OCTOBER 25, 2010 Secrets of the System
In Medicare's Data Trove, Clues to Curing Cost Crisis
By MARK SCHOOFS and MAURICE TAMMAN
Somewhere in the New York City area there is a family-practice doctor who, government records suggest, pocketed more than $2 million in 2008 from Medicare, the federal insurance program for the elderly.
That made her one of the best-paid family-medicine physicians in the Medicare system. But more noteworthy than the sum is her pattern of billing, which strongly suggests abuse or even outright fraud, according to experts who have examined her records.
This doctor didn't do typical family medicine. Instead, she administered a wide array of sophisticated tests, including polysomnography sleep analyses, nerve conduction probes and needle electromyography procedures—some of which have been flagged by federal antifraud authorities for special scrutiny. As a doctor of osteopathy, she has certifications for family practice and a hands-on treatment called "manipulative therapy," but none in neurology. She denies wrongdoing.
The Wall Street Journal is prohibited from naming this physician despite the fact that the paper detected her by mining a database paid for by taxpayers. Known as the Medicare claims database, it is a computerized record of the bills Medicare pays for medical treatment, and it is widely considered the single best source of information on the U.S. health-care system.
Other major insurance pools—including Medicaid, the government program for poor Americans, and individual private plans—provide insight into small slices of the American populace. Medicare, by contrast, insures virtually all Americans 65 and older—a population that consumes roughly a third of all of medical spending.
Federal investigators use the database to find fraud; academic researchers mine it to compare the cost and utilization of various services; and consultants make a business out of analyzing the data for a wide variety of health-care companies. In ways large and small, it offers an unparalleled look at why America's health-care costs are spiraling.
Test Case: Sifting the Medicare Database
The Journal focused on one family-practice doctor with extremely high billing and found she had performed in 2008 an unusual array of 29 sophisticated diagnostic tests. Some of these have been flagged by federal authorities as being frequently abused. The numbers here are derived from a 5% sample of Medicare beneficiaries.
But the Medicare data come with a severe limitation: While the services and earnings of hospitals and other institutional providers can be publicly identified, such information is kept strictly confidential for doctors and other individual providers. The reason is that the American Medical Association, the doctors' trade group, successfully sued the government more than three decades ago to keep secret how much money individual physicians receive from Medicare. The AMA has continued to defend this ruling, including in two cases in which federal appeals courts issued decisions last year.
This means the American public is barred from examining in detail how Medicare spends roughly an eighth of its funds, about $62.5 billion in 2009. While that may seem like a small piece, health-care experts point out that physicians have disproportionate power to direct spending in all the other areas of the system because they admit patients into hospitals, prescribe drugs and order procedures and equipment.
The AMA stands by its position and says little would be accomplished by publishing individual physician billing information.
What's more, the vast majority of doctors are honest and, far from getting rich off Medicare, tend to believe it pays poorly.
"The AMA has zero tolerance for health fraud, and we are working with the U.S. government to fight it," said Dr. Jeremy Lazarus, the Colorado psychiatrist who chaired the task force that developed the AMA's principles on use of doctor data. "We support the release of information that will help physicians improve the care they provide, but the release of personal physician payment data does not meet that standard, and physicians, like all Americans, have the right to privacy and due process."
The Wall Street Journal, in conjunction with the nonprofit Center for Public Integrity, attempted for nearly a year to obtain the database. As part of the effort, the CPI filed a lawsuit against the Department of Health and Human Services, which houses the Medicare program. The Journal and CPI wanted the data at no cost; the government wanted $100,000 for eight years of data. In a settlement, The Journal and CPI obtained the requested data at a substantially reduced fee. They later obtained a decryption key to identify individual providers but signed a contract agreeing not to publish such identities in most cases.
The database, technically known as the Carrier Standard Analytic File, focuses on doctors and others paid on a fee-for-service basis. It contains 5% of all beneficiaries, and includes all doctor claims that Medicare paid directly in association with their care.
But even with these limitations, the power of the database is clear. If it were fully available, with doctors clearly identified, the public could expose countless ways in which some health-care providers misuse or waste taxpayer dollars, health-care advocates say. The database could even provide some information on physician quality. Especially in the digital age, the database could be a powerful tool for holding the $500 billion Medicare program accountable.
"It's very hard to defend ignorance and willful hiding of data in the 21st Century," said former House Speaker Newt Gingrich, who has called for the database to be public as long as patients are kept confidential. "Our estimate is that the federal government, in Medicare and Medicaid alone, loses between $70 billion and $120 billion a year to crooks. You ought to be able to identify those."
Joseph A. Califano Jr., who tried to make Medicare payments to physicians public when he was secretary of Health, Education and Welfare under President Jimmy Carter, agrees. "Just the publicity, the embarrassment—aside from the actual prosecution of fraud—would have a tremendous impact" on fraudulent billing, he said. He added that opening the database could help state professional review boards "reveal incompetence in many cases, and I think that would improve the level of medical care." He said that patients should be kept private.
Mr. Califano's experience shows how effectively doctors have resisted such disclosure. In March 1977, amid a national debate over the cost of health care, the Carter administration released a list of all doctors who received Medicare reimbursements of $100,000 or more during 1975. The media covered it, publishing the names of highly paid doctors. The top earner was New York ophthalmologist Charles D. Kelman, practicing on East 58th Street, who billed $412,757, the equivalent of nearly $1.7 million today.
The AMA responded by saying the list was riddled with errors—a charge later upheld by the Comptroller General. A Michigan doctor was listed as earning $115,000 from Medicare, when he actually earned only $15,000. "My wife must think I have an apartment on the side and a mistress as well," he quipped at the time.
Reimbursements attributed to individual physicians often went to group practices. Dr. Kelman, for example, told The New York Times that two other doctors as well as optometrists and technicians shared in reimbursements attributed to him alone.
The AMA and individual doctors also fumed that publicizing physician incomes could stigmatize high earners and generally deter doctors from treating Medicare patients. They said the list provided no way to distinguish between a hard-working doctor and a crass fraudster.
The Carter administration issued an apology for the errors. But it also said it would release the names of providers who received Medicare payments during 1977, and the amounts.
To block publication, the Florida Medical Association filed suit in Florida, and that suit was joined by the AMA.
"It has long been a fundamental value in our society that, in the absence of a compelling state interest to the contrary, a person's financial affairs are nobody's business but his own," the AMA argued in court papers.
U.S. District Court Judge Charles Scott weighed two competing interests: that of the providers, whose privacy he said would indeed be invaded, and that of the public, which had an "important interest" in knowing how much taxpayer money was spent reimbursing Medicare providers. Judge Scott ruled that the public interest could be served by scrubbing disclosures of doctor names. So, he permanently barred the government from disclosing reimbursements that "would personally and individually identify" providers.
Technically, the ruling applied only to physicians licensed to practice in Florida and all members of the AMA who participate in Medicare and who would be identified. But in practice, the government has interpreted the ruling to bar identifying any individual provider, from nurses to physical therapists, in the Medicare claims database.
Over the ensuing three decades, this ruling has withstood all challenges. Last year, in a suit the AMA joined, a federal appeals court blocked identifying even the Medicare services individual physicians provided, on the grounds that publicly available fee schedules could be used to deduce how much Medicare paid the doctors.
In a case brought by the nonprofit group Consumers' Checkbook, the federal appeals court for the District of Columbia fortified the AMA's position. While ruling that doctors have a "substantial" interest in keeping secret the amount they receive from taxpayers, the court declared that the taxpayers' interest in knowing who was getting their money was, under the Freedom of Information Act, "non-existent" or "negligible at best."
Consumers' Checkbook argued that the database could be used to fish out Medicare fraud and abuse. Law enforcement officials and other anti-fraud experts widely regard the database as one of the best tools for identifying fraud, precisely because it can be mined for aberrant billing patterns. But the appeals court boxed Consumers' Checkbook into a Catch 22, ruling that the group had to have evidence of fraud before it could use the database to find that fraud. In the court's words, Consumers' Checkbook "has not provided any evidence of alleged fraud the requested data would reveal."
Today, a billion and a quarter claims pour into Medicare each year for Part A—which includes hospital, skilled nursing facilities and hospice—and Part B, including fee-for-service physician services, and durable medical equipment. There are more than 14,000 diagnoses and more than 7,000 medical procedures, most designated by Current Procedural Terminology, or CPT, codes. Code 75992, for example, is "Transluminal atherectomy, peripheral artery, radiological supervision and interpretation."
A full set of one year's data—with doctors' names encrypted and only 5% sampling available for physician claims and durable medical equipment—costs about $18,300.
Consultants, from one-person boutiques to large corporations such as Thomson Reuters, make a business out of putting this data into an easily understandable form and answering clients' questions.
Hospice care, for example, used to be provided mostly by local, not-for-profit outfits. Now, more than half of the Medicare-licensed hospice providers are for profit, according to Cordt Kassner, whose one-man consulting firm, Colorado-based Hospice Analytics, serves a variety of hospice providers and state hospice associations. Mr. Kassner said he spends about $10,000 per year purchasing Medicare claims data.
One of his clients is Michigan-based Great Lakes Home Health & Hospice, which served about 27,000 patients last year, according to the company, including 16,500 home health-care and 1,500 hospice patients. Great Lakes CEO William Deary says he used the data to identify relatively under-served cities. In 2007, Mr. Deary's company opened an office in one such city, Lansing, and in doing so "increased our hospice revenue by 88% in 36 months," he said.
For government fraud investigators, the database is a gold mine. Even the 5% sample obtained by The Journal illustrates its potential for highlighting unusual billing activity. The newspaper mined the database for 2008 outliers within specialties, including family medicine. Of the approximately 75,000 providers in this specialty, the family doctor who performed the battery of sophisticated sleep, nerve, and other tests was one of the top billers, with reimbursements of $142,522.12 on 44 patients.
Multiplying that 5% sample figure by 20 results in an estimated $2.8 million in total fees for this doctor for the year, although the wide statistical margin of error means it's impossible to pinpoint the exact number. A person with knowledge of the matter said the doctor's Medicare fees totalled $2.2 million in 2008.
Interviewed on a rainy Friday afternoon at work, the doctor acknowledged taking in more than $2 million from Medicare that year. She added that she ran a clinic with many employees and high overhead.
Within the 5% sample, this doctor is an outlier in several ways. Her billing shot up 16-fold from 2006 to 2007, and continued rising the following year. She averaged $3,239 in earnings per patient in 2008—nearly 18 times the mean for family-medicine doctors, and the 7th highest among family physicians with 10 or more patients.
Counting only diagnostic tests performed on at least two patients, the doctor performed or supervised 29 separate sleep, neurological, ultrasound and other diagnostic tests. Looking across all 811,785 providers in the 5% database, no other provider of any specialty conducted all 29 of those tests in 2008.
The Journal asked several fraud and billing experts to review spreadsheets showing this provider's billing from the 5% database. Her identity was kept private.
"The conspicuously large number of diagnostic tests appear medically improbable," said Kirk Ogrosky, a former federal prosecutor who specialized in Medicare fraud and is now a partner at Arnold & Porter.
The range of tests is "just so unusual, I don't see how that could be otherwise explained" than through abuse or fraud, said David Sand, medical director of HMS Inc., a company that helps numerous states control costs and root out fraud and abuse in the Medicaid system. The "breadth and depth of medical knowledge" required to do such an array of tests "defies comprehension," he said.
The New York-area physician, in the interview, denied any wrongdoing and said she only administered tests "recommended by the [medical] literature." She added: "I read a lot of literature."
After an audit by a Medicare contractor a year or two ago, she says she closed the office she was using at the time. She said she no longer does most of the 29 diagnostic tests she performed in 2008.
She does still practice medicine—in fact, she works out of at least three offices in two states. And she said she still has patients on Medicare and Medicaid. In the interview, she spoke about a range of treatment options, including one she used in her home country: leeches. She doesn't use leeches here "because of malpractice." In any case, Medicare doesn't have a specific billing code for leeches.
