Cancer Groups Give Part B Plan an F

Organized oncology isn’t ready to give up on ASP+6 payment for Medicare Part B drugs. But experimentation with value-based pricing has some supporters.

Change is hard and often unpopular. Change how people get paid, and they will get up in arms in a hurry.

The reaction from oncology groups was overwhelmingly negative last month when CMS proposed altering the way it will pay for Medicare Part B drugs administered in physician offices or hospital outpatient clinics. The Association of Community Cancer Centers lambasted the CMS plan for interfering with providers’ ability to “provide critical cancer care services in their communities.” The American Society of Clinical Oncology (ASCO) averred its support of payment reform but accused CMS for “using heavy-handed reimbursement techniques” to manipulate treatment choices.

Getting rid of the financial incentive

For more than 10 years, CMS has paid physician and hospital outpatient clinics for Part B–covered drugs at a price equal to the average sale price (ASP) of the drug plus an additional 6% of the ASP (ASP+6, for short). CMS has proposed replacing ASP+6 with a payment formula set at ASP plus just 2.5% of ASP and a flat fee of $16.80. Why? Because basing payment on 6% of ASP is believed (not everyone agrees) to create a financial incentive to prescribe more expensive drugs: 6% of a $1,000 drug means more money for the physician or the outpatient clinic than 6% of a $100 drug.

Other specialists would be affected by the move away from ASP+6. But oncologists would be among the most affected because of the buy-and-bill tradition in oncology and the high price of cancer drugs. In 2014, roughly $4 billion of the $20.4 billion in Part B payments based on the ASP+6 formula went to oncologists and hematologists, according to figures included in the CMS proposal. Switching to the new formula would, by CMS’s reckoning, reduce the Part B drug payments to oncologists by 0.6%, or by about $24 million—obviously not a huge percentage or sum but real money nonetheless.

One of the most interesting things about CMS’s proposal is the design. The agency wants to run it like a four-arm randomized clinical trial. CMS is proposing to randomize 7,048 Primary Care Service Areas (PCSAs) [Primary Care Service Areas (PCSAs) are clusters of ZIP codes that reflect primary care delivery patterns. There are 7,144 PCSAs in the country, but Maryland is not included in this proposal because of its all-payer model.] to one of four groups (see box). Physicians and outpatient clinics will receive their Part B drug payments according to which group their PCSA has been randomized to. Critics have complained that this CMS proposal is an overreach because it is nationwide and doesn’t give practices any choice about participating. But many political and public health scientists want new programs tested in just this way because of the inherent selection bias when a program depends on volunteers.

The four arms of the CMS Part B drug payment experiment

  • ASP+6 (the control group)
  • ASP+6 with value-based payment
  • ASP+2.5 with $16.80 flat fee
  • ASP+2.5 with $16.80 flat fee and value-base payment

One thing that has gotten a little lost in the we’re-against-it clamor is the second phase of the CMS proposal, which is scheduled to begin in 2017. During that part of the program, the agency wants to test a variety of value-based pricing schemes in the Part B program, including reference pricing (using a benchmark price to set the price for a group of drugs) and indication-based pricing (higher prices for indications for which a drug is more effective).

Michael Kolodziej, MD, Aetna’s national medical director for oncology, is no fan of the first phase of the CMS proposal, which he believes will be a flop. But experimenting with ways to tie payment to outcomes is “fascinating,” although difficult to pull off, he says.

“If we can promote a universe in which the better outcomes get rewarded better than the average or mediocre outcomes—what is not to like about that?” says Kolodziej.

Our most popular topics on