translational/postacute care strategies.
Several early adopters of RPM into transitional care describe success. Geisinger integrated AMC Health's (New York, NY) telemonitoring platform into its complex care heart failure program in 2008. Examination of claims data from enrolled Medicare patients showed they had a 33% reduction in all-cause hospitalization, 44% lower 30-day readmission rates, 38% lower 90-day readmission rates, 51 and 11.3% lower care costs than when they were not enrolled. The return on investment (ROI) was estimated to be 3.3.
The University of Virginia, in partnership with Locus Health (Charlottesville, VA), incorporated RPM into a comprehensive care coordination strategy for postacute care patients over age 65. 52 Launched in 2013, it initially enrolled only patients admitted with one of the CMS readmission penalty diagnoses and demonstrated a 46% reduction in 30-day readmission rates across all diagnoses with condition-specific reductions ranging from 20% to 50%. As a result of its success, the program expanded to include all diagnoses hospital-wide and uses RPM as one tool in a multicomponent postacute care strategy that includes each of the elements common to successful programs, such as CTI and TCM. Enrollment of eligible patients approaches 80%, and completion rates are 90%.
Early RCTs using RPM for postacute care have reached disparate conclusions. Comparison of negative and positive studies provides some insights about possible drivers for successful implementation of RPM in transitional care coordination.
In 2004, a collaborative of New England health systems tested the impact of the Specialized Primary and Networked Care in Heart Failure (SPAN-CHF) disease management program to reduce readmission among 200 patients discharged with congestive heart failure. 53 The nurse-run intervention used weekly coaching telephone calls for 90 days after discharge. The SPAN-CHF intervention reduced hospitalizations for heart failure by 52% and all cardiac admissions by 43%. In 2010, a follow-on study, SPAN-CHF II, 54 had 188 patients randomized to receive the original SPAN-CHF intervention or SPAN-CHF plus automated home monitoring (AHM) with an interactive scale, blood pressure cuff, text messaging system, and the Bosch Health Buddy. Adding AHM further reduced readmissions, with heart failure admissions reduced by 72% (compared to 52% in SPAN CHF I) and all cardiac readmissions by 63% (compared to 43%).
Similar outcomes were reported by Davis et al., using the Cardiocom device (Chanhassen, MN) in a postacute transitional care program for underserved patients discharged to home from the hospital with a diagnosis of heart failure or chronic obstructive pulmonary disease. 55 In addition to daily biometric monitoring, the program included an interactive educational component and communication with physician providers if concerning trends were identified. Care coordinators performed medication reconciliation, helped the patient develop a care plan, and coached patients to set up regular follow-up appointments. There were 149 intervention patients and 1,028 controls. The duration of the intervention was 90 days. Patients in the intervention arm enjoyed a 50% reduction in 30-day readmission rates and 13% to 19% reduction in 180-day readmission rates. There was no difference in emergency department utilization. Program completion rates were 85%, and adherence to daily monitoring averaged 70%.
The BEAT-HF trial 56 was designed to evaluate the effectiveness of a care transition intervention using RPM to reduce readmission rates in patients aged 50 or older who were discharged after receiving care for heart failure. The intervention consisted of predischarge education; remote monitoring of weight, blood pressure, and heart rate; and nine telephone coaching calls over the 6-month study period. The study found no difference in 30-day or 180-day readmissions between the intervention and control groups. However, intention to treat analysis was used, and only 57% of enrolled patients completed the study; among those who did, adherence to telemonitoring >50% of days was only 55%, and only 61.4% of enrollees completed >50% of the scheduled coaching calls. The low compliance and completion rates make it difficult to draw firm conclusions about the effectiveness of the intervention.
Intensive Care Unit Monitoring of Providers
RPM can support providers as well as patients. Tele-ICU, also known as eICU, refers to the use of an offsite command center staffed by intensivists and critical care nurses connected to inpatients at distant ICUs via real-time video, audio, and electronic means to support care rendered by onsite providers. 57 The concept was originally introduced in the 1970s as a video link and evolved with digital and Internet capabilities into a fully integrated system with continuous access to patient monitoring data and electronic records that provides a safety net of supplemental critical care expertise for onsite caregivers. 57,58,59 Contrary to the overall trend for inpatient units to downsize and shift more care to the ambulatory setting, ICUs are expanding. The Leapfrog Group found that, on average, length of stay was 3 days shorter and ICU mortality rates 40% lower in ICUs staffed by intensivists 60 and have made adequate staffing by intensivists one of the metrics by which hospital performance is measured. 57,61 In 2015, 44% of reporting hospitals met the Leapfrog ICU physician staffing metric, up from 10% in 2007. 60 However, a shortage prevents some hospitals from hiring enough intensivists to manage their ICUs full time. The demand for intensivists is expected to increase by 38% between 2000 and 2020, 62 and current training programs are not turning out providers in numbers sufficient to meet the demand.
The tele-ICU model is one way ICUs can stretch the available workforce to cover demand. The first integrated tele-ICU model was launched by the Sentara Hospital System in 2000 in Norfolk, Virginia, using a platform developed by Visicu (Philips; Amsterdam, Netherlands). 58 Since then Visicu was acquired by Philips, and Cerner and iMDsoft have joined the U.S. tele-ICU market. The average tele-ICU covers 138 beds across 4.6 hospital systems staffed by one critical care nurse for every 30 to 35 beds and one intensivist for every 100 to 130 beds. 62,63 In 2003, 16 (0.4%) hospitals used tele-ICUs. By 2010 the number had increased to 213 (4.6% of total). 64 During the same timeframe, the number of ICU beds covered by telemedicine increased from 598 (0.9% of total) to 5,799 (7.9% of total). 64 Despite penetration of tele-ICU support across the country, reported outcomes have been mixed. Proponents have reported reductions in length of stay, hospital and ICU mortality, and overall costs. 58,62 According to Mackintosh et al., most of the studies reported suffer from significant methodologic shortcomings. 65 They identified only two studies of sufficient rigor to be included in their review. One of the best before and after studies included 6,290 adult patients admitted to one of seven ICUs on two campuses of an 834-bed academic medical center over a 2-year period and showed that implementation of tele-ICU was associated with a 60% reduction in hospital mortality. 66 Following implementation of the tele-ICU there was also reduced ICU mortality, shorter length of stay, lower rates of preventable complications, and lower overall costs. 66 The second study included 118,990 adult patients from 56 ICUs in 32 hospitals from 19 U.S. health systems and observed a 16% reduction in hospital mortality, 26% reduction in ICU mortality, and reduced hospital and ICU lengths of stay following the implementation of the tele-ICU. 67 The aspects of the tele-ICU intervention associated with improved mortality and/or reduced length of stay included review of each case by the intensivist within an hour of admission, timely use of performance data, adherence to ICU best practices, and quicker alert response times. 67
Tele-ICUs have not worked in every health system and have been deactivated in some. Limitations include high start-up costs, inconsistent interoperability between the onsite EMR and established commercial tele-ICU platforms, unclear impact on patient outcomes in certain cases, and initial resistance and skepticism on the part of some care teams. 62 It is estimated that it costs approximately $30,000 to $50,000/ICU bed in one-time or capital costs to implement a tele-ICU program and an additional $30,000 to 40,000/year/bed in staffing costs for offsite tele-ICU providers. 58,62,63,68 Currently third-party payors, including CMS, do not reimburse tele-ICU intensivists and critical care nurses operating offsite. The limited financial analyses available from systems that have implemented tele-ICU programs indicate that the model is profitable; however, more data are needed to draw general conclusions. In some instances, the impact of the tele-ICU cannot be dissected from that of other quality improvements in the ICU. 63
Among systems reporting profitability, factors driving the improved margin include decreased lengths of stay allowing increased volumes of new admissions, onsite retention of patients at community hospital sites with tele-ICU coverage, and increased occupancy rate of ICU beds which more than offset the capital and operational costs of the tele-ICU. 58,62 In one system analyzed, the breakeven point was achieved within a year, subsequent to which the combination of the increased contribution margin per case and the increased case volume resulted in a higher total contribution margin. 58 Although payors do not currently reimburse tele-ICU staff, they stand to benefit financially from the decreased lengths of stay and decreased complications experienced by those whose lives they cover. In summary, although some studies show that the tele-ICU model dramatically improves patient outcomes, stretches resources, and reduces costs, taken as a whole, the available data are mixed. Most of the published peer-reviewed studies suffer from limitations in design, and most well-designed studies have been assessed as being at some risk for bias. Notwithstanding these limitations, the tele-ICU model holds considerable promise and is now operational in 11% of the nonfederal ICU beds in the United States. 58
ICUs are not the only venues providing RPM support for providers. RPM is increasingly being used to monitor the provider performance in postacute facilities that provide rehabilitation and skilled nursing services.