Much of the greatest innovation in industries other than health care deploys technology-enabled approaches to making services more accessible and convenient and lower cost or higher quality. For example, companies such as Blockbuster Video, which provided what seemed like essential services, was supplanted by Netflix, which offered on-demand, personalized viewing of videos from the convenience of home. Uber and Lyft have become quite popular, replacing hailing a taxicab. Similarly, in health care there has been a shift toward provision of home services, with substantial growth in areas such as provision of home infusion therapy and remote monitoring of blood glucose levels. The current coronavirus disease 2019 (COVID-19) pandemic has accelerated this shift because care in the home can reduce exposure to potentially infected individuals and help address capacity constraints in health care facilities. However, a common concern has been that payment models have not supported this type of innovation.

Traditional Medicare provides payment for services at home, but this payment is tied to recovery from acute events, and clinicians must certify that a beneficiary is homebound to be eligible. Primarily viewed as a substitute for more expensive facility-based care, care for recovery from acute events represents only a small portion of the type of care that could be provided at home that might be both less expensive and more effective than facility-based services. As experience in Sweden, central Texas, and northern California has shown, select patients with chronic kidney disease undergoing dialysis can be effectively self-treated at home, and outcomes such as blood pressure and mortality rates appear to be better.1 Patients with congestive heart failure, Parkinson disease and gait abnormalities, and other conditions could be remotely treated with automated assessment of and feedback on metrics that could be improved with support by clinicians, with office visits needed relatively infrequently.2 On March 30, 2020, the Centers for Medicare & Medicaid Services provided a waiver for homebound status in the setting of the COVID-19 pandemic, which, when combined with regulatory relaxation and the recent rapid adoption of telehealth, will allow for more rapid development of new clinical models at home.

The idealized care delivery model of the future would use a combination of remote monitoring sensors, bidirectional communication modalities to facilitate provision of care, and clinicians and nonmedical personnel working together to deliver coordinated services to address unmet clinical, behavioral, and social needs. Such a system could proactively keep patients healthy as opposed to reacting to disease events when symptoms progress. In this Viewpoint, we describe ways in which payers could think differently about catalyzing this type of health delivery system innovation.

Several principles are important to consider in this model. First, the payment model would incentivize clinicians and health care organizations to deliver services most likely to benefit the patient considering the cost and alternative uses of that resource. If a health system is only paid when a clinician sees patients in the office, the clinician is unlikely to enroll the patient in unreimbursed home-based services. However, services that currently are not reimbursed might be a better use of resources than services now being offered based on fee-for-service reimbursement; relative use is unlikely to change without changes in reimbursement. It is useful to consider what economists would think of as opportunity cost; resources used for one purpose are not available for another that might be of higher value. The payment system should encourage clinicians to recommend whatever service will produce the most health benefit for a given price without distortions, such as payment being offered only for visits in a medical facility.

Second, the proposed model would meet patients where they are. Whether clinicians are formally in a risk-sharing arrangement or not, system designers should view provision of care in the site that will be most efficient in improving health in developing care delivery innovations. Insurers only paying for in-person visits as a way of verifying that care was actually provided should become a thing of the past. Care should be simpler and more patient-centered, whereby patient preferences are factored into the site of care delivery when economically feasible.

Third, preventive and treatment services should be valued using a similar lens. The new payment model should be designed to encourage clinicians and health care organizations to assess whether care improves health at a reasonable price as opposed to saving money in the short term. Analytics should be used to systematically identify patients who are at elevated future risk to guide interventions that proactively lower risk. For instance, cancer treatment should consider the cost relative to the benefit, and smoking cessation program attendance should be encouraged even if it does not save money in the short term. A double standard for treatments and preventive services guarantees underinvestment in proactive approaches to preventing disease and overinvestment in managing acute illness.3

Fourth, the payment model should support customization. The concept of precision medicine should not just apply to using an individual’s genotypes to influence the choice of cancer treatments. Given the heterogeneity of patient clinical and social needs, preferences, and effectiveness of different interventions in different people, health systems need to continuously adjust what is offered to whom.4 There is a wide distribution of effectiveness of many interventions, and more health benefit for the amount spent can be attained by optimizing for each individual.

Technology is increasingly available to deliver personalized, convenient experiences providing health services at home. Analytics can help identify risk, anticipate clinical decline, measure biometrics remotely, and connect patients to appropriate clinicians virtually. The payment model needs to spur this innovation. Given the principles above, what might this look like?

One approach would be for payers to pay clinicians and health organizations through global capitation: health systems would take on full risk and assign each patient to a primary care clinician who could be the foundation of a system that provides high-quality lower-cost care. For individual clinicians or small groups, collaboration with firms that provide population management and financial accounting services could make participation possible. Capitated payments should be adjusted for both clinical and social risk, access, and other closely tracked quality metrics, and health care practitioners and organizations should have incentives to mitigate the rate of growth of health care costs. Increasing evidence suggests that progressive primary care payment models can spur increases in quality at no increase in cost.5 Consideration should be given to carefully account for the most severely ill 5% of patients who require extensive home services so that clinicians and health care systems are not de-incentivized from taking care of such patients.

A model for value-based purchasing in home care has significant embedded complexity beyond implementation in primary care, including its administrative complexity. For example, it is difficult to prospectively determine what population is appropriate for home care and to set benchmarks, and supporting the management of increasingly complex patients in the home requires significant logistical coordination. Value-based arrangements, in which primary care physicians and other clinicians take on full-risk contracts for patients, will encourage some types of care delivery innovations and not others. The experience with specialty care bundles successfully switching the site of postoperative care from skilled nursing facilities to homes suggests that shifts in payment will encourage clinicians to focus efforts on reducing costs from expensive hospitalizations or postacute facility stays by using approaches such as enlisting home care workers to support patients following hospital discharge. Although efforts to produce immediate health and economic returns are desirable, the economic returns from better management of diabetes or by enhancing functional status for a patient with Parkinson disease are less likely to be sufficiently immediate for decision-makers who are evaluated based on short-term performance and may be deprioritized by health systems being paid capitated rates. A focus on short-term returns is also likely given potential switching of patients between health plans or providers, and this is difficult to avoid.

New entrants in the market will likely step in to complement what existing health delivery systems provide if a payment model is established that rewards provision of services at home. A key factor for payers and health systems to agree on is what performance characteristics are most important, and to incent high performance on those metrics using a short list of behavioral economic principles to guide the design. For any type of clinician or health care system, the list of metrics should be limited to only the most important to preserve simplicity and the incentive system should be easy to understand, reward improvement and not just attainment of high thresholds, provide relatively frequent feedback, and selectively use loss framing to motivate greater effort.6 Social comparisons between and among clinicians and health care centers could be used as a way to further encourage better performance. The metrics should include assessment of patient satisfaction or a net promotor score (“Would you recommend this clinician or health care system to someone else?”). Incenting growth in total cost of care and providing a shared savings model that is shared with front-line clinicians would be appealing, but this should be “gated” by quality so that a clinician who attains scores of 50% or less might not be eligible for any total cost of care incentive and those who attain scores of 50% to 100% would receive the corresponding percentage of the total cost of care incentive, so that the full amount is only received if a clinician has perfect quality scores.

Although fee-for-service payment is often blamed for many problems in the health system, the lack of innovation in home-based services has not been due to the existence of fee-for-service per se, but rather that there is fee-for-service for facility-based but not remote monitoring services, communication, or care provision. This could potentially be solved by creating fee-for-service payments for each of these types of services. However, rather than micro-incenting numerous discrete activities, a more holistic approach—global payment—should be designed that establishes incentives for provision of services in the most appropriate setting.

Creating a payment model that encourages innovation in personalized home care is a necessary step for payers that would facilitate realization of the promise of home-based care in providing lower-cost high-quality approaches in areas such as home dialysis and chronic disease management. Rapid adoption of new clinical models in the setting of COVID-19 must be accompanied by payment models that will create sustainability. It is time to move beyond the limited definitions of home care currently used by Medicare and create payment models that support innovation of care provided in home-based settings.

Back to topArticle Information
Corresponding Author: Kevin G. Volpp, MD, PhD, Perelman School of Medicine, Department of Medical Ethics and Health Policy, University of Pennsylvania, 423 Guardian Dr, 1120 Blockley Hall, Philadelphia, PA 19104-6021 ([email protected]).

Published Online: May 21, 2020. doi:10.1001/jama.2020.1036

Conflict of Interest Disclosures: Dr Volpp reported receiving grants from Humana during the conduct of the study and grants from Vitality/Discovery, WW (Weight Watchers), Humana, Hawaii Medical Services Association, and Oscar; personal fees from the Center for Corporate Innovation; and personal fees from and being the co-owner of VAL Health outside the submitted work. No other disclosures were reported.

Funding/Support: Supported by a grant from Humana to the University of Pennsylvania.

Role of the Funder/Sponsor: The funder had no role in the preparation, review, or approval of the manuscript or decision to submit it for publication.

References
1.
Shinkman R. Is “empowered dialysis” the key to better outcomes? NEJM Catalyst. March 15, 2018. Accessed May 18, 2020. https://catalyst.nejm.org/doi/full/10.1056/CAT.18.0232
2.
Asch D, Terwiesch C, Volpp KG. How to reduce primary care doctors’ workloads while improving care. Harvard Business Review. November 13, 2017. Accessed May 14, 2020. https://hbr.org/2017/11/how-to-reduce-primary-care-doctors-workloads-while-improving-care
3.
Pryor K, Volpp K. Deployment of preventive interventions: time for a paradigm shift.  2018;378(19):1761-1763.PubMed
4.
Volpp KG, Krumholz HM, Asch DA. Mass customization for population health.  JAMA Cardiol. 2018;3(5):363-364.
ArticlePubMedGoogle ScholarCrossref
5.
Navathe AS, Emanuel EJ, Bond A, et al. Association between the implementation of a population-based primary care payment system and achievement on quality measures in Hawaii.  JAMA. 2019;322(1):57-68.
ArticlePubMedGoogle ScholarCrossref
6.
Emanuel EJ, Ubel PA, Kessler JB, et al. Using behavioral economics to design physician incentives that deliver high-value care.  Ann Intern Med. 2016;164(2):114-119. doi:10.7326/M15-1330PubMedGoogle ScholarCrossref