I have received a fair amount of criticism for my recent oped “Universal Coverage Kills,” which appeared at National Review Online and in the Orange County Register: Fun excerpts include:

• A philosophically sympathetic eICU medical director emails that I am “off the mark.”
• A “free-market believer and an attorney” emails that he is “upset about the increasingly anti-physician rhetoric of Mr. Cannon. It is extremely biased and counter-productive to the free market…The DOCTORS are the ultimate producers of health care. Get off their backs.”
• A “libertarian doctor” emails that I “insult our intelligence with well-meaning but fundamentally obtuse, ignorant opinions. If you don’t know, and you clearly don’t know, ask a practicing doctor.”

Moving beyond my philosophical allies…

• In a letter to the editor, an Orange County pediatrician writes, “Cannon’s preaching from his high horse that hospitals and doctors are ‘rewarded’ for medical errors is irresponsible and wrong. He needs to walk in our shoes and, perhaps, he will be less cynical and be more grateful that we, as a medical profession, are here to heal with our hearts and souls, to cure with our mind and body and to always do no harm.”

Bloggers have also offered their two cents:

• WhiteCoat writes, “It amuses me when people possessing little knowledge of the inner workings of the practice of medicine write articles as if they are ‘in the know.’ Mr. Cannon’s article is one such work…there are several statements Mr. Cannon makes that are either purposely inflammatory or that show a fundamental lack of insight…I’m glad he links to a few statistics, but the conclusions he comes up with are flat out wrong.”
• Ezra Klein mocks as usual, yet (damn him!) fails to provide any juicy quotes.

So what’s all the hubub about? I claimed that America’s grand experiment with universal coverage – Medicare – has stifled innovations that reduce medical errors, and has thus caused unnecessary deaths.

Here’s my argument:

One way to discourage people from injuring each other is to force Smith to bear the cost of any injuries he inflicts on Jones. That’s the idea behind tort law (including medical malpractice liability), even if it sometimes doesn’t work very well (such as with medical malpractice liability).

Medicare tends not to do that. It generally pays health-care providers on a fee-for-service basis. That means that if a provider makes a mistake that injures a patient and the patient requires follow-up care, Medicare will pay for both the service that caused the injury and the follow-up services. Medicare’s payment system thus rewards medical errors with extra payments. The Medicare Payment Advisory Commission agrees: “At times providers are paid even more when quality is worse, such as when complications occur as the result of error.”

Medicare recently admitted that this is a problem, and has attempted a remedy. Medicare announced that as of October 2008, it will no longer pay for:

• Hospital services attendant to any of a list of “hospital-acquired conditions,” such as catheter-associated urinary tract infections, or
• Hospital or physician services attendant to “never events” such as operating on the wrong body part, operating on the wrong patient, or performing the wrong surgery on a patient.

In terms of using financial incentives to discourage medical errors, however, Medicare is late to the party. Markets have long since developed another payment system that forces providers to bear 100 percent of the (financial) costs of any medical errors.

Prepayment gives the provider a flat amount of money per patient. When combined with an integrated delivery system to form a “prepaid group plan,” where all the hospitals and doctors work for the same entity, prepayment places the cost of avoidable medical errors squarely on providers. Following a medical error, the providers themselves have to pay the cost of any additional care. Under fee-for-service payment, providers lose money if they reduce medical errors; under prepayment, providers profit by reducing errors.

That may explain why prepaid group plans are ahead of the rest of the market when it comes to deploying electronic medical records and other error-reduction efforts. As Lucian Leape and Donald Berwick note:

Several large, integrated health care systems, notably Kaiser Permanente, Ascension, and [yes] the Veteran’s Health Administration, have been leaders in implementing new safe policies and practices.

According to Paul Starr, prepaid group plans have been around since at least 1929, though the largest and best-known (Kaiser Permanente) dates to the 1940s.

(Prepayment is not perfect, of course. Providers get to keep whatever money they don’t spend on medical care, which creates obvious incentives to skimp. Evidence from and since the RAND Health Insurance Experiment, however, suggests that patients in prepaid group plans have fared no worse than other patients.)

So despite the fact that alternative payment systems have been available since 1929, it took that lumbering giant Medicare (created in 1965) more than 40 years to notice and then attempt to remedy the fact that its payment system rewards providers for medical errors. When pressed, some members of the Church of Universal Coverage will admit that government control of health care will lead to less medical innovation – i.e., fewer cool new drugs and medical devices. They less often admit that government suppresses innovation in the financing and delivery of medical care, even though that can be just as harmful. Indeed, considering that Medicare is the largest purchaser of medical care in the nation – so large private insurers tend to ape its every move – Medicare bears responsibility for an awful lot of deaths due to medical error.

Here’s what my critics had to say:

A couple of doctors argued I should not cite the Institute of Medicine’s estimate that medical errors kill up to 100,000 patients in U.S. hospitals every year. I am aware of the problems with the IOM’s method of identifying medical errors, and have written about those problems previously. Unfortunately, I’m not aware of a better estimate, and my critics do not offer one.

More important, the precise number of deaths due to medical error is beside the point. If we can agree that medical practice responds to financial incentives, then I’m sure we can agree that Medicare’s payment system has made providers less vigilant about avoiding medical errors than they would be in a market free to experiment with different payment systems.

Mostly, however, the doctors argued against Medicare’s new policy. Some illustrative excerpts:

“Some of these complications, such as wrong-site surgery, incompatible transfusion, and air emboli, are admittedly indefensible. However, others, such as pressure ulcers, ventilator-associated pneumonia, and catheter-associated urinary tract infections, are not necessarily reflective of poor care.”

“There is an irreducible minimum for post-operative infections…It is impossible to prevent all surgical wound infections.”

“While it certainly sounds reasonable to not pay for care that is the result of a patient injury, some injuries such as a post-surgical wound infection or bed sores, cannot be prevented in all patients. You can reduce the risk, but even with what is considered optimum care, bad things can happen because every patient is different and how the patient reacts to some insult to his or her body is different.

“When payment is denied for care for these type of complications, what will eventually happen is that the no physician or hospital will desire to care for the sickest patients…What occurs in those situations, is that the physician and the hospital is being punished for the patient’s poor health.”

The problem with those arguments against Medicare’s new payment policy (or any type of prepayment) is that they assume there is no room for improvement in error reduction – despite considerable evidence to the contrary.

1. There may be “an irreducible minimum for post-operative infections.” But how do we know what that minimum really is? If all hospitals have an infection rate of 5 percent and some rogue hospital cuts its rate to 1 percent, Medicare’s old policy would penalize that hospital. That seems more like a recipe for protecting incumbents than reducing medical errors.
2. Medicare’s new policy may encourage providers to avoid doing some procedures on subgroups of patients when the cost of those procedures (including the expected cost of complications) exceeds the payment. But the new policy also creates profit opportunities for providers who can reduce that expected cost by reducing the probability of infection within that subgroup.

Ultimately, those arguments tell us nothing about whether Medicare’s old payment policy is better than the new one. Prepayment creates incentives for providers to skimp and to avoid high-risk patients, but it also creates financial rewards for innovations that enhance patient safety. Fee-for-service payment may be better for patients, or it may be the last refuge of a scoundrel who just isn’t trying hard enough to improve patient outcomes.

More important, those arguments are also completely beside the point. The point is not whether Medicare’s new payment policy is better than the old one. We need experiments with different payment policies to see which produce the best outcomes for patients, and the rigidity that government brings to that process is downright harmful.

WhiteCoat, an ER doc, makes two interesting observations. First, he accuses me of “advocating a national HMO.” Yeah, that’s so me. Second, he writes:

if people like Mr. Cannon and our beloved government are so sure that all of these “errors” are preventable, then provide all of us overpaid brain dead doctors with a way to prevent the errors, then. Give me some links to those articles, there, Hippocrates. Put up or shut up.

I accept WhiteCoat’s challenge and draw his attention to the “About” page of his own blog, where he describes himself as having “undiagnosed ADD”:

I can literally be eating lunch, talking on the phone, admitting a patient and typing an entry on this blog all at the same time.

So here’s my advice on how to avoid medical errors. Put down the sandwich, hang up the phone, step away from the blog, and pay attention to your patient.

The Orange County pediatrician writes, “we do not allow surgical sites to get infected, air bubbles to enter a patient’s bloodstream or operate on the wrong body part so we can get ‘rewarded’ with extra fees.” As if I suggested such a thing.

And finally, Ezra Klein, who gets a little confused about dates and payment systems. Klein apparently thinks the market contributed no solutions to this problem until a couple of years ago when the Minnesota HMO HealthPartners started refusing to pay hospitals for “never events.” Surely, Klein would agree with Paul Starr that 1929 saw not just the creation of Blue Cross, but also the creation of the Ross-Loos Clinic and the nation’s first medical cooperative in Oklahoma, both of which were prepaid group practices. Surely, Klein would also agree that markets had therefore developed, by 1929, a payment system that forces providers to bear the financial costs of their errors. (But I should thank Klein for helpfully pointing out that Medicare’s new policy was also developed first by the private sector.)

Yet Klein also argues that government beats markets at finding the best payment system:

The vast majority of doctors are private. The vast majority of doctors remain on fee-for-service plans. Kaiser Permanente and Group Health Cooperative have not pioneered a new norm of salaried physicians. Not even close. Meanwhile, the Veteran’s Administration, which is the country’s largest socialized health care system, has its doctors on salary. Indeed, as the VA demonstrates, when government takes over health care, it salaries its physicians. So do most other socialized systems around the world. When government simply pays for health care on the private market, as with Medicare, it does not salary the physicians because it does not employ them. Instead, it pays them under private market norms. Weird how that works – it’s almost as if the market, where physicians retain their fee-for-service status, is failing us, but socialized systems worldwide have figured out the salary thing.

Again, Klein is confused. First, prepayment is not the same thing as having physicians on salary. For example, Kaiser Permanente is offering large employers access to its hospitals and medical group on a fee-for-service basis, yet physicians within that group still draw a salary. The distinction is important.

Second, Medicare is not a market-driven health care system. (Heads up: when health-policy wonks tell of the elderly woman who said, “Tell the government to keep its hands off my Medicare,” that’s a laugh line.) Medicare isn’t just government-run, it is socialized medicine.

Third, it’s a little odd to attribute the dominance of fee-for-service payment to market forces when one considers all that government has done to inhibit prepayment. Everything from Medicare, to medical licensure, to insurance-licensing laws, to corporate-practice-of-medicine laws, to laws prohibiting prepayment, to ERISA, to the tax exclusion for employer-sponsored health insurance premiums has inhibited the growth of prepayment broadly and prepaid group plans in particular. It’s almost as if the physicians have used the law – and their useful compatriots on the Left – to block competition from better ways of financing and delivering medical care. And the Left then responds by lamenting the market’s lack of progress.

I repeat my prediction: Ezra Klein will die a libertarian.