Philadelphia Reflections

Malpractice: Reported Medical Errors

Wall Street Journal Errors

A recent article in the Wall Street Journal reports the state of Minnesota releasing a survey of 378,544 surgeries performed in the hospitals of that state in a 15-month period. During that time, 99 serious medical errors were found, 21 of which resulted in the death of the patient. One error in about four thousand surgeries, and one death in twenty thousand. Because of an understandable reluctance to self-report mistakes, assume this incidence of negligent errors is a bare minimum level of the true incidence. But arbitrarily doubling that to a rate of one death in ten thousand, would be a safety record that calls the malpractice crisis to account. Minnesota is somewhat special, to be sure, frequently held out as demonstrating the lowest medical costs in the country. So go on, then, say the rest of the country has one death in five thousand surgeries. You might be very risk-averse to feel such an incidence warrants destroying our present medical system, imposing some new one with unknowable risk content. That's especially true if you recognize that average American life expectancy got four months longer during those fifteen months. One way to reduce the number of errors would be to perform fewer surgeries, but there are ways of measuring the harm that would do. It's not ever comfortable to defend any error, but it really is necessary to examine the full consequences of any proposal about them.

Let's extend these examples. Since there were only five fatal medication errors in 400,000 Minnesota surgeries, then fatal medication errors must be impossibly rare. An event that only occurs once in eighty thousand cases must represent very special circumstances when it does happen. If people are of the mindset to take 79,000 successful surgeries for granted while applying the most pejorative scrutiny to one case that was unusual, there is no way statistics can change a mindset that the medical system is riddled with error.

My own reaction to these bare statistics is that if there really was only one death from a medication error, there must have been five hundred near-misses. I would conjecture the persons making these mistakes probably caught them before they got serious most of the time, perhaps four hundred times. And then one or two other people caught it most of the rest of the time, leaving the last few cases to escape by pure luck, and one unlucky person making it through to the statistical report. Over a period of two centuries, the hospital has developed systems for catching errors, and most of the systems depend on redundancy. We in hospitals do almost everything three times, screening out a huge amount of human error under stress. Any efficiency expert worth his stop-watch can see repetition and overlap, redundancy, and\waste. Focused factories, as Professor Herzlinger of Harvard styles them, can easily save money by enforcing a discipline of doing it once, and doing it right the first time. That saves money, and that's not a minor issue. But if we yield too far to this pressure, some of those other five hundred medication errors are likely to prove fatal.

A modern hospital employs several thousand employees, of varying levels of skill and training, with a great deal of employee turnover. An occasional incorrigibly incompetent employee can occasionally do real damage before being identified and dismissed. In recent years professional shortages in critical areas can force some substandard employees to be tolerated longer than they should be. By encouraging longer stretches of employment, a congenial slow-paced environment can reduce the incidence of errors but then look out for those efficiency experts. And crowded, tense anthills of activity must perform all day and all night, weekends and holidays. Emergencies appear out of nowhere, the door to the ambulance entrance banging open to a cluster of shouting excited firemen. They can appear at a moment when every single employee is lashed to a heaving deck of other necessities, or wearily starting out the door at the end of a tumultuous day. Some other pious professor or earnest newspaper columnist offers the non-helpful suggestion that we would perform better if we got more rest. What suggestions might be valuable to reduce stress when the loudspeaker blares, Code Blue?

A Fair Plan for Fire Insurance (and Health Insurance, too?)

Penn Mutual

The Fair Plan (sixth and Chestnut, Philadelphia) is a fire insurance company with unusual features. Some day, it is to be hoped some scholar will write a book about the highly mixed motives of the people who created it, compared with the unexpected ways it did or did not fulfill original expectations, of both its creators and its enemies. The Fair Plan only issues fire insurance on houses, if other insurance companies have turned that house down as a bad risk. Such risky houses would normally draw higher premiums for fire insurance, but the Fair Plan insures these risky houses at normal rates. Therefore, it loses money, which is made up by the other regular fire insurance companies in the state in proportion to the business they do, obviously thus raising the price of fire insurance for everybody. But in this way, Pennsylvania guarantees that everybody can get fire insurance. Is this a good idea? Might this be a way to give health insurance to all those people who can't get health insurance? Let's talk about the Fair Plan.

We'll set aside discussion of whether the Fair Plan was a product of cynical politicians pandering for votes, or whether it was a noble gesture for fairness and equality for our poorer citizens. It very likely had elements of both motives in it, but that doesn't matter any more. It's a form of hidden taxation, of course, and it has the result of making Fire Insurance seem more expensive in Pennsylvania than in other places that do their social work with real taxes. Go too far with that, and people will end up buying their insurance in Bermuda instead of boarded-up former fire insurance companies in Pennsylvania.

As the story is now told, the regular insurance companies had a choice of taking the "substandard" applicants in turn ("Assigned Risk") or creating a new company (Joint Underwriting Association). They decided they preferred the JUA. So a company was formed which specializes in nothing but bad risks, including a few arsonists and other unmentionables, but mostly poor people in bad neighborhoods. If we are ever thinking about following the Fair Plan model in health insurance, it would run the risk of being accused of creating a two-class healthcare system. But no one seems to bring up that rhetoric about fire insurance, primarily because there is comparatively little intrusion of politics in the matter, This system is given orders to spread the extra cost of universal fire insurance out to the policy holders of all fire insurance, and it does it very efficiently, without extending its mandate into setting firefighter wages, running fire departments or repainting scorched woodwork. The fundamental decision was whether to spend Society's money this way. Once that decision is taken, the task is to do it efficiently. Notice, this is not compulsory fire insurance; it is compulsory availability of fire insurance.

After the Fair Plan had been running for ten or so years, a funny thing emerged. There were years when the Fair Plan made a profit! The fire insurance industry had absorbed the Fair Plan into their scheme of things, and felt free to increase the number of applicants they rejected, during years when money was tight or business was bad. If you had compulsory availability of fire insurance, the provision of a company which could not refuse an application made it possible for every other company to refuse when it pleased. When the economy encouraged rejection, a better class of applicant came to the Fair Plan, which made the plan more profitable. When economic conditions reversed, this reversed, and the Fair Plan again lost money. For this reason, the insurance industry is very anxious to prevent the Fair Plan from becoming political, or getting tangled up in worthy but extraneous ventures. And that's probably a good model, too, if we are considering adopting a similar system for the health insurance world: stick to your mission.

Since this simple, tested idea never seems to get into the discussion phase of present agonizing over health insurance for the uninsured, it's one clear sign that such discussions at present are not terribly serious.

Utilization Review

When medical care was entirely a private two-party arrangement, the patient and doctor negotiated what was to be done, and often the price; if the patient could not pay, some embarrassed workaround was figured into the equation, including a vague promise to pay the doctor when he could. A surprising number of patients did pay later, and because the doctor's bill contained very little overhead, even partially paying it off created an unspoken promise that the patient would now be welcomed back.

It was the appearance of insurance forms which created secretarial overhead, even though of course it also brought along some revenue. Skeptics may doubt the extent of this, but at my advanced age as a patient I now visit four doctors of different specialties. Collectively, they seem to have fourteen assistants and fourteen computers I can count. That remains the situation even if a patient delegates decision-making to members of the family, or a hospital wants to know what is being given away to charity patients, or a bewildered patient seeks a second opinion. It changed abruptly when third-party payment seemed to entitle a government or an insurance company to be reassured that claims were legitimate. This uneasy and often resented intrusion into a private matter began to cause friction even in the 1920s, soon after third-party reimbursement made an American appearance, and as physicians gradually recognized that "utilization review" generated more overhead cost than simple claims submission. At first, health insurance companies were restrained from brisk business-like behavior, by the realistic possibility that the patient might switch back to paying bills in cash. Somehow that possibility now seems so remote that when the owner of a Korean auto manufacturing company tried to pay his hospital bill with hundred-dollar bills, the hospital simply had no procedure for handling such an unexpected request.

As long as a secretary was sitting there filling out forms, she might as well answer the phone for appointments. In this way, appointment systems became routine, including hidden extra revenue loss generated by broken appointments. Insurance pre-payment is always complicated by the realities of emergency care, where considerable expense must sometimes be incurred before the third-party has a chance to review the facts; in this lies the origin of the term "reimbursement". The third-party acquired the ability de facto to delay or deny reimbursement, an action immediately regarded as high-handed, since expenditures had already been made in good faith and could not be recovered. The right to deny such claims was implicit, and at first it was used gingerly by insurance companies trying to establish themselves. Now that insurance is considered normal, any incurred costs must be handled in some way, so other subscribers are charged extra to keep the hospital solvent. The cost to the system is exactly the same this way, but no other way has been suggested to maintain the credibility of insurance oversight. When I once told the owner of a department store that I had 2% bad debts in 1955 (ten years before Medicare) he replied that my bad debts were less than his and my bookkeeping was far more elaborate. True, some hospitals could not survive without the insurance system, but if others in better locations examined their experience extensively, they might discover their accident rooms are net losers for the insurance. Third-parties slowly retreated, hospital prices crept upward and the system readjusted, but no one is entirely satisfied with this showmanship masquerading as a balance of terror. When you see a sick or injured person turned away from an emergency department for lack of insurance proof, you can be sure there is either not enough slack in the system, or not enough administrative imagination.

Senator Wallace Bennett

Early in the 1970s, Senator Wallace Bennett of Utah devised a solution to this problem, and persuaded the rest of Congress to endorse a nation-wide system of Professional Review Organizations (PSRO), run by the medical profession but paid for by Medicare. Although many physicians muttered about the "creation of new elites", and hospitals were particularly uneasy about the migration of power, the new system worked reasonably well. If the physicians in the local PSRO approved a service, it was paid for. If payment was denied by a local PSRO, an appeal to a state-wide body of physicians was provided. In that way, prevailing local standards were respected, while appeal to distant physician judges was also provided, to guard against local vendettas ganging up on someone they disliked. Inevitably, a few localities would elect an unsatisfactory PSRO, so a voluntary national organization was formed to educate, inform and defend the process, and better able to discipline it with peer pressure than might be supposed. The national organization was pleasingly benevolent, sometimes apologizing for its role as a necessary one to weed out constituent PSROs whose bad behavior might discredit the others and start up the usual chatter about professional self-government: The foxes were guarding the hen house, and must be replaced by wolves. The PSRO gradually adhered to an unspoken rule of rarely boasting of success; it saw its role as redirecting unsatisfactory behavior, not vanquishing wrong-doers. It soon became evident that successful PSROs were of two types, rural and urban. In localities where a single hospital served a county or more, physicians were of all types mixed together, and the need was to strengthen the emergence of the natural leaders, often by placing them on the PSRO board. In large cities, however, a sorting-out process had usually resulted in strong hospitals and weak ones; birds of a feather flew together. The weaker hospitals were soon identified and urged to elect stronger leadership; sometimes it was necessary to suggest that the institution might serve better as a nursing home. Throughout the process ran the threat of exposure; successful PSROs usually relied far more on peer pressure than the threat of withholding payment.

What ultimately defeated the PSRO was its success, not its failures. Many pre-existing elites were content to see a new governance structure fumble, but felt highly threatened by a successful one. Moreover, the AMA leadership seemed particularly concerned about the direction of "regulatory capture", and was not completely certain whether the PSRO was beginning to dominate the Washington bureaucracy, or the AMA House of Delegates, or a little of both. The PSRO fraternity was definitely a large national organization of doctors who knew and respected each other, and many of its leaders were in the AMA House. Matters finally came to a head in a famous vote opposing the PSRO, by 105 to 100. The opinion of organized medicine meant a great deal to Congress. When representatives of the PSRO next testified before the Senate Finance Committee, the Chairman of the Committee wryly announced the next speaker from the AMA was one who was "presumably here to explain to us the difference between 105 and 100." The PSRO law was soon amended, and insurance review took its place.

AMA House of Delegates

For the time being, utilization review in the future is apparently currently limited to Medicare, under the Medicare Accountable Care Act. With the foregoing bit of history to warn the Accountable Care Organizations about what they are getting into, a simple set of principles can be stated. Utilization review encounters two components of cost: the volume of care, and the price of the services. Doctors control the volume, hospital administrators control the prices. Determining the proper volume of services is a job for practicing physicians, alone and unhampered. To some extent, under the new law an incentive to keep it that way was created by sharing with the provider members a portion of any cost savings associated with the patient group. But the ability to change the sharing will probably be a a one-sided government decision unless the physicians fiercely insist that it remain negotiable. Agreeing that the physician voice should be dominant in the issue of service volume should not imply agreement to exclude their inspection of the prices of services. When physician income becomes linked to their choosing less expensive alternatives, physicians must not continue to be blinded to what the true cost of components truly is. While it remains conceded that hospitals and vendors must retain some flexibility in adjusting the ratio of charges to costs, any deviation from a standard bandwidth must be negotiated with the physicians affected. Past experience should be ample proof of the fairness of this demand.

The volume of most medical services at present is about right, but prices will keep rising when they bear little relation to the underlying direct costs. Furthermore, the Medicare (or similar) Cost Report of every hospital must annually report the ratio of cost to charges, probably by item rather than department. And this ratio must be monitored. Complexity is no argument against doing so; the Chargemaster system has never been resisted because of its complexity. It has been no secret for thirty years that a twenty-dollar aspirin tablet has a very high ratio of charges to its actual cost. Other items are provided at a loss, and the mixtures are highly variable between hospitals. Everyone concerned knows this situation is unsustainable, but everyone recognizes that total denial of reimbursement is entirely too destructive if no slack remains in the system. If it is successful an appeal mechanism will become necessary, because new treatments can be resisted by old treatments, but the optimum rate of change will vary by local situation.

Rather than overcomplicate matters, the goal should be total denial of reimbursement for denied services, because the public will demand it. It must however, be coupled with an adequate profit margin on approved services, to service the costs of retrospective review. Designation of the optimum ratio of cost to charges is a critical decision, much like the function entrusted to the Federal Reserve, of picking the best average interest rates for the national economy. If the agencies selected by Obamacare get this point very wrong, very often, the public must quickly be in a position to let everyone know of its displeasure.

The other source of circumvention is the hospital system of enforcing a physician hierarchy reporting to a physician chieftain, but insisting that the chieftain himself be financially beholden to the hospital administration. In that way, financial health of the institution sometimes dominates the health of the third parties, who then find a way to retaliate. if this issue is neglected a deadlock could affect the health of the patients. Since a three-way balance must be preserved, governance must be fairly shared three ways. Those who believe this is a minor issue because the third party obviously has final power, have a great deal to learn.

Academia Gets Its Share of Blame

In 1910, Abraham Flexner produced a book-long report on reforming medical education, under the sponsorship of the American Medical Association and the funding of the Carnegie Institution. The report was a product of the progressive era of reform which ended the Gilded Age, and can fairly be described as the handbook of a revolution in American medicine. The book had an impact on closing 106 of the 160 medical schools in the United States and Canada. To Flexner that was a disappointment; he thought only 31 were worth saving. To be a good doctor, in his opinion, required between six and eight years of scientific training. The quality of student, as well as the quality of schools, was important, justifying a vast shrinkage of the student body. He liked research and unleashed an avalanche of medical research which subsequently transformed medical care in a hundred ways.

Having said all that, it must be acknowledged, Abe Flexner had been educated in Germany, and his efforts mainly crystallized the transfer of German academic medicine to America, and from here onward to the world. What's being acknowledged is that for a time, German Medicine was far ahead of us, in an environment where we aspired to be the leaders.

Flexner was a charismatic figure, supported by American Medical Association reports from its Council of Medical Education, the wealth of the Carnegie Institution, and the enormous resources of the Rockefeller Foundation. American academic medicine pointed to the example of the Johns Hopkins Medical School, formed before the ferment of pre-World War I, and now being given prominence by Teddy Roosevelt, Woodrow Wilson and the rest of the muckraking era who aspired to replace the Gilded Age with something better. America was aching to take charge of something, had great gobs of philanthropic money at its disposal, and the wind in its sails. Abraham Flexner was a towering figure, all right, but he was not a physician, an academic, or a scientist. He was more the agent of revolution than its originator, ending up as the patron saint of several political factions, individually fighting academic wars with each other. He was, in short, a rain-maker.

Abe Flexner, the Rainmaker

German Model for Medical Schools

Abe Flexner and his book aimed to reduce the number of medical schools from 160 to 31, demanding the survivors look like German schools. This was not so strange; like Japan and China today competing with America, Flexner wanted to compete with Germany more than he wanted to worship it. The students of his dreams were required to be college graduates, the curriculum was to be four years along with two years of science followed by two years of hands-on learning. If the school could afford it, the professors would be on a full-time salary, free of any need to practice in order to make a living. On this last point, many begs to differ, arguing you have to get on a boat to learn how to sail one. There were many echoes of the ferment to come, in the upheavals of the 1960s, and the Humboldt's University of Berlin was the source of quite a few, in both cases.

Immediately there arose a town-and-gown competition. The gown group had the point they could afford the time to do some research. The town group had the point you couldn't teach doctors if you weren't one, yourself. The distinction is probably best understood by comparing medical students with law students. Both were originally taught by practitioners; schools were late arrivals. Many lawyers have remarked that a law school graduate now knows almost nothing about the practice of law until he joins a firm which teaches it to him. A medical student (post-Flexner) can do a pretty good job with most medical problems, the day he graduates. The practicing physicians have retreated into specialty training; a doctor has trouble becoming a specialist if he didn't have the right residency. The rest of the awe-inspiring march of medical progress in the Twentieth century, is the consequence of pouring unbelievable amounts of money into the system, thereby attracting a glittering array of talented medical students. Some of the talents are unbelievable. My own medical school, as an example, has a symphony orchestra made up of students in their spare time, performing on a truly professional level for their own amusement. In a sense, all of this was due to Flexner. In another sense, Flexner himself had little to do with originating it.

Now, forget the yellow journalism, or muckraking, quality of what I am about to relate. A book was recently published relating that a particular medical school was able to support its operations without touching a penny of medical student tuition for ten consecutive years. Instead, the tuition money was transferred to the University's undergraduate schools, and the medical school subsisted on research and other funds. The undergraduates continue to protest about rich doctors, while the medical students complain about going a hundred thousand into debt -- to pay their tuition. But forget that part. The most undesirable situation it reflects is that for ten years, the school was totally able to exclude any parent and student influence in an area that unfortunately has a growing power over events, the school's finances. If the students, families, and alumni of a university have no power to influence its decisions, who do have such power?

George Schaefer:

When the City of Philadelphia found itself with 8000 houses abandoned for taxes, it set up a program to get them back onto the tax rolls, and provide low-income housing as well. The idea was for the city to give the house or sell it at a low price to someone who would rehabilitate it and bring it into conformity with the building code. "Rehab and bring it up to code."

Unfortunately, the process required the owner to satisfy the Bureau of Licensing and Inspections, and that usually required a great emphasis on the washbasins. The process was as follows. A number of twenty dollar bills would be placed in an envelope and scotch taped underneath a wash basin. The inspector was asked to pay particular attention to the underside of the wash basins. As he left, he was asked if he found the washbasins to be satisfactory.

Well, reputable contractors mostly wouldn't touch this sort of thing. It was George's impression that new construction did not go through this process, but rather involved a licensed architect certifying that the code had been followed. Just why this process is less corrupt is not clear, but there may be reasons.