CMS's Dubious Medicare Data Dump
Imagine you're a small-business owner (or maybe you don't have to). The federal government decides to publish all your tax returns and those of other small-business owners on the Internet, but with all the expenses blacked out, leaving only income to be considered. Then, echoing the classic Mike Myers SNL sketch "Coffee Talk with Linda Richman," the government simply tells the public, in a heavy New York accent, "Talk amongst yourselves."
Would you, as Linda might say, be verklempt?
That is essentially what the Centers for Medicare and Medicaid Services (CMS) did just over a week ago, when it released a database containing the amounts Medicare Part B paid each health-care provider for various services, procedures, and in-office-administered medications. This came after Dow Jones, Inc. -- whose Wall Street Journal highlighted fraud and abuse in the Medicare program in 2010 based on limited claims data released by the government -- successfully overturned a decades-old injunction preventing the release of provider-specific numbers.
I am not against this move, at least to the degree it achieved CMS's goal, as stated in the related press release, to "help beneficiaries and consumers better understand how care is delivered through the Medicare program." Unfortunately, as implemented, this seemingly well-intentioned effort will produce more confusion than clarity, at least in the short run.
Yes, this information may ultimately prove useful to researchers capable of performing detailed statistical analyses. But, in their raw form, the aggregate amounts received by providers -- the numbers that drew the interest of most journalists -- are misleading to the public.
Why? Well, for starters, including the cost of highly expensive drugs (some up to thousands of dollars per dose) in the published totals dramatically increased the numbers for those who administer such drugs. For no other reason, this tended to vilify oncologists, rheumatologists, dermatologists, and -- yes -- ophthalmologists, who must use these drugs whenever they feel they are the best option for a given patient.
Excluding (or more clearly segregating) those amounts -- which CMS could have easily done -- would have brought the totals of most of those providers closer to earth. But even the more reasonable remainder is still misleading.
That is because, as every businessperson knows, no one sees a dime of this money until so-called "overhead" expenses -- staff, rent, supplies, etc. -- are taken care of. Interestingly, CMS could even have addressed that, since payments under Medicare are actually calculated based on standardized amounts for "work" and "practice expense." So if you happen to know that fact, and are capable of finding the needed values on the CMS website and do the required calculations, you're home free!
Except that patients -- at least those who don't also happen to be health-policy researchers at the RAND Corporation -- can't reasonably be expected to do that. In fact, they likely rely on -- indeed expect -- their government to make information "public-friendly" if it goes to the trouble to make that information -- er -- public.
And even those adjustments don't give the whole picture. For example, some providers may treat an unusually large number of patients who need expensive services, and generate higher revenue for that reason alone. So a provider treating a paper cut may have received less than, say, another treating the spurting artery of the overly-chef's-knife-happy Julia Child as played by Dan Aykroyd in another memorable SNL sketch. CMS's inclusion of specific claim data on disease processes, and attributing them to individual patients in a non-identifiable way (not 100 percent telling, but available), may have helped make such distinctions.
On a more basic level, some providers may have simply seen more Medicare patients than their colleagues. And, of course, none of this tells you anything about what providers -- particularly those receiving lower amounts -- may have received from other sources like private insurers and patients paying cash.
I suppose I can see how CMS could have decided this approach -- which did create headlines about the more egregious-sounding amounts -- might be more effective in raising awareness of the magnitude of Medicare outlays than, as had been done in the past, just releasing the overall totals (e.g., $77 billion in 2012 for Part B), because it adds a "context" to which the average person can relate.
But since CMS is mandated by law to pay for "medically necessary" services under Medicare -- and nothing in this data facilitates determining whether or not that occurred for a given patient encounter -- generating "buzz" based on almost uninterpretable provider-identified aggregate totals seems patently unfair. Furthermore, it tends to undermine the trust of seniors in the program without warrant.
I am not saying this data shouldn't be available. But CMS owes it to patients and -- yes -- providers to make the data it releases both persuasive and -- more important -- truly useful in decisionmaking. If a far more sophisticated analysis suggests problems beyond "Medicare is spending the money it is required to be spending" -- including possible fraud -- those problems should indeed be addressed.
But just like Emily Litella, marvelously played by the late Gilda Radner of SNL's "Not Ready for Primetime" Players (a label one might apply to this data), for now CMS may need to be willing to say: "Never mind."
Craig H. Kliger is an ophthalmologist and executive vice president of the California Academy of Eye Physicians and Surgeons.