In a recent Health Affairs blog, Bowers et al¹ discussed the rationale for Hospital at Home (HaH) programs. Part of their argument for HaH included a mention of the scope of hospitalizations in the Medicare population² and the shortcomings of the Oncology Care Model (OCM). The OCM is a program developed by CMS to provide incentives to transform chemotherapy use and improve quality in cancer care in the Medicare population. If there are savings compared to non-program controls, a portion of those savings may be paid to the participating practice.

In this commentary, we address three issues related to this article: 1) the misperceptions in the discussion of the scope of the problem and the performance of the OCM, 2) the impact of the OCM on the performance and resultant savings achieved by The US Oncology Network (The Network), a group of independent, community oncology providers, and 3) the potential place of HaH in not only the OCM, but also in Medicare Advantage programs.

Restating the Problem

The authors make two key arguments for considering HaH. The first is that hospitalizations are expensive – accounting for 48% of costs.² However, when further examined, surgical admissions account for the same percentage of cost as does medical oncology. With medical oncology care, much of the expense occurs in the last 30 days of life.³

A study conducted by The Network using OCM data confirmed the last 30 days of life with ICU admissions is the most costly.⁴ In addition, in-hospital drugs and Part D drugs were not accounted for. A previous pilot study with Aetna Medicare Advantage indicated that 25.8% of costs were hospitalization related, 2.5% ER related and 71.7% were outpatient, predominantly drug costs. Drug savings were 21.4% over three years. In the same study, inpatient savings for solid tumors were 5.3%. Savings increased year over year in this program for both drugs and inpatient use.⁵ A recent analysis of The Network (unpublished) data in the Oncology Care Model showed costs per episode of hospitalization to be 10% of the total, with Part B drugs at 41% and Part D drugs at 21%.

The second argument is the notion of the OCM “falling short” of its promise. The authors mention a three-year study⁶ of the OCM performance, when in fact this was a study done after three years of the program but evaluating only the first three 6-month performance periods (PP 1-3) or the first 18 months. A second study referenced measured only the first 12 months.⁷ A recent paper catalogued the changes made by successful OCM practices half-way through (PP1-5). Key practice changes reducing hospitalization in those early years included patient navigation, targeted follow-up calls following chemotherapy, patient education, advance care planning, same day clinic visit, “call us first” campaigns, and patient triage.⁸

Hospital care is an important but not dominant cost center. There are ways to address hospitalization by restructuring the practices to follow patients more closely, get them into the clinic more expeditiously, identify those at high risk, and pay particular attention to end-of-life and hospice care. The restated problem then is – given that some of the changes made by practices in the OCM have taken up to a year to implement, when should we expect changes and what else can be done to improve performance?

Cost Savings in the OCM

The US Oncology Network has 14 practices participating in the OCM. Of these, over 50% have elected two-sided risk, meaning that if losses exceed a certain threshold, the practice would pay back those costs to CMS. In response to the external evaluation of PP 1-3⁶, The Network communicated a summary letter to CMS in August 2020 and issued a press release regarding this experience on November 11, 2020. The results as outlined below are otherwise unpublished.

Firstly, The Network compared participating practices’ performance for PP 1-3 with the published report. These Network practices achieved total savings (after deducting the supplemental OCM payments to the practices) of $231,000. This is compared to losses in the report of $154 million for PP 1-2. For The Network’s participating practices, hospitalization rate decreased by 6% with incremental improvement from PP1 to PP3. Hospice utilization increased 7%. This increase occurred in PP3.

Abt Associates recently evaluated the OCM from PP1-PP5⁹. Again, once the incentive payments were subtracted, there were no net savings. However, the losses were less for PP2 and 3 with a bump in PP4 to about $76 million related to the distribution of performance-based payments. Notably, costs for lung, lymphoma, colorectal cancer and high-risk breast cancer all decreased. Overall, Part A and Part B costs decreased slightly ($212 and $287 per episode, respectively). Hospice use was flat.

The Network then looked at savings from PP1 to PP5 (first 2.5 years). Savings were $78 million with hospitalization reduction of 7%, ER visit reduction of 4% and hospice utilization increase of 5%. In the latest press release, The Network calculated the savings through PP6 at more than $122 million based on experience with over 100,000 patients.

Additionally, The Network has not yet accounted for the rapid increase in the use of telemedicine seen since the advent of the COVID-19 pandemic. Telemedicine visits for The Network practices were negligible in January 2020. From March 2020 through December 2020, The Network averaged approximately 36,000 telemedicine visits per month. The cost implications of this use on hospital and end-of-life care are not yet determined, but there is potential to reduce hospitalizations and unanticipated ER and office visits.

These are preliminary numbers as the data have not been curated by CMS. However, the OCM program provides feedback on raw data on a timely basis. These feedback reports have correlated closely with the final reports. Figure 1 (unpublished data) shows the total cost curve for a representative participating practice in The Network through PP8. There is clear bending of the cost curve consistent with the projected savings. Figure 2 (unpublished data) represents the changes in hospitalization percentages over time. This same practice has lower rates than the average for both OCM practices and non-OCM practices. Notably, the rate of hospitalization has dropped for all groups.

Figure 1: Trends in Total Medicare Expenditures per Beneficiary per Month
Figure 2: Trends in Inpatient Admissions to Short-term Acute Care Hospitals and CAHs, All Cause Per 100 Beneficiaries
Source: Texas Oncology (The US Oncology Network) OCM Practice Feedback Report for the Period - April – June 2020

The Changing Landscape of Medicare Advantage Plans

The experience of the OCM will surely have an impact on the successful expansion of Medicare Advantage (MA) plans. Successful OCM practices may be considerable contributors to the efficient delivery of cancer care in these programs. In addition, MA plans will be able to incorporate hospice care into their programs such that payment for end-of-life care will fall all under one payment system. This may lead to further hospice use and hospital avoidance in the last 60 days of life.


Hospitalizations for chemotherapy-related cancer care as measured in the OCM are not the major cost driver, rather this was Part D and Part B drugs. The OCM has proven cost savings and has the potential for more mainly by focusing on drug spend. The reduction in hospitalization requires a multifaceted strategy. This includes better patient navigation, symptom management, access to outpatient treatment facilities and early hospice care. Hospital at Home may have a role, but it is far from clear what that role may be given the extensive changes seen in successful OCM practices in The Network.

CMMI Disclaimer: The statements contained in this document are solely those of the authors and do not necessarily reflect the views or policies of CMS. The authors assume responsibility for the accuracy and completeness of the information contained in this document.

About the Contributors
Russell Hoverman, MD, PhD
Russell Hoverman, MD, PhDVice President of Quality Programs, Texas Oncologyand Medical Director of Managed Care, The US Oncology Network
Stuart Staggs
Stuart StaggsSenior Director of Strategic ProgramsThe US Oncology Network
Lalan Wilfong, MD
Lalan Wilfong, MDVice President of Value-Based CareTexas Oncology
Marcus Neubauer, MD
Marcus Neubauer, MDChief Medical OfficerThe US Oncology Network
Lucy Langer, MD
Lucy Langer, MDChair of the National Policy Board Executive Committee, The US Oncology Networkand Practice President, Compass Oncology

  1. Bowers JT, Scott CA, Mooney KH et al. The CMS Oncology Care Model is falling short of its promise. Could Oncology Hospital at Home be the remedy? Health Affairs December 23, 2020.
  2. Brooks, GA, Li L, Uno H et al. Acute hospital care is the chief driver of regional spending variation in Medicare patients with advanced cancer. Health Affairs. 2014; 33:1793-1800.
  3. Chastek B, Harley HC, Kallich J et al. Health care costs for patients with cancer at the end of life. J Oncol Pract.2012 Nov;8:75s-80s. doi:10.1200/JOP.2011.0000469.
  4. Hoverman JR, Mann B, Phu S et al. Hospice or Hospital: The costs of dying of cancer in the Oncology Care Model. Palliative Medicine Reports. 2020;1:92-96.
  5. Hoverman JR, Neubauer, MA, Jameson M et al. Three-year results of a Medicare Advantage cancer management program. J Oncol Pract. 2018;144: e229-e237. https://doi:10.1200/JOP.17.00091.
  6. Abt Associates. Evaluation of the Oncology Care Model: Performance Periods 1-3.
  7. Brooks GA, Jhatakia S, Tripp A et al. Findings from the Oncology Care Model. J Clinicl Pract. 2019;15: e888-e896. DOI:10.1200/JOP.19.00265.
  8. Klein Rm, Brow M, Buescher N et al The Centers for Medicare and Medicaid Services Oncology Care Model halfway through: Perspectives from diverse participants. JNCI. 2019;111:764-771.
  9. Abt Associates. Hassol A Project Director. Evaluation of the Oncology Care Model. Performance Periods 1-5.