Bad Medicine

I have, apparently, had a comment (innocuous, I thought) deleted at Catron’s site. He’s up in arms over a piece in Slate by Jesse Pines and Zach Meisel* that suggests that hospital administrators have a financial incentive to bed insured patients coming from clinics over uninsured ED walk-ins, and that a good solution would to create performance incentives that reward short wait times for ED patients.

Here’s him:

The authors of this disgraceful piece of agitprop would have their readers believe that the people who run hospitals deliberately allow people to languish in their ERs for financial gain. The suggestion is not merely slanderous. It is absurd on its face.
I have worked in hospital finance (at institutions large and small) for more than two decades and have never met an administrator or finance person (not one) to whom such an idea would even occur. Where I have worked, suggesting such a policy would get you fired.

As a free-marketeer, Catron could have chosen to point out competing incentives that the Slate authors ignore (for example, a commenter points out that diverting ambulance traffic when the ED is overfilled is costly), or explain why rewarding shorter wait times would be counterproductive. In any case, the relevant response would have contained data or logic, rather than moral outrage.

* Note: I have worked with Jesse Pines in the clinical setting and greatly respect his academic work

**UPDATE**
A commenter at kevinmd writes:

Their hospital sounds similar to my urban academic center. No one’s intentionally stated that the goal is to shaft the urban poor in favor of the well-insured, but the policies are set up to accomplish the same thing. Patients routinely wait 48 hours in the ER for floor beds after being admitted (tying up ER beds while up to 50 patients sit in the waiting room) while direct transfers from suburban hospitals arrive on the floor. We constantly hear about how the hospital hopes to “improve the payor mix,” i.e. that they’d rather take the well-insured suburban patients over the Medicaid/insured locals who come in via the ER.”

Lesson: you must be well-informed to make accusations of slander. White Coat calls the Slate piece a “conspiracy theory,” an accusation which I think indicates economic illiteracy. As I’ve said here in the comments already, it’s all about the incentives, baby. Beyond that, though, White Coat has a shaky grasp of the power of small differences in probability to make huge differences in profit (don’t let that man into Vegas). And, a final (and perhaps clearer) restatement of my point, to attack the Slate piece without addressing why hospital administrators don’t explicitly reward shorter ED wait times (which everyone has done) is a non-sequitur. Meanwhile, back at kevinmd, dermatologists are accused of cherry picking their patients for financial gain and nobody seems to mind. Hospital CEOs don’t face the same choices?

Borges thought it a fantastic nightmare, but it turns out that a perfect autobiographical memory just nets you second place on Jeopardy and an inability to forget family vacations.

I’m off to sub-Saharan Africa tomorrow to practice medicine in a way they’d never let me do here in the states (uh, until next month). I may try to write something about it here and there, but there are probably better blogs out there for that kind of stuff already. Tamara Todd’s blog, for example, is interesting to me for the meta-narrative.

Addendum: Some pre-trip reading on the nature of medical imperialism.

There’s uproar amongst opponents of single payer health care over a recent survey published in the Annals of Internal Medicine (letters section, thus avoiding peer review) suggesting that a majority of doctors (59%) support “universal health care.” Critics point out, rightly, that the survey (which consisted of two relatively straightforward questions and unknown accompanying material) was sent to members of the AMA, which is pro-universal health care but not pro-single payer and perhaps not a representative sample of physicians across the board, and that just over half of the surveys sent out were returned, which opens the door for serious selection / response bias. And certainly, the media coverage seems a bit much.

The authors address some of the methodological concerns in a similar study conducted five years ago (opens in pdf format), to which this survey is intended as a follow-up. There’s no mention in the current one of whether the trend towards more support is significant or even amenable to statistical analysis, but the clear implication of publication is that it is meaningful.

More interesting to me than the merits of the survey itself (my opinion is that it’s relatively worthless) are some of the comments by critics that the data is untrustworthy simply because its author, Aaron Carroll, is a member of Physicians for a National Health Program. As best I can tell, Catron goes so far as to lie about Carroll’s credentials.

The lead author of this “survey” is Aaron E. Carroll, a single-payer zealot on the Board of the activist group “Physicians for a National Health Program.” This guy routinely produces “studies” and “surveys” that somehow always show that Americans in general and the medical community in particular want government-run health care.

I will note that a pubmed search for “carroll ae” turned up only the two surveys already mentioned, and Catron only links to a pro-single payer op-ed written for an Indiana newspaper.

Greg Scandlen, who appears to be much more reasonable, still manages to startle.

The letter was written by Aaron Carroll, MD and Ronald Ackerman, MD, both of the Indiana University School of Medicine. I don’t know about Dr. Ackerman, but Dr. Carroll is a member of the board of directors of Physicians for a National Health Program (PNHP), so is hardly an unbiased researcher. Interestingly, the Annals requires the disclosure of financial conflicts of interest, but not political conflicts or biases.

As a fan of overcoming bias of all types, I wholeheartedly support the idea that scientists should try to perform research which disproves their own hypotheses, and I think the restriction of conflict of interest disclosure to direct financial matters absurd (”Dr. Blue discloses that his sense of self worth depends on the positive results presented in this study”) . I doubt that we could trust many papers if all potential conflicts were disqualifying — how many papers in support of HSAs are written by people who don’t support HSAs? Pragmatically, it’s hard enough to get people to disclose their financial conflicts, and it seems that disclosure of political conflicts or personal biases is a nebulous concept that would be impossible to enforce. What biases are relevant? (”Ok, well he’s not really for PNHP, but he did vote for Clinton in the democratic primary”). Interestingly, in papers I could find on pubmed, Scandlen is listed as affiliated with “Consumers for Health Care Choices,” but no disclosure of his own personal biases or political views are made explicit.

Update: Don McCanne of the PNHP comments on Healthcare BS to say that (1) the article was peer reviewed (don’t know the Annals letter policy), and (2) that the doctors surveyed were selected from AMA masterfile, not member file.

If you’re wondering where I’ve been, you must be an odd and lonely person. Anyways, taking my last test before graduation next month, then flying off to darkest Africa to play doctor one last time before, uh, the real thing. Oh, and I find out where I’m going to be spending my next three (+?) years this Thursday. Indeedy.

Anyways, between the studying and the packing and the intense anxiety about the future, blogging seems a poor investment of my time. But what the hell.

So, anyways, I’ve been thinking that people who complain about overdiagnosis aren’t complaining about overdiagnosis, but that our diagnostic categories and therapeutic options are crude and impractical. There’s no intrinsic difference between mastering your emotions through force of will or force of pill, and anyone who tells you different just hasn’t had the right pills. Because they don’t exist.

Similarly, a common complaint about modern medicine you hear from the holistic types is that it treats the symptom, not the disease. This isn’t entirely accurate, but who cares? Without symptoms, there is no disease. The problem with modern medicine is that it treats symptoms poorly, and with gross side effects.

This video is pretty neat. Found through TGGP.

The progress of medical science is plagued with embarassment. Recently, the intensive blood-sugar lowering arm of the ACCORD trial was halted due to an increase in mortality, primarily cardiovascular. Prior to that, Avandia, a commonly used oral hypoglycemic, was temporarily pulled from the market after it, too, was shown to increase cardiovascular mortality. Tolbutamide, an older older hypoglycemic, was found in 1970 to increase cardiovascular mortality, and one wonders whether it is a more general phenomenon than a particular class effect (tolbutamide belongs to the class “sulfonurea,” Avandia to the class “thiazoladinedione;” the increase in cardiovascular mortality in the ACCORD trial was not attributable to the increased usage of Avandia in that group). In any case, the two more recent studies come as something of a surprise.

Another recent surprise involved Vytorin, a combination of simvastatin, a drug lowers that cholesterol by altering liver metabolism, and ezetimibe, which blocks the absorption of cholesterol in the stomach. The ENHANCE trial showed that the drug does in fact lower LDL (”bad”) cholesterol levels and decrease the size of atherosclerotic plaques, as current theory would predict. But the drug did not lower overall mortality compared to simvastatin alone, and in fact increased mortality from stroke (which theory would not predict).

And a last study, not so much a surprise, but an embarassment nonetheless, showed that systematic bias towards the publishing of positive anti-depressant trials and the squelching negative ones has led to vastly overstated effects for those drugs.

These studies have certain flaws that are starting to be addressed by the clinical trials registry. For instance, to be published in the New England Journal or JAMA, trials must be registered, industry or government sponsorship declared, and outcomes measures defined ex ante to prevent publication bias (anti-depressants) and the post hoc tampering with measures (vytorin, though the attempt was abandoned). These do not prevent bias from creeping in by any means, but they are a start.

But there is another problem with which the clinical trials registry can help, and that is the problem of doctors’ innumeracy. In particular, we are not very good at determining whether a given study is true or false. One problem is overreliance on p-values (for the untrained, the p value represents the frequency with which you would get a particular result even if the hypothesis you are testing is not true): we are trained to use this as a shortcut around actual analysis of a study; p values less than .05 are “true” or “proven” or “significant”, while those above are not. A better method would be to use Bayes’ theorem to interpret clinical tests. But Bayes’ theorem is hard for most clinicians to work with, not because they are stupid, but because of the difficulty in finding an agreeable prior (a subjective estimate of a study’s a priori chances of being true). Most studies test hypotheses generated by a vast body of previous research, and those hypotheses are held with a certain degree of certainty by experimenters, but that information is inaccessible to the non-expert. For a generalist doctor or specialist who hopes to interpret a study outside of his domain of expertise quickly and cheaping, the difficulty of reasonably estimating priors make Bayesian interpretation of the literature unworkable.

Prediction markets, however, are good at quickly aggregating expert opinion (much better than doctors), and can be easily designed to produce probability estimates, as they are in the presidential markets. So a prediction market set up to buy shares of a given trial in the registry outcome being positive or negative with a payoff of $100 if that outcome is met would provide an accurate estimation of prior probability that is easily retrieved by the non-expert. This would eliminate a big barrier preventing most doctors from using Bayes theorem to arrive at a quicker, better estimation of a study’s truth value. So, why isn’t it being done?

Finally getting around to reading Prescription for a Healthy Nation by Tom Farley and Deborah Cohen. The first chapter details how we overspend on healthcare, while confusing ourselves that we are spending on health. The two are not synonymous, as many, many, many others make clear. Farley and Cohen cite estimates that medical errors kill 50-100,000 persons per year, whereas iatrogenic deaths that result from appropriate medical care kill upwards of 250,000. Amongst these statistics are fun anecdotes like this one:

One of the more heavily publicized studies on this followed doctors’ strikes in 1976. Over five weeks, in a dispute about malpractice insurance, about half of the doctors in Los Angeles County cut back on services, and because anesthesiologists were the most militant, hospitals cancelled most elective surgery. For those five weeks no surgeons were doing non-emergency hysterectomies, tubal ligations, knee repairs, or face-lifts. As much as the strike must have caused panic in people who could not see their doctors, researchers afterward found that it actually prevented more deaths than it caused.

To some degree, the medical community has reacted to this information by moving towards evidence-based medicine; furthermore, physicians are coming around to the idea that price-transparent rationing is necessary for any rational health care system, universal or not. Nevertheless, while questions like this are batted around amongst economists and doctors with an interest in policy, my informal sense (talking to fellow students, and judging by the hubris of most of the doctors I’ve actually met) is that most physicians grossly overestimate the amount of good that they do (even if we ignore any one doctor’s place at the margin).

Why is this? One reason is that it’s easy for us to remember the patients we help, the ones for whom prospects were grim but recovery was profound, or even those whose life was never in danger but we helped recover more quickly or pass the illness more comfortably. On the other hand, amongst those 250,000 who died due to iatrogenic causes, a great many would have died anyways, and it’s easy to use that fact to elide those patients from our consideration. And as Farley and Cohen point out, it’s not just the memory of doctors that is so biased:

The newspapers trumpet the uplifting victories–such as Lance Armstrong’s beating testicular cancer before dominating the Tour de France–but muffle the losses in small obituaries.

Over at medrants, DB asks how we can improve the learning climate in teaching hospitals. I suspect this crap is not what he had in mind. Spillover at KevinMD, where I get into things old-school fray style, even (hi Dawn, and a hat tip to Moloch).

I admit a morbid fascination with this type of injury. The idea that someone could shove a reasonably sized foreign object into their own brain, with massive bleeding but no external trauma, seems just . . . bizarre. Here’s this week’s New England Journal image of the week, a fifteen year old boy assaulted with a pencil: