publication date: Mar. 8, 2016
In an Experiment, CMS Will Vary Part B Drug Payments by Providers’ ZIP Codes
By Paul Goldberg
Is Average Sales Price plus 6 percent the right amount to pay doctors under the Medicare Part B program?
Would a smaller margin diminish what may be an incentive to prescribe the most expensive drugs on the market? With clinical performance being equal, or close enough to equal, is it not better for the doctor’s wallet to bill 6 percent of the highest ASP available?
In a move that immediately set off an explosion in the cancer field, the Centers for Medicare and Medicaid Services is on the verge of announcing a Part B Drug Payment Demo, a project where ASP add-ons would vary based on ZIP codes.
The objective of the system-wide experiment is to track utilization patterns without looking at patient benefit. The study was apparently developed without direct input from cancer groups—and without their knowledge.
Though the experiment was slated to be made public this week, a summary appeared on the CMS website recently, apparently by mistake. This flawed rollout—more of a drip than a leak—allowed oncology groups to blast the experiment even before it was announced.
“Based on preliminary information that was released, we strongly urge you to withdraw any consideration of implementing this initiative,” a sign-on letter from more than 60 groups of physicians and oncologists wrote in a letter to Andy Slavitt, CMS acting administrator.
“We are deeply concerned this risky, unproven experiment to Medicare Part B drug payments will jeopardize the health of millions of Medicare patients with cancer,” the groups said in the letter dated March 7. “The proposed experiment to be implemented by the Center for Medicare & Medicaid Innovation appears simply to focus on Medicare drug spending rather than on patients and the quality of medical care they receive.
“Any CMMI experiment that forces these vulnerable Medicare patients to abandon treatments that are working and improving their quality of life is misguided and ill-conceived.”
The full text of the letter as well as the CMS documents are posted here.
This experiment reflects reality of what CMS can and cannot do, observers say. The agency can try to manage costs by cutting the margins that go to oncologists for administration of chemotherapy. It can also experiment with prescription patterns. What it cannot do is regulate drug prices directly.
The quality and value of care are not even the endpoints of the study, at least in the version that was inadvertently made public. The list of groups that signed a letter opposing the CMS study includes the American Society of Clinical Oncology, the Community Oncology Alliance, the Association of Community Cancer Centers, the US Oncology Network, Cancer Support Community, CancerCare, and a large number of state oncology societies.
CMS did not respond to questions by deadline.
The Design of the CMS Experiment
The prematurely released document, a transmittal to Medicare contractors, states that the decision to start the experiment is final. It would begin on July 5. The mechanism would test “ASP payment limit values” based on ZIP codes. Hospital outpatient departments, physician offices, and pharmacies would be affected.
The transmittal’s description of the experiment follows:
CMS is providing this document in order to assist contractors as they implement the Part B Drug Payment Model. Although the model has not yet been finalized through notice and comment rulemaking, the information below can be used as interim guidance treated to the implementation CR.
Updates to this information will be made available on a flow basis. Medicare pays for most drugs that are administered in a physician’s office or the hospital outpatient department at Average Sales Price (ASP) plus 6 percent, with additional separate payment for drug administration under the Medicare physician fee schedule (MPFS) or the hospital outpatient prospective payment system (OPPS).
The ASP is calculated on a quarterly basis using manufacturer-submitted data. Medicare pays providers such as hospitals, physicians, Rural Health Clinics, Federally Qualified Health Centers and pharmacies for Part B drugs. Every quarter, CMS creates a new ASP file for MCS, FISS and VMS systems with a national price listed by Healthcare Common Procedure Coding System (HCPCS) code. To our knowledge, currently, the claims system does not make any other adjustments to the payment rates listed on the ASP file.
CMS restates some prices on the ASP file with retroactive pricing to a previous quarter and ASP payment amounts frequently change from quarter to quarter. CMS relies on updated pricing submissions from drug manufacturers each quarter and provides a quarterly pricing file for the Medicare Administrative Contractors (MACs) after validating and updating the manufacturer-submitted data. Because of the pricing submission schedule, the final contractor ASP file each quarter is frequently released very close to production deadlines. CMS is considering testing different versions of ASP price files in different areas of the country across all Part B drug settings using a mandatory participation model.
For example, there may be two alternative ASP price files each quarter, in addition to the ASP price file that is currently released each quarter. We anticipate that the payment value determination for a given claim in a given geographic area that is participating in the model would be accomplished by overlaying payment limits that appear on the current “ASP” price file with model-derived payment limits in the geographic areas that are being evaluated. Certain model payment approaches may also require separate add-on payments; in such a situation, we may utilize a G-code in addition to a drug HCPCS code. CMS anticipates that most Part B drugs, particularly the drugs that appear on the quarterly ASP price file, would be included in the model.
However, CMS would develop a list of drugs that would be excluded from the model. CMS anticipates that drugs that are currently contractor priced would continue to be contractor priced. Testing would include monitoring Medicare Part B spending and providers’ prescribing patterns; implementation is anticipated to begin on August 1, 2016. This implementation date would require changing the ASP pricing files in the middle of a quarter for some geographic areas that are assigned to model-based payment limits. The geographic areas may be as large as an entire MAC, but would more likely be a subset of the MAC, based on ZIP codes or perhaps MSAs or similar units.
In other words, the MAC may have to accommodate several versions of the ASP price file that would be used in different areas of the MAC jurisdiction. Monitoring and testing of the model would be performed by separately by CMS and its partners.
Cancer Groups Object
In their sign-on letter objecting to the plan, cancer groups argue that there is no data that would suggest that payment changes would either improve the quality of care or reduce costs.
In fact, a project by UnitedHealthcare implemented within community oncology practices designed to eliminate any “incentive” proved the exact opposite to the CMS assumption. According to the study, “eliminating existing financial chemotherapy drug incentives paradoxically increased the use of chemotherapy.” The spending on drugs increased by 179 percent.
The paper, “Changing Physician Incentives for Affordable, Quality Cancer Care: Results of an Episode Payment Model,” was published in the Journal of Oncology Practice.
In the paper, by Lee Newcomer et al., five medical groups that agreed to take part in the project were compared with a large national payer registry of fee-for-service patients with cancer. The objective was to examine the difference in cost before and after the initiation of the payment change. Between October 2009 and December 2012, the five groups treated 810 patients with breast, colon, and lung cancer using the episode payments.
Based on the registry, the predicted fee-for-service cost of the episodes cohort would have been $98.1 million. The actual cost was lower by almost a third: $64.8 million.
The cost of chemotherapy drugs would have been expected to be at $7.5 million, but was in fact higher, at just under $21 million.
“Modifying the current fee-for-service payment system for cancer therapy with feedback data and financial incentives that reward outcomes and cost efficiency resulted in a significant total cost reduction,” the paper concludes. “Eliminating existing financial chemotherapy drug incentives paradoxically increased the use of chemotherapy.”
That paper is posted here.
Also, the margin CMS is using in an effort to either test or influence physicians’ behavior is in reality lower than 6 percent. In their letter, cancer groups said the add-on is closer to 2.3 percent.
“CMS must understand the actual Part B reimbursement rate before implementing fundamental changes that may have serious consequences for patients and providers,” the letter states. “The ASP methodology currently includes a customary distributor prompt pay discount which reduces Part B reimbursement to approximately ASP plus 4 percent.
“Furthermore, Medicare applied the Budget Control Act of 2011 mandatory 2 percent sequester cuts to Part B drugs in such a way that the actual payment set by Medicare, after the prompt pay inclusion, is equivalent to approximately ASP plus 2.3 percent. It is imperative CMS understands and evaluates this current reimbursement rate and its outcome–especially as practices continue to close or consolidate with large health-systems, increasing costs for both patients and Medicare–and engage multiple stakeholders before implementing any initiative that would further reduce reimbursement rates.”