10 Reasons Why CMS's Cancer Payment Model Could Fail

The Oncology Care Model's purpose is to contain costs and improve care, but some oncologists and practice analysts say it is laced with perverse incentives.

The goals for the Centers for Medicare & Medicaid Services' Oncology Care Model are laudable: to decrease costs while increasing efficiency, to support the practice or evidence-based medicine, and to support earlier palliative care.

But some oncologists and practice analysts say the payment model is laced with perverse incentives and presumptions that may doom it to fail.

When it starts in spring of 2016, the OCM's primary purpose will be to hasten interventions and prevent hospital admissions for patients who experience chemotherapy complications such as fever or diarrhea.

CMS Announces Bundled Care Payments for Oncology

Those savings will come "from reductions in complications, emergency department visits, and hospitalizations that are expected from enhanced care coordination, use of evidence-based guidelines and improved end of life care," wrote Patrick Conway, MD, CMS chief medical officer, and other CMS officials in the Journal of Oncology Practice last month.

Starting the day of the first chemotherapy dose, CMS is offering to pay qualifying oncologists $160 per Medicare beneficiary per month for the duration of each six-month episode. In exchange, practices must establish triage clinics that respond around the clock.

Under the model's two tracks, practices that save at least 4% or 2.75% during the six-month episode, compared with that same practice's prior Medicare spending, will receive shared savings.

But the CMS model, the first in a series, is fraught with peril, says oncology practice expert Harold Miller, an analyst with the Center for Healthcare Quality and Payment Reform in Pittsburgh. Miller has worked on alternative models for the American Society of Clinical Oncology (ASCO), but "CMS has elected not to pursue our strategy," says ASCO Chief Medical Officer, Richard Schilsky, MD.

In ASCO's Journal of Oncology Practice, Miller argues that the CMS model has several major flaws.

1. It's Vulnerable to Being Gamed
The six-month "episode," which begins not with diagnosis, surgery, or radiation, but "arbitrarily" with the Medicare beneficiary's first chemotherapy dose, may prompt oncologists to game the system by timing treatments to last longer than six months so that a new six month-series will kick in, Miller and others say.

"Let's suppose I have a patient and they're on a six-month chemo regimen. Do I have a perverse incentive to delay one of the treatments so the drug costs show up on the next six-month episode? This is a serious problem," Miller says.

What if appropriate care requires a more expensive second regimen at month five. But would the doctor delay until month seven, when a new episode begins?" he asks.

"One possibility is that oncologists… might increase the number of episodes of treatment they initiate due to the availability of sizeable care management payments and performance based payments," says report prepared for CMS by two non-profit research organizations, RAND Corp. and federally funded MITRE.

2. Oncologists Can't Control All Costs
Participating oncologists would only see a financial upside if all costs for their patients during a six-month episode of treatment, including those over which they have no control, averaged lower than those in the practice's recent history. Those costs include any hospitalization or ED visit, drugs, skilled nursing care, home health, radiation oncology and all other Medicare-covered services.

But oncologists argue that they usually can't control those costs, especially if their patients go elsewhere for care, or require expensive drugs.

3. Lack of Risk Adjustment
The Oncology Care Model model offers no risk adjustment methodology to account for a practice that accepts patients that are sicker and costlier than others.

Barbara McAneny, MD, an Albuquerque oncologist who runs a federally funded demonstration project using 24/7 triage nurses at seven physician practices across the country, also believes the CMS model is flawed. "You expect to see three or four patients with very aggressive inflammatory breast cancer during the year, and then all of a sudden you get 12, and you're toast," she says.

Better Chemo Care Keeps Patients Out of the Hospital

And, she adds, CMS has not told oncologists how it will calculate newly approved, more expensive drugs that evidence says their patients should take.

4. No Real-Time Tracking
Miller says one of the strangest parts of the Oncology Care Model is that a practice won't know how it's doing until a year down the line.  

"CMS is literally going to come up with some regression formula based on claims data to figure out how to calculate what the right amount of spending was, sometime in the summer of 2017? That's nuts."

5. Unrealistic Expectations
McAneny says OCM is not suitable for practices that are already saving money, like hers.

"We've dropped hospitalizations by about half compared with the rest of our service area. Now Medicare wants to compare me to how I did last year? That's not fair. It should compare us to what the [Albuquerque oncology] market would look like without us in it. It shouldn’t compare me to me."

6. Little Incentive to Reduce Drug Costs
Medicare pays oncology practices "the average sales price (ASP)" of a drug plus a 6% markup. Commercial payers pay practices the ASP plus up to a 25% markup. Those markups provide revenue that helps practices pay for shipping, storage, and administration, as well as other office overhead costs, Miller says.

But it can also incentivize practices to buy more expensive drugs.

Miller says the range of choices is huge. "There are different drug regimens for the same type of cancer that range form $400 in cost to $64,000 in cost with the exact same survival rates."

The issue is further complicated because practices get discounts from purchasing organizations for buying in bulk, and they often choose one or two recommended drugs to use for all similar patients regardless of payer.

Do they buy cheaper drugs to glean shared savings from Medicare's oncology model and use them for their commercial patients too, giving up the 25% margin on a higher priced drug? Or do they avoid the oncology model altogether and continue buying the costlier drugs for the commercial insurer's 25%?

"That's also nuts," Miller says. "You don't want to force doctors to think about the financial impact on their practice of using one drug vs. another. You want them using the drug that works best for the patient."

7. Cherry Picking
The RAND/MITRE report cautions that CMS's Oncology Care Model might tempt some practices to "cherry pick" less-sick patients to avoid higher costs.

8. Stinting
Another unintended consequence highlighted in the RAND/MITRE report is stinting on care, or "switching patients to less expensive, but inappropriate, treatment regimens or recommending against clinically beneficial services, in order to receive performance-based payments."

9. Meaningful Use Attestation Required
To be eligible, practices will have to attest to Stage 2 Meaningful Use of EHR-certified technology, which many smaller or solo practices may not yet be able to do.

10. Oncologists' Behavior
McAneny says that perhaps the biggest obstacle to such a program's success may lie in changing the behavior of all physicians involved in a patient's cancer care.  

They shouldn't reflexively send patients to the hospital without first considering if prompt care can be rendered at home or in the doctor's office.

The RAND/MITRE report also mentions this challenge. For the model to break even, it says, oncologists "would have to reduce utilization and intensity by roughly 4%. Based on research on other similar payment models, a behavioral response of that magnitude may be possible, but it is by no means certain."