In December 2021, CMS' Center for Medicare and Medicaid Innovation (CMMI) released its latest evaluation of the Oncology Care Model (OCM), covering performance periods 1- 6, with episodes beginning July 1, 2016 through July 1, 2019. As a reminder, the OCM is CMMI's first cancer-specific alternative payment model that aims to decrease Medicare spending while improving quality of care for cancer patients. This blog reported on the results from the first two evaluations of the OCM covering performance periods 1- 3 and performance periods 1-5. We’ve updated the takeaways below to reflect OCM results through performance period 6, which didn’t vary much compared to the results through performance period 5. The evaluations compare OCM episodes with a matched group of comparison episodes attributed to oncology practice physicians that aren’t participating in the Model.
Keep reading to find out our top five takeaways from the latest report or jump to a specific takeaway below:
- ED visits, hospitalizations, and spending at acute care hospitals didn’t change
- Spending at the end of life decreased
- Practices only reduced low-value use of non-chemo drugs and imaging
- Medicare lost $377.1 million during performance periods 1-5
- Results varied based on beneficiary tumor site, risk, and race
1. ED visits, hospitalizations, and spending at acute care hospitals didn’t change
OCM practices implemented several new strategies to decrease ED visits and hospitalizations, such as offering access to urgent care, closely monitoring high-risk patients with outreach and follow-up calls, and expanding patient navigation and phone triage. But unfortunately, these efforts have proved futile thus far.
Although the latest report does not touch on ED utilization, results from the previous report show there was no reduction in ED visits or hospitalizations across performance periods 1-5, and patients didn’t show improvements in symptom management or adherence to oral chemo. The evaluation of performance periods 1-5 also found there were no clinically meaningful declines in ICU admissions or 30-day unplanned readmissions, even for visits or admissions due to chemotherapy toxicity. As such, spending on hospitalizations at acute care hospitals didn’t change either.
Though CMMI reported a $124 decrease in payments per beneficiary per six-month episode to “Other Inpatient Hospitals” (such as PPS-exempt cancer hospitals) during performance periods 1-3, it didn’t report on this metric in late evaluations. The slight, but statistically significant, 0.184-day decline in hospital length of stay originally noted during performance period 3 didn’t extend into performance periods 4 and 5 and was not reported for performance period 6.
CMMI hypothesizes that some ED visits and hospitalizations may be unavoidable, or that heightened contact with patients due to proactive outreach and rapid responses to patients’ phone calls may have led providers to identify additional situations to send patients to the ED. Non-OCM practices may also have focused on reducing admissions in response to other pressures, negating any comparative progress by OCM practices.
2. Spending at the end of life decreased
In the evaluation of performance periods 1-5, OCM practices saw a 1.1% decline in hospitalizations in patients’ last 30 days of life, leading to a $539 decrease in spending at the end of life per beneficiary per episode (compared to a $672 decrease reported over performance periods 1-3). This was mostly due to decreased Part A spending. Interventions such as increasing access to palliative care and expanding advance care planning might have contributed to these improvements. However, there was no change in the number of ED visits or use of Part B chemotherapy in the last two weeks of life or in the use, timing, or duration of hospice care. The evaluation of performance periods 1-6 did not include any additional end-of-life utilization data.
3. Practices only reduced low-value use of non-chemo drugs and imaging
In the latest evaluation of performance periods 1-6, OCM practices demonstrated more cost-conscious use of Part B non-chemo drugs, especially supportive care drugs like white blood cell growth factors. This led to an overall relative reduction of $161 per beneficiary per episode, which increased across the performance periods. The number of imaging services used declined slightly (but statistically significantly) as well in performance periods 3-6. OCM practices also exhibited faster adoption and greater use of biosimilars of filgrastim in the previous evaluation of performance periods 1-5, and made more clinically appropriate and less expensive decisions regarding antiemetic treatment without leading to greater symptom burden for patients. Information about cancer treatment choices was not reported for performance period 6.
However, despite national guidelines promoting higher-value alternative treatment options in many cases, use of radiation therapy services didn’t change through performance period 6. E&M visit patterns also remained the same. Similarly, physicians continued to prescribe high-cost chemo regimens, even when less expensive and equally effective options were available through performance period 5, though data on treatment selection was not reported for performance period 6. The silver lining is that this means practices didn’t restrict access to high-priced beneficial chemo regimens, limit adoption of new treatments, reduce clinical trial involvement, or avoid treating high-risk or high-cost patients, which were initial concerns. But there is evidence that physicians could have done more to limit spending while maintaining equivalent patient outcomes, such as employing hypofractionation.
4. Medicare lost $377.1 million during performance periods 1-5
Unsurprisingly given the results outlined above, the OCM didn’t reduce care spending enough to offset the monthly enhanced oncology services (MEOS) payments and performance-based payments (PBPs) paid to participating practices. Consequently, Medicare lost between $65 million and $100 million during each of performance periods 1-5, though losses tended to decrease across the performance periods. Medicare spending data for performance period 6 wasn’t available at the time the latest report was published.
Encouragingly, the most recent CMMI report shows that excluding the MEOS payments and PBPs, Medicare achieved gross payment reductions that grew with each performance period, peaking at nearly $50.2 million in savings in performance period 5. There was an average $298 reduction in total Medicare spending per beneficiary per episode during performance periods 1-6 although the reduction was only significant in performance periods 1-5. Part B payments remained the same across performance periods 1-3, but declined by $182 per beneficiary per episode across performance periods 1-6. This is primarily attributed to decreased spending on supportive care drugs, as Part B chemo payments didn’t change, and was the biggest driver of the total spending reduction reported per beneficiary episode. Part A payments also declined by $104 per beneficiary per episode across performance period 1-6, similar to the reduction documented in the previous evaluation. There was no change in Part D spending across performance periods 1-6, despite findings that Part D spending increased in the performance period 1-3 evaluation.
It's possible that OCM practices' success in reducing spending over time resulted in greater PBP payments, which were highest in performance period 4, as they met cost and quality goals and ultimately contributed (along with the flat-rate MEOS payments) to CMMI losing money.
5. Results varied based on beneficiary tumor site, risk, and race
Spending decreases were highly concentrated among patients with certain types of cancer. Total spending per beneficiary per episode across performance periods 1-6 was significantly reduced for beneficiaries with lung cancer (-$1,112), lymphoma (-$934), high-risk breast cancer (-$885), and colorectal cancer (-$865). Interestingly, the drivers of these savings varied by tumor site. Part B spending per beneficiary per episode was also reduced for beneficiaries with high-intensity prostate cancer, but this reduction was offset by a relative increase in Part D spending.
Spending also differed by beneficiary risk across performance periods 1-6. Beneficiary risk was defined using Hierarchical Condition Category (HCC) scores, which are assigned based on patients’ demographics and diagnostic histories, including comorbidities. Spending decreased among high-risk beneficiaries but increased among low-risk beneficiaries. Interestingly, the spending decrease among high-risk patients was not significant in performance period 6, which the most recent evaluation attributes to an increase of lung cancer immunotherapy payments for OCM episodes.
In the evaluation of performance periods 1-5, results varied by race, albeit inconsistently. Spending per beneficiary episode significantly decreased among white beneficiaries but didn’t significantly change among Black, Hispanic, or other-race beneficiaries. This is the opposite of the trend reported in the evaluation of performance periods 1-3, and further data was not reported in the most recent evaluation of performance periods 1-6. Additionally, Black beneficiaries were the only racial group to exhibit both a slight increase in unplanned readmissions and improvements in treatment adherence across performance periods 1-5.
It’s unclear why results varied across beneficiary groups. It’s possible that practices targeted their cost reduction efforts toward certain patient populations or that the OCM requirements were more conducive to reducing costs among particular beneficiaries.
The OCM’s consistently mediocre results so far suggest major methodology changes for future models
The most recent evaluation only covered the first six of the OCM’s nine completed and 11 total performance periods, which are now scheduled to continue through June 2022 due to the Covid-19 pandemic. However, since OCM practices had largely finished making practice changes to succeed in the OCM by the end of performance period 5, CMMI indicated that the results reported in the most recent two evaluations are indicators of what to expect in future performance periods.
Given the failure of the OCM to reduce Medicare spending so far, we expect CMMI will change its methodology for future medical oncology payment pilots. For example, CMMI is considering excluding low-risk episodes from total cost of care responsibility in future models since efforts to decrease spending were most successful among high-risk episodes. CMMI has also indicated an interest in assessing whether there should be enhanced service payments in any future model and, if so, what level of payment would be appropriate. But, nothing is set in stone yet, and results from future performance period evaluations could still influence the methodology CMMI chooses for its next medical oncology payment pilot.
Since CMMI has yet to unveil a plan to replace the OCM, it’s likely that it will let the OCM end in June 2022 without a new medical oncology payment pilot in place. Recently, the Community Oncology Alliance asked CMMI to extend the OCM until at least the end of 2022 while it considers a potential future payment model, as participating community oncology practices have made significant investments in process changes, staff, and infrastructure. However, CMMI has not yet responded to this request.