It's no secret that the American Medical Association has many objections about last month's unprecedented government release of Medicare payment data. But now that the data is out there, health care has to determine its usefulness – and that debate is just beginning.
In a letter last month to CMS, the association outlined a series of concerns, particularly about whether the data painted an accurate picture. “Instead of new insights into health care, the recently-released data have brought a series of sensationalist news stories, the majority of which inaccurately reported on the data, confused the public, and in some cases may have encouraged patients to make care changes that were not in their best interest," Dr. James Madara, the AMA's executive vice president and CEO, wrote in a letter to CMS Administrator Marilyn Tavenner.
The letter outlined a series of AMA contentions, largely focused around concerns about incomplete sets of data. In addition, some physicians have raised privacy concerns which compete with the basic premise that supporters of the release have echoed: Performing services paid for by the government opens you up to fair scrutiny.
Some of these limitations have been heavily publicized. For example, ophthalmology was quickly identified as the most “expensive” specialty, and just as quickly justified by the expense of drugs used to treat macular degeneration. That fact highlights one issue: Reimbursement for drugs is co-mingled with other physician payments, making it difficult to determine what percentage of costs is attributable to the cost of the drugs. It is also a billing reality than many providers file under a single NPI — another heavily reported factor that clouds the data.
Perhaps the most severe limitation is that the data set is only a slice of a physician’s whole patient population because it solely covers Medicare patients. That makes it difficult to create comparative figures, like the total number of times a physician performs a procedure.
So given all of these limitations, how can the data be useful? There are several ways, say supporters of the release.
The groups that could benefit the most are those creating networks of providers: Payers, employers and ACO participants. For example, the data provides information that insurance companies could use in helping determine the right mix of cost and coverage.
Within the data set, insurers can identify the E/M charge codes that rank the severity of patient conditions on five levels, with Level 5 being the most intense. Insurers may avoid including those physicians that are billing high volumes of Level 5 cases (taking into account, of course, that the physician in question may be a specialist who is only referred severe cases).
An employer with a self-insured employee health plan probably has the same concerns — and can benefit in the same way. So can ACOs, which adhere to the old proverb that a chain is only as strong as its weakest link: All providers within the organization stand to be penalized by high-cost physicians, and have the same incentive to exclude inefficient or expensive links.
This theory of the value of the Medicare data is gaining some traction. Tom Wadsworth, the senior vice president at MedeAnalytics, the largest privately owned health care analytics company in the country, told Healthcare Dive about it. Also, The New Yorker made a passing reference to it in an April blog post about the release, quoting Harvard health economist Amitabh Chandra. “This type of tiered networking, on a grand scale, could actually improve the efficiency of our delivery system. It is this version of transparency-driven tiering, Chandra believes, that could assist in our [health care] cost-containment efforts," The New Yorker wrote.
Of course, that doesn’t mean that the data is perfect, or complete. Wadsworth points to the data on ordering physicians that Medicare compiles. For example, you can see who performed a CAT scan, but not who ordered the procedure. Releasing that information, according to Wadsworth, would be valuable to everyone.
And yes, Wadsworth acknowledged, there are certainly people who are going to interpret it inappropriately. But there are always going to be non-expert parties who are going to interpret any data “inappropriately.” As a result, the real question may be whether the data is good enough to create meaningful change.
“If you could create a more selective data set that helps patients and physicians make better choices and drives better examples of what the data could do, then I’m all for it, so long as you’re not watering it down and losing a lot of value of the data,” Wadsworth told Healthcare Dive.
The AMA wants to see a number of changes in how the data should be treated. The group has called for the delay of further additions to the set and for physicians to have the opportunity to edit their data before it’s released. While these are both seemingly logical suggestions, they may be unrealistic at best, and at worst, capable of hamstringing the entire release.
“[Delay]," said Wadsworth, "is just another way of saying, 'never release it, or take so much out of it that it’s worthless,'" pointing to the impossibility of making any date set beyond repudiation. And while there’s nothing wrong with physicians correcting their data, this is a huge data set. Realistically, how can the organization accept and file those corrections without crippling delays?
The Medicare data set still may require a lot of refining — but that doesn’t mean it’s not useful as is, and, according to Wadsworth, it doesn’t mean that it should be buried until some far-off, indefinite date where all journalists, patients and other interested parties are Medicare experts. Time in the public eye and critical input from the AMA and others will improve the quality of future releases, backers say.