On January 30th, a 3-judge DC appeals court overturned a lower court decision that would have forced public release of Medicare physician data. Writing for the majority in a split 2-1 judgment, Circuit Judge Karen LeCraft Henderson declared that
“The requested data does not serve any (freedom-of-information-related) public interest in disclosure. Accordingly, we need not balance the nonexistent public interest against every physician's substantial privacy interest in the Medicare payments he receives.”
But in a strongly worded dissent, Judge Judith Rogers, the third member of the ruling panel, found that the request by the consumer group, Consumer Checkbook, represented “a commanding and important public interest in disclosure of the data.”
She added:
“The crux of the court’s determination today that physicians’ privacy interests outweigh the public interest in disclosure is its conclusion that the requested data cannot assist the public in assessing either the quality of Medicare services or HHS’s efforts to combat fraud and waste…In reaching this conclusion the court overstates the inviolability of the privacy interest and overlooks the near undeniable fact that the requested data can be of some assistance to the public’s evaluation of how HHS is carrying out its initiatives aimed at measuring and improving health care quality and its efforts to combat Medicare fraud and waste.”
In 2007, Checkbook sued HHS under the Freedom for Information Act (FOIA) to release the Medicare physician data from four states and DC, and promised to make the data publicly available so consumers could understand physicians’ relative pricing and performance. The Bush Administration’s Department of Health and Human Services (HHS), in league with the AMA throughout this case, argued that doctors have a right to privacy that precludes the public’s right to know how government works. When the lower court found with Checkbook, the consumer group promptly sued for the data in the remaining states and territories, and HHS filed an appeal that was joined by the AMA. Amicus briefs supporting Checkbook’s position were filed by several prominent health care groups: AARP, the Center for Medicare Advocacy, Consumers Union (publisher of Consumer Reports), the National Business Group on Health, the Pacific Business Group on Health, and Judicial Watch.
This court’s decision creates an interesting set of dynamics. On the AMA’s site, a brief press release triumphantly states that
“The American Medical Association successfully fought to preserve protections against public disclosure of Medicare payments to physicians.”
“The court found that physicians have a significant right to privacy, and there is no public interest in the disclosure sought by Consumers’ Checkbook. The court clearly found that the release of personal physician payment data does not meet the standard of the Freedom of Information Act, which is to provide the public with information on how the government operates."
But the statement is attributed to a specific AMA Board member, Jeremy Lazarus MD. It must be awkward for the AMA’s new President, Nancy Nielsen, who has worked as a practicing internist for 23 years, served as Chief Medical Officer for Buffalo, NY health plan Independent Health, and represented the AMA at the National Quality Forum. Dr. Nielsen told the Wall Street Journal that she “endorses insurers’ use of report cards but wants doctors involved in what is rated.”
I am not a lawyer, so may not appreciate the full power of the ruling, but several observations come to mind.
First, the decision appears to be a very narrowly drawn response to the question of whether the release of physician-specific Medicare payment data meets the standard of the Freedom of Information Act. If that is the case, then it does not seem to be a very definitive or strong barrier to future efforts to make Medicare physician data publicly available.
And while the AMA undoubtedly will use the ruling to fight future efforts, the Obama Administration’s ascent presumably is accompanied by the belief that greater pricing and performance transparency are critical to the public interest, including efforts to identify and reduce health care waste and inappropriateness. It is difficult to imagine that this decision could withstand the overwhelming pressure to make transparency a core value of national health reform.
So the remarkable thing is that the judgment appears to have been made outside the context of the data’s potential impact on health care cost and quality. Checkbook’s President Robert Krughoff highlighted this point in his response:
“The inexplicable thing about the majority’s opinion is that the two judges in the majority didn’t understand, or acknowledge, any counter-balancing public benefit that would result from allowing the public access to the requested data."
A question now is whether HHS or CMS can simply make the Medicare physician data available. After all, physicians delivering services through Medicare and Medicaid are vendors taking public dollars. Hospitals providing services through those programs do not enjoy the same special protection, and their data are released to the public.
More to the point, while few credible physician data sets are publicly available at this time – the commercial health plans typically hold their data close, treating them as proprietary – a great deal of evidence has been developed over the years to show dramatic differences in the cost and performance of physicians by specialty and market. Jerry Reeves MD, the Chief Medical Officer of the UNITE HERE health plans, presents a slide showing a 6x-8x difference in resource consumption – that is, cost – between the least and most expensive physicians in a specialty and market to obtain the identical outcome.
At this point, patients have virtually no objective data available to guide their physician choices. Organizations like HealthGrades and Emily’s List aggregate peripheral information about physician education and training, patient experiences and brushes with medical malpractice incidents, but there is no substitute for credible data that would point to, say, rates/ranking on post-surgical infections, re-admissions or adherence to guidelines.
One irony here is that the Bush health team constantly trumpeted the idea of transparency, while blocking its fruition at every turn. This ruling is a last gasp of the old paradigm, an attempt to protect the entrenched interests that have brought American health care to its current sorry state.
My guess is that the Appeals Court’s decision for physician privacy at the expense of patient knowledge will be extremely short-lived, and end up being nothing but a minor negative footnote in the steady march toward better health care in America. Certainly, there is good evidence that some progressive health plans increasingly understand the value of using their data to drive better patient decisions, and to make physicians aware of their own performance.
Dr. Nielsen at the AMA has the experience, background and opportunity to own up to this reality as well. It will be interesting to see whether she possesses the strength of leadership to seize that opportunity.
But one thing is clear. This ruling is hardly the last word on physician transparency.
Categories: Uncategorized
The term Public works environment has recently been expanded to include digital public infrastructure projects and health care.
I think physicians should want this data to be public. What do they have to hide?
I don’t think enough people (including some judges) understand what disclosure of medical records implies.
Medical records may contain intimate details about sexual practices (eg with animals), criminal offenses (most common one is illicit drug use), mental health, and references of such regarding 3rd parties (eg as part of a family history).
If we selctively only disclose records not containing such details, we woudl imply that such details exist when we don’t release the information.
Collecting such information is essential for providing optimal care, especially in mental health. If patients start withholding such information or give dishonest accounts of themselves in fear that this information might leak out of the doctor’s office medical care will be seriously compromised. We are talking risking actual human lives here.
I have no objections to individual patients requesting access totheir *own* files (very few exceptions with some mental health patioents) – but I too would rather go to jail than releasing it to some snooping bureaucrat. I owe that much to my patients.
If the public is interested in taxpayer’s expenditure, the problem can be solved through government medical officers – sworn to and understanding the same privacy principles as the rest of us health professionals, and actually involved in patient care understanding the intricacies and difficulties of keeping medical records.
Yes, there are probably a few dozen or even hundreds of black sheep among my colleagues rorting the system – what we should ask ourselves is that if it is really worth risking human life for the sake of going the easiest path in rorting those few out? I think not.
Dr. Klepper,
I would love if you could elaborate a little more and maybe follow up on the literature you offered in your earlier post.
If you are talking about constructs that sound a lot like DRGs but that I have never heard of (“DxCGs, Clinical Risk Groups (CRGs), Episode Treatment Groupers (ETGs)”), I would be extremely curious how the risk adjustment works, and how this has been verified. For instance, it seems that the medicare data has to rely on the coder’s diagnosis … and this is already where the imprecision starts. A doctor who quickly codes for hypertension, for instance, would be doing much better than a colleague who is more hesitant to make this diagnosis, and the early diagnosis maker will have good outcomes or “cures” just due to regression to the mean.
I am all for transparency and quality control (and I am not an AMA member), but I would like to hear how this set of medicare data is supposed to deliver.
Joe Blow,
Sorry, but your statements are factually incorrect. There are entire industries that have been built around the credibility and reliability of this process.
While billing information does not contain lab scores and many other important bits of data about the clinical process – only clinical encounter data has that robustness – it is still chock full of clinical information that can inform both quality and cost evaluations. It can tell us, for example, which procedures were performed, whether people experienced infections and/or needed re-admissions following surgeries. It can help us identify procedures that might not fit with the known condition. It can tell us which care pathway was followed to obtain a specific outcome, and whether that path to the outcome was more or less efficient than another physician’s approach. And on and on.
At this point, claims are also typically the only available data sets that are full continuum. This allows us to view a case episodically rather than simply at the encounter level.
So while I understand your bias, from the perspective of understanding what’s buried in the data, you’re dead wrong.
OK, back to reality here. I’m all for releasing Medicare data on physicians, but that data set DOES NOT INCLUDE OUTCOMES MEASURES. It only includes billing codes. It can give you a list of how many times a given doctor used a certain Medicare billing code, but what it CANNOT TELL YOU is what antibiotics were used in which clinical scenarios, how often clinical guidelines were followed for sepsis, MI, pneumonia, etc.
The Medicare data set only tells you that Dr. X billed 2,343 instances of code ABCD, it cant tell you anything meaningful about doctor quality.
Dr Klepper I echo the questions and underlying concerns expressed and posted by rbar on Feb 6th. If you can back-up the accuracy of your physician profiling and outcomes based on Medicare data, I would be interested in additional background that supports your claim. I am working on developing a data base to steer patients to high performing doctors for major employers in NY.
The fact that “you cannot do simple outcome research with existent medicare data” and “you need high numbers” to do profiling based on outcomes does not excuse the AMA’s actions in fighting transparency.
The AMA, which is nothing more than a labor union for physicians, trots out the same tired arguments against openness, transparency, meaningful metrics and outcomes based compensation at every turn.
They do the entire population and the practice of medicine in this country a disservice, and it’s been clear time and again their only constituency is their physician membership. They are no different than the lobbyists for oil companies who would sell their country down the river to protect their membership from any risk of standards or measurement.
Brian,
If you have a high quality review article about physician profiling, I would be eager to read it (I have a good library backing with inst. subscriptions). To me, you seem to make an extraordinary claim … the claim that you can accurately do physician profiling just based on MC outcome data.
2 thoughts:
1) If you really want to do profiling based on outcomes, don’t you need high numbers, and doesn’t even the one or other oddball bad outcome influence things?
2) How come, myself being a fairly open minded physician working in a giant MSG with full EMR and powerful QC iniatives, that we (actually, my colleagues from IM since I work in a different specialty) work with selected test- (not outcome) parameters like HbA1c, normotensive BP readings etc?
In other words, a whole MSG/health system with excellent data access does things like HbA1c etc. for quality control, while you claim one can reliably profile MDs just with some medicare data?
And the medicare hospital iniative (I am sure you know the correct name of what I am talking about), don’t they rely on similar probes like: time to IV ABx in pneumonia, stroke patient d/ced on ASA etc.? I am not debating that these measures might be useful, but don’t they indicate that you cannot do simple outcome research with existent medicare data?
I would be appalled by this ruling, but what happens if Consumers Checkbook sets its sites on bloggers and consultants next? 🙂
This is an awful ruling that really puts some cold water on the whole transparency movement. While there are several issues (e.g., coding limitations, coding creep, etc) with trying to use ICD-9 procedure and diagnosis codes for quality/cost profiling, the reality is that claims are likely going to be the dominant method by which individual physicians are evaluated for the next several years.
Even a dominant Blues plan may only have 50-60% of an individual physician’s payer mix. It still leaves a notable gap in the record especially when trying to analyze their cost/quality patterns. Most of the national plans may have much larger membership but it is usually so diffuse that it is usually difficult to come up with the sample sizes to come up with meaningful profiles for specific conditions treated by an individual physician.
Self-funded employer coalitions such as Bridges to Excellence realized that they largely have the same numbers problems when trying to analyze individual physicians. Even if you get several large employers in a local geographic area, it is hard to get meaningful sample sizes for individual physicians.
Without widespread access to the Medicare physician files, meaningful transparency efforts are going to continue to languish even if the stimulus bills passes and a ton of EMRs are installed in individual physician offices over the next 3-5 years as currently outlined in the House version of the bill.
rbar,
All modern algorithms associated with physician profiling (e.g., DxCGs, Clinical Risk Groups (CRGs), Episode Treatment Groupers (ETGs)) risk adjust, which eliminates the old worry that “my patients are sicker.” At this point, there is a great deal of experience with profiling so, while there are occasional mistakes – and there MUST be avenues of redress – the analyses are pretty accurate. If you like, I can refer you to physician experts and literature about this very evolved discipline.
The pricing of each procedure – which as you point out, is set – is far less important than the utilization patterns, and therefore the episodic costs, required to obtain a desired outcome. I am involved in a company that builds and manages comprehensive onsite primary care clinics for employers, and we routinely evaluate the performance of the specialists in each market to identify which docs consistently obtain the best outcomes at the lowest cost. Then we use that data to steer toward the high performers and, equally important, away from the low performers. As you’d expect, this approach produces much healthier patients for significantly fewer dollars.
As you also note, we live in a capitalistic society. So long as data are unavailable and performance is opaque, the market can’t work. Reasonable people have every reason to apply the best tools available to understand who does a better and worse job, and to allow the most positive attributes of markets to work.
I am all for transparency, but I have a few questions after reading this:
(1) how can pricing be judged based on medicare data (since prices are set)? (As a side note, we have an at least semicapitalistic health care system right now … price tables for average self pay rates in a doctor’s office or online, as much as I would personally hate that, are only logical).
(2) Certainly, one can estimate practice patterns such as test ordering behavior, certain complications etc. based on medicare data, but who can guarantee that the data is correctly interpreted and not the source of misinformation (take 2 general internists, one seeing a younger, more healthy and affluent, the other one having a rather geriatric practice with ample NH pat … you get my drift. You can adjust for certain factors, but who guarantess that this is done right? That the geriatrician who might be doing excellent (and certainly hard) work is not punished by the numbers because his patients … uhmmm … tend to be sick and die?
I am afraid that some transparency advocates underestimate the difficulty judging physician quality. It certainly should be attempted, but it should be done right.