Everybody hates HMOs. They're inflexible, they deny care, they turn medical professionals into bean counters. Conceived decades ago as a way to focus on preventing illness, they now dump the sick.
But is it possible that everybody's wrong? Could it be that what's wrecking American health care isn't HMOs but the profit motive driving many of them?
Few people ask this radical-sounding question these days. Fewer still try to answer it. On July 14 the Chicago-based Journal of the American Medical Association published an article by four doctors arguing that profit is indeed the problem. The article was newsworthy because authors David Himmelstein, Steffie Woolhandler, Ida Hellander, and Sidney Wolfe used data to make their case, not rhetoric.
They got their numbers from an industry group, the National Committee for Quality Assurance. Every year the committee gathers information from HMOs around the country in a standardized format known as the Health Plan Employer Data and Information Set, familiarly known as HEDIS. Himmelstein et al analyzed data submitted by HMOs in 1997 covering the year 1996. For that year HMOs were asked to track 14 indicators of good health care, ranging from preventive measures to the treatment of serious illnesses. How many 2-year-olds and 13-year-olds enrolled in each HMO had received all their immunizations? How many women aged 52 to 69 had had mammograms in the preceding two years? How many heart-attack sufferers had got beta-blockers afterward? How many diabetics had had eye exams in the past year? How many mental patients had got outpatient follow-up within 30 days of being discharged from the hospital? And so on.
You might expect that for-profit HMOs would give better care. After all, competition between profit-seeking retailers serves the common good in other walks of life. It gives us plenty of nutritious food, ever-cheaper computers, and--once the government put its foot down--cleaner cars far cheaper than anticipated. The investors may get rich but we all end up better off.
On second thought, you might expect that for-profit and not-for-profit HMOs would act about the same. As several economists told me, they all operate in the same regulated environment. The University of Chicago's Richard Epstein, author of Mortal Peril: Our Inalienable Right to Health Care? wrote in an E-mail interview, "No health organization can afford to operate at a loss, so the nonprofits have some strong market incentives....As a theoretical matter there is no clear priority between regulated charities and regulated for profit organizations."
But the numbers didn't come out that way. The JAMA article shows that for-profit HMOs gave lower-quality care than nonprofits in all 14 ways measured. Of two-year-olds in not-for-profits, 72 percent had had all their shots, compared with 64 percent in investor-owned HMOs. Of women in their 50s and 60s in not-for-profits, 75 percent had had mammograms in the previous two years, compared with 69 percent in for-profits. Of heart patients in not-for-profits, 71 percent had been given beta-blockers, compared with 59 percent in for-profits. Of diabetics in not-for-profits, 48 percent had had eye exams, compared with just 35 percent in for-profits.
All but one of the 14 disparities are statistically significant--i.e., very unlikely to be due to chance. They're medically significant too, according to the JAMA authors. "If all 23.7 million American women between ages 50 and 69 years were enrolled in investor-owned, rather than not-for-profit plans, an estimated 5925 additional breast cancer deaths would be expected," because fewer women would have had their mammograms.
These findings may well understate the problem--because the indicators of quality used are easy to "game." The cheapest way for an HMO to improve its rating isn't to give better care, but to work on the things that are being measured. "For instance," the authors write, "HMO administrators may push clinicians to increase mammography rates, but deny them the time needed to perform optimal clinical breast examinations, patient education, or other clinical activities that HEDIS does not measure." For this reason, said coauthor Steffie Woolhandler in an interview, "These indicators work best the first time you use them."
The authors conclude their article emphatically: "The decade-old experiment with market medicine is a failure. The drive for profit is compromising the quality of care, the number of uninsured persons is increasing, those with insurance are increasingly dissatisfied, bureaucracy is proliferating, and costs are again rapidly escalating. We believe national health insurance deserves a second look."
Of course, the four authors didn't move from a state of indecision to advocacy of national health insurance in the course of conducting this study. They had held this belief before. Himmelstein and Woolhandler had proposed a national plan in the New England Journal of Medicine ten years ago. Wolfe directs the Nader-founded Public Citizen Health Research Group. Hellander is executive director of Physicians for a National Health Plan, based in a small office on South Michigan Avenue.
But none of that makes their analysis wrong. Their most cogent critic, health-services-management professor David Dranove of Northwestern University's Kellogg Graduate School of Management, doesn't resort to cheap shots. He sympathizes with their concerns but has problems with their data and methods. He points out that the data they used didn't allow them to control for important factors such as how large the HMOs were, how long they'd been in existence, or who their patients were. So it's conceivable that these factors make for-profits look worse than they are. His logic is that if for-profit HMOs happen to have more hard-to-serve members (people who don't see the point of prevention), then they might score worse for that reason rather than because they're investor owned.
The JAMA authors acknowledged this possibility in their article, but they didn't tone down its conclusions. Dranove's argument, after all, is hypothetical. No one knows whether for-profit HMOs have a different clientele. And Dranove can't point to studies contradicting the JAMA article. (Meanwhile, new research has found fault with for-profit companies in another area of medicine. On November 25 the New England Journal of Medicine published a multiyear study of more than 3,000 kidney-dialysis patients. Those being served in for-profit dialysis centers were less likely to be put on a waiting list for a kidney transplant and were 20 percent more likely to die.)
JAMA author Ida Hellander--who broke an elbow in March and is still trying to get her for-profit HMO to pay the emergency-room bill--says that written responses sent to JAMA didn't refute the article's fundamentals. (The letters haven't been published yet.) One HMO official, she says, wrote that the market had just begun to understand quality health care and hadn't had an adequate opportunity to respond rationally. Hellander isn't buying it. "The HMOs say they'll get better as the market matures. But we don't see things getting better. Physicians would tell you things have only gotten worse."
Regardless of what the JAMA numbers mean exactly, they highlight the uncomfortable tension that has always existed in medicine between professionalism and getting paid. A generation or two ago, most doctors worked alone and were paid a fee for services rendered. When in doubt, they had a financial incentive to overtreat patients (more services equal more fees). Cost played little or no role in their thinking. The earliest managed-care programs were community-based cooperatives that sought something better, like the Group Health Cooperative of Puget Sound, founded in 1945. In essence, a local group of patients would hire a group of doctors on salary. Such a system reduced doctors' incentive to overtreat. More important, it encouraged everyone to save money through preventive medicine, patient education, and cooperation among doctors to deliver the best care.
Such alternatives were few--and repeatedly vilified by mainstream medical professionals as "socialized medicine." But in the early 1970s, as employers and government sought to control the rising costs of fee-for-service medicine, managed-care co-ops were Republicanized and reinvented as HMOs. HMOs could be for-profit or not-for-profit, cooperative minded or bottom-line minded, but the balance gradually began to shift toward the bottom line. Historian Paul Starr writes in his book The Social Transformation of American Medicine, "The socialized medicine of one era had become the corporate reform of the next." Corporate reformers seized on a third way to pay doctors--by "capitation," a flat fee per patient per year. If the patient's care costs more than the flat fee, the excess comes out of the doctor's pocket.
The fee-for-service system had encouraged overtreatment in some cases, and the salary system was more or less neutral. Capitation encouraged undertreatment in some cases. Gordon Schiff, a doctor at Cook County Hospital, painted one example in a recent interview: "In a lot of HMOs you [the doctor] get $500 to take care of a patient. If the cost is less, you make money. And if the cost is more, it costs you. This greatly complicates an already uncertain process, such as, do you [the patient] need a CAT scan for a headache? In most cases you don't. But if I make money when you don't get a CAT scan, and if you know that I've got a kid starting college, then it's hard for you to trust my recommendation." That loss of trust strikes at the heart of medical professionalism, because a professional by definition is (among other things) someone you trust to put your interest above her own.
Even though medical professionals are now in a nasty bind, it's important to remember that they haven't been saints throughout this history. When doctors' fee-for-service arrangements encouraged them to overtreat patients (and to be overpaid), there was nothing like the current outcry against the market's debasement of professional ethics. Now that doctors' market arrangements through HMOs encourage undertreatment, they're protesting that professionalism is being pitted against the market. But in both cases market incentives encourage doctors to act unprofessionally.
As HMOs have grown, for-profits have been edging out not-for-profits. In 1985 for-profits had 26 percent of all HMO members; by 1998 they had 62 percent. Not only are for-profits taking over, but the old socially minded HMOs (which may or may not be not-for-profit) are under pressure to act like businesses--otherwise large employers will drop them for being too expensive. What worries the JAMA authors is that over the long haul Richard Epstein may be right after all about where the health-care world is moving. In the near future it may no longer be possible to detect a difference between for-profit and not-for-profit HMOs. The disparity in care that the authors found may vanish--not because for-profits have become more conscientious but because they've dragged everyone down to their level. Hellander describes what Puget Sound has become: "They signed up 750,000 senior citizens in the Spokane area, and now they're dumping them because they aren't making money on them. Now it has to be driven by the bottom line. That would've never happened ten years ago in group health. Of course back then they weren't impelled to expand--they were serving the people who were on their board."
Why has it been so hard to make the market work for everyone's benefit in health care? For one thing, says Woolhandler, there's not that much competition between HMOs: "At best you have oligopolies," a small number of companies that carve up the market among themselves. And even when HMOs do compete, the buyers in the marketplace are employers, not individuals. In theory your employer would like you to be healthy, but in practice what your employer shops for may not be what you would shop for. So you rarely have a chance to send a signal to the marketplace by choosing United instead of Humana or vice versa--your bosses already sent their own signal.
In economic theory, HMOs with good quality ratings would gain customers--and those providing low-quality service would lose them. But that isn't happening, and we don't need to depend on national-health-insurance advocates to tell us so--we can see it documented in a study released November 30 by the Schaumburg-based Society of Actuaries (the report is available on the Web at www.soa.org/library/mancare.htm and soon will be published in the North American Actuarial Journal). Authors Dennis Scanlon of Penn State and Michael Chernew of the University of Michigan write, "There is currently little evidence that consumers use such information in the manner hypothesized."
But the problem runs deeper than the absence of a truly competitive market. Would we even want to have one in medicine? In his 1997 book Everything for Sale: The Virtues and Limits of Markets, Robert Kuttner looked at health care from the individual consumer's side. Markets make winners and losers, he wrote. If you get a good job or get lucky in the stock market you can buy a BMW; your friend who doesn't do so well will make do with an old Nissan or the CTA. Most of us can live with that most of the time. But what if your less successful friend loses a breast or a baby because she couldn't get preventive care?
In health care we just don't like the idea of winners and losers. Kuttner cites a Harris poll from the 1980s in which 91 percent of respondents agreed that "everybody should have the right to the best possible health care--as good as a millionaire gets." Obviously that's impossible--there would be no way to pay for it. But it's a clear sign that most Americans don't want health care to be determined by ability to pay. Nor would it be a wise idea even if they did. As long as TB can find its way into gated communities, some basic level of health care for all is a public good. And, adds Woolhandler tartly, "No nation has ever achieved universal health care through markets. Markets aren't there to provide universal anything."
Woolhandler sees HMO market medicine as a social experiment gone wrong. "When you try a new therapy, by random events it may not look good at first. Later it can turn out to be OK. But you always do define a point--hopefully you do this ahead of time--where the experiment is considered to have failed." Beyond that point it would be malpractice to endanger the patients any longer.
Art accompanying story in printed newspaper (not available in this archive): illustration/Kurt Mitchell.