Author + information
- Received July 19, 2016
- Revision received October 24, 2016
- Accepted November 15, 2016
- Published online March 27, 2017.
- aDepartment of Medicine, Duke University Medical Center, Durham, North Carolina
- bDuke-Margolis Center for Health Policy, Durham, North Carolina
- ↵∗Address for correspondence:
Dr. Zubin J. Eapen, Duke Clinical Research Institute, P.O. Box 17969, Durham, North Carolina 27715.
Telehealth offers an innovative approach to improve heart failure care that expands beyond traditional management strategies. Yet the use of telehealth in heart failure is infrequent because of several obstacles. Fundamentally, the evidence is inconsistent across studies of telehealth interventions in heart failure, which limits the ability of cardiologists to make general conclusions. Where encouraging evidence exists, there are logistical challenges to broad-scale implementation as a result of insufficient understanding of how to transform telemedicine strategies into clinical practice effectively. Ultimately, when implementation is reasonable, the application of these efforts remains hampered by regulatory, reimbursement, and other policy issues. The primary aim of this paper is to describe these challenges and to outline a path forward to apply telehealth approaches to heart failure in conjunction with payment reform and pragmatic research study design.
- alternative payment models
- clinical research networks
- health care policy reform
- Medicare Access and CHIP Reauthorization Act (MACRA)
- Merit-Based Incentive Payment System
According to annual report data acquired by the American Heart Association, the Centers for Disease Control and Prevention, the National Institutes of Health, and other government entities, approximately 5.7 million Americans over the age of 20 had heart failure (HF) diagnoses in 2012, which is expected to increase by a projected 46% to >8 million American adults by 2030 (1). With regard to financial burden, an estimated $30.7 billion was spent on HF in 2012, more than two-thirds of which was spent on medical care. By 2030, this estimated total cost is anticipated to increase by 127% or roughly $69.7 billion (2). Telehealth, the use of medical information obtained via nontraditional means of communication from one location to another for the goal of patient care (3), conceptually offers an approach to redesign and improve care outside the doors of the hospital and clinic. Traditional technologies such as landline telephones and emerging technologies including remote monitoring via implantable and wearable devices have been studied as part of management strategies aimed at reducing hospitalizations associated with increased morbidity, mortality (4), and costs for patients with HF. Yet telehealth has not been broadly adopted for the management of HF.
Widespread implementation of telehealth practices in HF has been slow because of a variety of obstacles. Evidence of benefit is not consistently demonstrated across studies of telehealth interventions in HF, which limits the generalizability of such interventions. Where promising evidence exists, there are logistical challenges to broad-scale implementation because of an inadequate understanding of how to translate telehealth strategies into practice efficiently. Finally, when implementation is reasonable, the application of these efforts remains encumbered by regulatory, reimbursement, and other policy issues. The primary aim of this paper is to describe these challenges and to outline a path forward for telehealth approaches to HF.
Existing evidence for the use of telehealth in HF includes a variety of interventions studied, ranging from telephone-based interventions to telemonitoring via implantable devices, but the generalizability and implementation of these results are encumbered by small scale, mixed outcomes, and limited understanding of the feasibility of translating interventions from clinical trials to the real world. To develop conclusions regarding the state of the evidence for the use of telehealth in HF, a search of published research was performed in a comprehensive way as follows. The search was focused on randomized controlled trials, systematic reviews, and meta-analyses. Using the PubMed, Cochrane Library, and ClinicalKey databases, primary search terms included “telehealth,” “telemedicine,” and “heart failure,” with the following Medical Subject Headings terms: “telemedicine”; “heart failure”; “heart failure, diastolic”; “heart failure, systolic”; and “congestive heart failure.” By studying relevant meta-analyses, a manual search of the reference lists for significant, pertinent primary randomized controlled trials was performed. Finally, landmark trials of interest to the authors were included. Only published studies were included in the narrative review.
In 2010, Inglis et al. (5) published a meta-analysis of 25 peer-reviewed, randomized controlled trials comparing structured phone support and/or noninvasive telemonitoring with usual care of patients with HF. Noninvasive telemonitoring interventions entailed objective monitoring of blood pressure, weight, electrocardiograms, or rhythm strips for review. The primary outcomes were all-cause rehospitalization, HF rehospitalization, and mortality of any etiology. The included studies tended to be small and of variable quality. With regard to HF rehospitalization, both interventions—telephone support and telemonitoring—had positive effects compared with usual care, with relative risks (RR) of 0.77 (95% confidence interval [CI]: 0.68 to 0.87; p < 0.0001) and 0.79 (95% CI: 0.67 to 0.94; p = 0.008), respectively. Telemonitoring reduced all-cause mortality (RR: 0.66; 95% CI: 0.54 to 0.81; p < 0.0001), but the positive effect of telephone support on all-cause mortality was not statistically significant (RR: 0.88; 95% CI: 0.76 to 1.01; p = 0.08).
In contrast to this meta-analysis, larger, high-quality randomized clinical trials have recently shown no benefit to noninvasive approaches to telehealth in HF. The Tele-HF (Telemonitoring to Improve Heart Failure Outcomes) trial randomized 1,653 recently hospitalized patients with HF to a telephone-based interactive voice-response system that collected daily information about symptoms and weight that was reviewed by the patients’ clinicians (6). There was no significant difference between the intervention and usual-care groups with respect to readmission for any reason or death of any cause within 180 days after enrollment. Of note, adherence to daily calls to the interactive voice-response system had dropped to 55.1% by the end of the study.
Similarly, the BEAT-HF (Better Effectiveness After Transition–Heart Failure) randomized clinical trial assessed the effectiveness of pre-discharge education, systematized telephone calls, and remote telemonitoring among 1,477 patients hospitalized for HF across 6 academic medical centers (7). Wireless electronic devices were used to transmit daily data, including patient blood pressure, heart rate, and weight to a nurse triage center, triggering a nurse response if clinical attention was required on the basis of pre-defined parameters. The transitional care intervention did not reduce all-cause hospital readmission at 30 and 180 days, nor did it decrease mortality at 180 days. The only clinically significant positive result of this study was an improvement in quality-of-life score on the basis of the Minnesota Living With Heart Failure Questionnaire (8).
Like Tele-HF, another takeaway from BEAT-HF is limited patient adherence to telemonitoring and telephone calls, with slightly more than one-half of patients using the intervention within the first 30 days of the trial. Nonrandom interview responses from only half of the study population creates a reporting bias that likely skews the study results in favor of the intervention. Beyond these study challenges, poor patient engagement creates further challenges in implementing remote monitoring strategies. In Tele-HF and BEAT-HF, the burden of interfacing with the technologies used may have outweighed the value patients perceived. Trials investigating telemedicine interventions that require a significant degree of “patient activation,” a term that comprises the insight and knowledge of patients as well as their ability to perform the action despite external stressors (9), may generate varied conclusions as seen in these large-scale trials.
Notably, Inglis et al. (10) published a 5-year update to the 2010 Cochrane meta-analysis assessing the effectiveness of structured telephone support and noninvasive monitoring for patients with chronic HF. The meta-analysis now includes 17 new peer-reviewed, randomized controlled trials, such as Tele-HF, in addition to 24 studies reviewed in the original paper. The impact of both structured telephone and noninvasive telemonitoring interventions on all-cause mortality and HF-related hospitalizations was positive overall with the addition of the new studies. Yet the attempts of these meta-analyses to present an overall effect remain limited by the inclusion of such a wide variety of studies with variable quality, size, intervention type, adherence, and results. Evidence from large-scale studies of high methodologic quality, including Tele-HF and BEAT-HF, characterize the difficulty for telehealth strategies to improve outcomes in HF.
Invasive approaches have also been investigated in several multicenter, randomized trials, which have produced inconsistent results. COMPASS-HF (Chronicle Offers Management to Patients With Advanced Signs and Symptoms of Heart Failure) was a multicenter trial of 277 patients with New York Heart Association functional class III and IV HF symptoms with implanted continuous intracardiac pressure monitoring (11). In the treatment group, information including right ventricular systolic and diastolic pressure and rate of change in right ventricular pressure, estimated pulmonary artery diastolic pressure and right ventricular pre-ejection and systolic time intervals, heart rate, body temperature, and activity were used to direct HF management. The primary outcome of this randomized, parallel-controlled and single-blinded trial was a reduction in combined HF-related events, including hospital, urgent clinic, or emergency department visits during which intravenous diuretic therapy was required. The study did not identify a statistically significant difference between patients undergoing implanted hemodynamic monitoring and the control group receiving usual medical therapies.
In contrast, the IN-TIME (Influence of Home Monitoring on Mortality and Morbidity in Heart Failure Patients With Impaired Left Ventricular Function) was a randomized controlled trial examining the impact of telemonitoring data from cardiac resynchronization therapy defibrillators and implantable cardioverter-defibrillators on HF management in 664 patients (12). Clinical management guided by telemonitoring data from these devices significantly reduced a composite clinical endpoint consisting of HF-related hospitalization or death, worsened New York Heart Association class status, or patient-reported decline in health (odds ratio: 0.63; 95% CI: 0.43 to 0.90). The investigators also reported a 1-year all-cause mortality improvement (hazard ratio: 0.36; 95% CI: 0.17 to 0.74).
Similarly, CHAMPION (CardioMEMS Heart Sensor Allows Monitoring of Pressure to Improve Outcomes in NYHA Class III Heart Failure Patients) was a single-blinded, multicenter, randomized controlled trial, which included 550 patients, assessing the ability of the CardioMEMS implantable device, a pulmonary artery pressure sensor, to reduce HF hospitalizations (13). The primary outcome results revealed a 37% decrease in the rate of HF hospitalizations in the treatment group. It should be noted, however, that most readmissions for HF are for non-HF reasons. Among readmissions within 30 days of a HF hospitalization, 63% are for reasons other than HF (14). All-cause rehospitalization, which represents a more patient-centered outcome and the subject of financial penalties, is considerably more frequent than HF readmission. Therefore, the impact of implantable hemodynamic monitoring devices on HF-related hospitalizations should be thoughtfully interpreted (15).
Establishing robust conclusions from meta-analyses featuring telemonitoring trials is limited by mixed results and strength of evidence. Feltner et al. (16) published a systematic review and meta-analysis of transitional care interventions to prevent readmissions for patients with HF. Telemonitoring was 1 of the interventions and was investigated in 8 of 47 randomized controlled trials. In the telemonitoring intervention group, the effect of telemonitoring on all-cause 30-day readmission was not statistically significant, and the effect of telemonitoring on 3- to 6-month all-cause readmission and HF readmission was without benefit in the context of evidence with moderate strength.
Regarding the cost of telehealth interventions, in a meta-analysis analyzing 11 randomized controlled trials including more than 3,700 patients, the investigators compared implantable device telemonitoring to standard of care for HF management and concluded that device monitoring was associated with a reduction in health care utilization, which was considered a substitute for direct health care costs, by an estimated 15% to 20%. Notably, because of inconsistent cost analysis methodology among studies, the investigators were unable to evaluate cost using a meta-analysis (17). Moreover, Grustam et al. (18), in a subsequent systematic review evaluating the cost-effectiveness of telehealth in HF, described their findings with the following statement: “We cannot conclude whether telehealth interventions in [congestive HF] are cost-effective or cost-ineffective: papers are heterogeneous and of poor methodological quality. Still, there is a political need for strong evidence.”
Logistical, Reimbursement, and Policy Challenges
Results from select studies suggest that noninvasive or invasive approaches to telehealth could improve outcomes in HF with the appropriate infrastructure to monitor data and manage patients. However, the limited uptake of CardioMEMS illustrates the challenges of translating telehealth trial design to real-world clinical practice. As with any diagnostic device, benefit can be demonstrated only when the diagnostic data can enable a treatment paradigm different from usual care. These alternative management strategies each require compatible information systems to collect and present telemonitoring data, personnel to assume monitoring and management responsibilities for patients outside the hospital or clinic, and protocols that enable these personnel to use these data as part of an alternative and standardized care pathway. Health systems often have to fund this infrastructure out of existing resources or seek grants and external capital, making it difficult to expand and sustain the necessary support.
In addition to these logistical barriers, which may slow adoption by providers, technologies used in telehealth are generally not widely covered in fee-for-service (FFS) payment systems and have had to prove their value to gain reimbursement for particular categories of patients. In the case of CardioMEMS, cost-effectiveness estimates range from approximately $13,000 to $56,000 per quality-adjusted life-year gained in a 5-year horizon (13,19,20). Additionally, the Institute for Clinical and Economic Review, a nonprofit entity that analyzes the costs of medical therapies, formulated a value-based price benchmark for the device ($10,665) for eligible patients and concluded that higher prices are associated with excessive costs (20). At the time of this publication, St. Jude Medical had filed for a national coverage decision from the Centers for Medicare and Medicaid Services after 2 regional Medicare administrative contractors chose not to reimburse the device, leaving it without coverage in several states.
In addition to reimbursement, policies governing telehealth may differ at the state level. With state-level differences in medical licensing and practice laws, there is considerable variation in practice and payment structure across the nation. For example, in Maine, “services provided via telehealth must utilize equipment that is capable of two-way video and audio (i.e., telephone, facsimile interactions and electronic mail delivered services are not reimbursable) . . . benefit must be related to physical, social or geographic issues that make delivering the service in person difficult” (ME ADC 10-144 Ch. 101, Ch. I, § 1). In contrast, Mississippi state law more broadly includes collecting remote patient data for condition monitoring and digital uploading of those data (Miss. Code. Ann. § 83-9-353). Interstate initiatives that involve data transfer and physician involvement across state lines are limited by these differences in regulations, physician state-based licensing, and ultimately reimbursement (21,22). These interstate differences in the interpretation of telehealth have limited both the scalability of standardized models across states and the development of centralized models in which the management of patients crosses state lines. Moreover, the Physician Quality Reporting System has not included telehealth services in the definition of patient-facing encounters, thereby limiting its reimbursement potential. Thus, in the current FFS model, coverage of telehealth tends to be limited.
Changing the U.S. Policy Landscape
Recent health care legislation will change reimbursement and, therefore, adoption of telehealth services. The Patient Accountability and Affordable Care Act aims to support telehealth in the context of accountable care organizations, which can use savings from lowering the cost of care to pay for services not generally covered in the traditional Medicare program. This alternative payment model (APM) is intended to foster evidence-based medicine, high-quality care, cost saving, and coordination of care “through the use of telemedicine, remote patient monitoring and other such enabling technologies” (42 U.S.C. § 1395jjj). The Medicare Access and CHIP Reauthorization Act (MACRA) of 2015 will further accelerate the shift from the traditional FFS payment structure to value-based care (Public Law No. 114-10). The proposed rule replaces the Medicare sustainable growth rate methodology for updates to the physician fee schedule and with the new Merit-Based Incentive Payment System (MIPS), which consolidates former reporting programs including the Physician Quality Reporting System, the value-based modifier, and meaningful use. MIPS has 4 performance measure domains: quality, advancing care information, and clinical practice improvement activity. Within the clinical practice improvement activity domain, health care providers must report on the provision of various practice-level activities, which can include telehealth services and the provision of 24/7 access. Unlike the Physician Quality Reporting System, MIPS will consider telehealth services to be patient-facing encounters, thereby improving reimbursement. Moreover, the Creating Opportunities Now for Necessary and Effective Care for Health Act, a bipartisan bill pending in the Senate (S. 2484), could further the aspiration of MACRA to reduce telehealth policy and reimbursement restrictions (23). In addition to these federal efforts to facilitate Medicare reimbursement for telehealth, 29 states and the District of Columbia have passed legislation to promote telehealth payment parity and encourage coverage by commercial insurers (24). A range of legislative proposals have sought to expand direct payment for telehealth services. However, because such policies are often viewed by actuarial and budget experts as cost increasing when not targeted well, these payment expansions have been limited.
MACRA establishes a pathway for additional advanced APMs that are voluntary but also provide bonus payments and exemption from MIPS. These advanced APMs require participating providers to share significant risk, not just savings, thereby making them accountable for losses as well as gains (25,26). Currently, 30% of Medicare payments are already associated with APMs. By 2018, the Centers for Medicare and Medicaid Services hopes to have 50% of reimbursements tied to APMs (27), although only a small fraction of these are likely to qualify as advanced APMs (28). The Health Care Payment Learning and Action Network, a Centers for Medicare and Medicaid Services network of private and public entities with the objective of fostering quality care through payment redesign (25), has created a framework in which the transition from traditional FFS to APMs can be easily understood (Figure 1). In their APM framework, 4 categories are used to describe health care payment models: the traditional FFS approach unrelated to quality or value (category 1), FFS with a link to quality and value (category 2), APMs that are built on FFS architecture (category 3), and APMs that feature population-based payment (category 4) (26). APMs in categories 3 and 4 can be condition specific, with benchmarks calibrated for costs and quality of care within a condition-specific population. Examples of APMs include bundled payments for patients hospitalized with HF (category 3) and population-based payments for HF care (category 4) (25,26).
These APMs present more of a financial incentive for health systems to invest in non-encounter-based chronic disease management programs and telehealth services depending on their cost-effectiveness (29).
These changes in health care policies create new opportunities for telehealth services to gain traction in the United States. Strategies that can improve the coordination of care outside of the hospital will be critical in helping clinicians provide patient-centered care to populations at a lower cost, a principal motivation for moving away from FFS payment systems that have had difficulty keeping up with innovations such as telemedicine. To date, the evidence base includes randomized clinical trials of various, dissimilar strategies that paint a mixed picture of the overall effectiveness of telehealth. As a result, clinicians are left without a clear direction for which strategies they should implement and how to implement them. Taking into account differences in systems-based practice from health care system to health care system, extrapolating the evidence for telemedicine on a global scale becomes even more challenging.
Translating telehealth into real-world practice could require large-scale, pragmatic trials to close the implementation gap. The National Institutes of Health Health Care Systems Collaboratory supports partnerships between health care systems and clinical investigators to promote and execute such trials (30). The National Patient-Centered Clinical Research Network also creates an avenue for patient-centered clinical research through the use of partner networks such as hospital systems, safety net clinics, and patient-led organizations to amass sizable datasets (31). By directly engaging patients, research networks such as the National Patient-Centered Clinical Research Network can evaluate not only traditional telehealth strategies but also emerging digital health approaches that use smart phones and wearable devices in the care of patients with HF. If supported with the evidence required for adoption, these digital health approaches could be even more attractive to providers participating in APMs because they could create even cheaper and more scalable ways to monitor and manage a diverse patient population. Telehealth as a tool to increase access to health care for medically unserved populations supports this goal (32).
Large-scale, randomized clinical trials across research networks will be neither feasible nor necessary for every variation of telehealth that a health system may seek to deploy. The stepped wedge design, in which an intervention is rolled out sequentially to an increasing number of patients, is a pragmatic and complementary approach that is increasingly being used in the evaluation of care delivery-oriented interventions such as telehealth (33). This design is particularly applicable in the field of telehealth, in which interventions: 1) may have limited evidence of effectiveness in particular settings; 2) are perceived to generally do more good than harm; and 3) can feasibly be deployed in only a segment of the overall patient population initially. Research networks or individual learning health systems could use the stepped wedge design to test and then expand the use of telehealth services that improve clinical outcomes and cost measures emphasized in APMs.
Risk sharing with telehealth vendors provides an opportunity to adopt interventions that have been appropriately evaluated. Pharmaceutical and telehealth services are establishing payment arrangements in which vendors share health care costs if outcome expectations are not met. For example, the large insurer Cigna negotiated with 2 manufacturers of PCSK9 inhibitors to require drug companies to discount costs for patients and also formulary prices if low-density lipoprotein cholesterol goals are not met (34). Cardiac device makers are also negotiating with health care systems to share risk on the basis of patient outcomes such as rehospitalization (35). Such approaches could be extended to partnerships involving telehealth services.
Telehealth strategies that shift components of HF care from the hospital to the clinic and ultimately the home can increase the value of care by lowering costs while maintaining or improving quality. However, more evidence on the effective use of telehealth in conjunction with payment reforms is needed first. The current evidence base includes the evaluations of varied approaches to telehealth in HF. A mixed picture of overall efficacy, implementation challenges, and uncertainty regarding the future policy and reimbursement landscape have left many providers without a sufficient level of confidence to incorporate telehealth strategies as part of redesigned care engaging APMs in telehealth work. To tackle the challenge of diverse and conflicting evidence, research efforts should evaluate accessible interventions with careful attention toward the practicality of such interventions. In the future health care landscape, telehealth strategies designed to advance high-value HF care in new value-based payment models provide opportunities to develop pragmatic solutions for patients, providers, payers, and policymakers alike (Central Illustration).
Dr. Eapen consults and serves on the advisory boards for Novartis, Amgen, Cytokinetics, Janssen, Medtronic, Myokardia, SHL Telemedicine, and Equity–Pattern Health Technologies. Dr. McClellan is a board member for Johnson & Johnson; and serves on the advisory board for American Well. Dr. Fraiche has reported that she has no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- alternative payment model
- confidence interval
- heart failure
- Medicare Access and CHIP Reauthorization Act
- Merit-Based Incentive Payment System
- relative risk
- Received July 19, 2016.
- Revision received October 24, 2016.
- Accepted November 15, 2016.
- Mozaffarian D.,
- Benjamin E.J.,
- Go A.S.,
- et al.,
- for the American Heart Association Statistics Committee and Stroke Statistics Subcommittee
- Heidenreich P.A.,
- Albert N.M.,
- Allen L.A.,
- et al.,
- for the American Heart Association Advocacy Coordinating Committee, Council on Arteriosclerosis, Thrombosis and Vascular Biology, Council on Cardiovascular Radiology and Intervention, Council on Clinical Cardiology, Council on Epidemiology and Prevention, and Stroke Council
- ↵American Telemedicine Association. What is telemedicine? Available at: http://www.americantelemed.org/about-telemedicine/what-is-telemedicine#.VylRcMfOdFJ. Accessed May 3, 2016.
- Inglis S.C.,
- Clark R.A.,
- McAlister F.A.,
- et al.
- Rector T.S.,
- Kubo S.H.,
- Cohn J.N.
- Hibbard J.H.,
- Stockard J.,
- Mahoney E.R.,
- Tusler M.
- Inglis S.C.,
- Clark R.A.,
- Dierckx R.,
- et al.
- Bourge R.C.,
- Abraham W.T.,
- Adamson P.B.,
- et al.
- Blecker S.,
- Paul M.,
- Taksler G.,
- Ogedegbe G.,
- Katz S.
- Sandhu A.T.,
- Goldhaber-Fiebert J.D.,
- Owens D.K.,
- Turakhia M.P.,
- Kaiser D.W.,
- Heidenreich P.A.
- ↵Ollendorf DA, Sandhu AT, Chapman R, Heidenreich PA. CardioMEMS HF system (St. Jude Medical, Inc.) and Sacubitril/Valsartan (EntrestoTM, Novartis AG) for management of congestive heart failure: effectiveness, value, and value-based price benchmarks. Institute for Clinical and Economic Review 2015. Available at: http://icer-review.org/sites/default/files/u148/CHF_Final_Report_120115.pdf. Accessed May 3, 2015.
- ↵Thomas L, Capistrant G. State telemedicine gaps analysis: coverage and reimbursement. May 2015. Available at: http://www.americantelemed.org/docs/default-source/policy/50-state-telemedicine-gaps-analysis---coverage-and-reimbursement.pdf. Accessed August 29, 2015.
- American Health Lawyers Association
- ↵Center for Connected Health Policy: The National Telehealth Policy Resource Center. Federal legislation: S 2484—Creating Opportunities Now for Necessary and Effective Care Technologies (CONNECT) Act. Available at: http://cchpca.org/federal-legislation-s-2484-creating-opportunities-now-necessary-and-effective-care-technologies. Accessed March 31, 2016.
- ↵Bryant M. Telemedicine and alternative payment models: an apt fit. Healthcare Dive. March 24, 2016. Available at: http://www.healthcaredive.com/news/telemedicine-and-alternative-payment-models-an-apt-fit/416274/. Accessed on April 19, 2016.
- ↵Alternative Payment Model Framework and Progress Tracking (APM FPT) Work Group. Alternative payment model (APM) framework: final white paper. January 2016. Available at https://hcp-lan.org/workproducts/apm-whitepaper-total.pdf. Accessed May 14, 2016.
- ↵U.S. Department of Health and Human Services, Centers for Medicare and Medicaid Services. CMS fact sheet: better care. smarter spending. healthier people: paying providers for value, not volume. January 2015. Available at: https://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2015-Fact-sheets-items/2015-01-26-3.html. Accessed May 14, 2016.
- ↵National Institutes of Health. NIH Health Care Systems Collaboratory: about us. Available at: https://www.nihcollaboratory.org/about-us/Pages/default.aspx. Accessed June 3, 2016.
- ↵National Patient-Centered Clinical Research Network. About PCORnet. December 2015. Available at: http://pcornet.org/about-pcornet/. Accessed May 24, 2016.
- Hemming K.,
- Haines T.P.,
- Chilton P.J.
- ↵Loftus P, Matthews AW. Health insurers push to tie drug prices to outcomes. The Wall Street Journal. May 11, 2016. Available at: http://www.wsj.com/articles/health-insurers-push-to-tie-drug-prices-to-outcomes-1462939262. Accessed June 3, 2016.
- ↵Lee J. Devicemakers explore risk contracts with hospitals. Modern Healthcare. December 6, 2014. Available at: http://www.modernhealthcare.com/article/20141206/MAGAZINE/312069964. Accessed June 3, 2016.