Better brush up on your hospital corners

Well, this should be fun.

In line with the goals of national health care reform, New York City’s public hospital system has embarked on a radical change in the way it will pay doctors. Instead of granting automatic pay increases, it will pay them based on how well they reduce costs, increase patient satisfaction and improve the quality of care. The Health and Hospitals Corporation, which runs the city’s 11 public hospitals, deserves praise for an ambitious proposal that will need to be refined as it is put into practice.


The corporation has now jettisoned that approach in new contracts with three “affiliates” — the New York University School of Medicine, the Mount Sinai School of Medicine and the Physician Affiliate Group of New York — which together employ more than 3,300 doctors who deliver care in the public hospitals. There will be no cost-of-living increases for the next three years. Instead, the contracts will give annual bonuses to the affiliates if their doctors do well in meeting performance goals, like improving the coordination of care and reducing the average length of stay.

One flaw is that there will be times when the doctors who actually provide the services cannot through their own efforts meet some of the standards by which they will be judged. One standard holds emergency room doctors responsible for reducing the time between deciding that a patient needs hospitalization and sending the patient out of the emergency room. Yet the patient cannot leave the emergency room unless there is a bed available, the room is clean, the head nurse on the floor agrees to accept the patient and an orderly arrives to transport the patient.

First of all, a disclosure: I have old professional connections with two of the three listed “affiliates.”  More on that in a second.

But on to quality.  Who doesn’t agree that better quality deserves higher reward?  You pay more for a Lexus than you do for a Scion.  The egg roll at Shun Lee costs more than the one that’s been sitting under a heat lamp all day at the hole-in-the-wall ptomaine factory around the corner.  The bill from Debevoise & Plimpton will set you back a lot more than the one from the legal office you found through an ad on the bus.

Nobody disputes the rightness of any of this.  And why should it be any different for medicine?  As we prepare the winding cloth for the expiring fee-for-service carcass, why shouldn’t some kind of quality model take its place?

Well, first of all, who determines quality?  Under our current spaghetti-pile of a healthcare “system,” third-party payers determine what gets rewarded.  Unlike with luxury automobiles or affordable home furnishings or the like when consumers themselves decide whether they’ll pay more for a particular good or service, with Acura or Ikea or what have you rising or falling based on how happy it ends up making people with their purchases, with healthcare it ends up being Blue Cross (to pick a random example) who doles out the dough.  Sure, “patient satisfaction” is an oft-referenced quality measure, but some satisfaction survey is at least one step removed from patients simply deciding if the care they receive is worth the money they’re paying for it. [Edited to add: For a further beef with patient satisfaction surveys as quality measure, see roto_tudor’s comment below.]

And how does one determine quality?  Which measures does one use?  How accurate are they?

I happen to know one of those “affiliated” public health hospitals all too well.  And I can tell you that every single step in the sequence described in the last sentence of the above quote is fraught with delay and complication.  In some cases, said difficulty or delay can be mitigated or circumvented.  (I learned real fast not to wait around for the patient transport orderly to take my patients where they needed to be, at least if time was of the essence and I didn’t want to get yelled at because they’d missed their scheduled CT scan.  I wheeled them there myself.)  But heaven help you, Mr. Emergency Physician, if you got too testy that one time with Miss Smith on 8 (and yes, you had better call her by her last name, and yes, she might well go by “Miss”), because damned if that room doesn’t take forever to get cleaned.  And now, if your paycheck is on the line?  My friend, you’d best march up there with the Windex yourself.  (Conciliatory flowers will be cheap at any price.)

I’m hardly immune to this out here in Private Practice Land, either.  And I assure you that the menu of quality measures we can shoot for is arbitrary and confusing.  (As of now, at least, I believe we’re still allowed to a certain extent to select which quality benchmarks to meet.  We try to select goals we can hit, as opposed to pie-in-the-sky attempts at solving intractable problems.  Obesity reduction?  Forget it.)  If our asthmatic patients are supposed to get a set number of check-ups per year, do we have to have them schlep back to our office even if they’re seeing an allergist for asthma management?  If I’m supposed to have a certain percentage of my teenage patients screened for a common STI, must I order a test on my Tanner II boy who evinces utterly no interest in sex at this time?  (Particularly when that test will then freak the crap out of the parents when it appears on their Explanation of Benefits statement sometime in the future?)

What’s the solution to this problem?  Hell if I should know.  Since the Affordable Care Act seems to have doubled down on our current hash of a system, clearly we’re stuck with profit-based outside entities deciding what kinds of healthcare are paid for at what level for everyone for a while yet.  Do I agree with the Times editorial board that doctors should be compensated in a manner that takes account of how good a job they’re doing?  Sure.  But I think we’re a long way from a system that will reliably do that.

Russell Saunders

Russell Saunders is the ridiculously flimsy pseudonym of a pediatrician in New England. He has a husband, three sons, daughter, cat and dog, though not in that order. He enjoys reading, running and cooking. He can be contacted at blindeddoc using his Gmail account. Twitter types can follow him @russellsaunder1.


  1. I don’t know about anyone else, but I had to look up “ptomaine.” I anticipate that this recent vocabularial annex may not get a lot of play at cocktail parties.

    • I actually spent a little bit of time pondering whether or not to use that term, since it is outmoded and reflects a discarded notion of food poisoning. But it seemed too amusing a word to replace, so I kept it.

      • I was not familiar with the term, and when I read the post I initially misread “ptomaine” as “poutine.”

        I was extremely amused to discover how close I had been.

        • I had the chance to have my first order of poutine. Turns out 8 million Quebecians(?) CAN be wrong.

          • “Quebecois” (or, more crudely, “Quebecers”).

            I’ve only ever had poutine when I’ve been… uhh… intoxicated. At such times, it is WONDERFUL.

          • Did you get it from McDonald’s or Wendy’s or something?

            If it’s not from a cart, it’s not the one you should be using for judgment.

          • Carts are fine, but diners are the way to go… or places that are devoted to just poutine. (There’s a place in or around Quebec City that offers massive plates of various recipes of putine, and it’s like all they do. Of course, I doubt most of our readers will be spending much time in QC.)

          • I’ve had poutine at several different places, and love it.

            A couple places in Victoria do a poutine with pulled pork, which is probably the least healthy thing in the world and is delicious.

      • Ptomaine was a classic term in my house; as was dreading a visit from those unwanted houseguests, Sam and Ella.

    • Na Allan Sherman fans?

      “Hello Mudda, Hello Fadda,
      Here I am at Camp Granada.
      Camp is very entertaining,
      And they say we’ll have some fun if it stops raining.

      I went hiking with Joe Spivy;
      He developed poison ivy.
      You remember Leonard Skinner;
      He got ptomaine poisoning last night after dinner.

    • Ha. Being from Maine, the place that invented the ‘Idaho Potato,’ I thought it had something to do with potatoes in the diet; it doesn’t. But to avoid ptomaine I’d avoid putting animal by-products in the compost used on those tatters.

        • No, he was from MA. But the farms that did his test growing and grew the first crops were in ME; he developed the potato to resist potato blight which was a problem here at the time. Or so our potato-historians tell us.

          And if potatoes were people, Mr. Russet Potato Head would like Tim Russert, which is how you pronounce ‘russet’ if you’re from ME.

    • There was a pile of Mad Magazines from the late 60’s? Early 70’s? in the corner of one of my friend’s garage and “ptomaine” was one of the words that one of the usual gang of idiots (could it have been Al Jaffee?) thought was funny in its own right.

      Our little circle grew up saying “puh-to-mane”.

  2. I read an article somewhere that examined the trouble with assessing doctors based on patient outcomes. They noted how things like the timing of shifts (both in terms of time-of-day and day-of-week) can drastically impact outcomes when other factors are normalized. If you regularly work the overnight shift on a Friday, you are going to see different types of patients with different types of ailments than the guy who regularly works the daytime shift on a Wednesday. At least, that is what the article concluded.

    Is this accurate? Or at least a real issue when it comes to evaluating doctors? It makes sense to me, in part because I analogize it to the trouble (but not impossibility) of evaluating teachers based on student outcomes when teachers are working with a wide range of students in a wide range of contexts.

    • It matters where your emergency department is situated. It matters what shifts you work. It matters what your patient population comprises. Etc etc etc in saecula saeculorum.

      I happen to think a great deal of what passes for “quality measures,” at least at this point, is tedious busywork and/or orthogonal to patient care.

    • Evaluating teachers was the first thing that the OP brought to mind, and all the perverse incentives both kind of ratings engender.

      “I’m very sorry, but your prognosis isn’t good. Anyway, I’m going to write you up as a sprained wrist that responded beautifully to treatment. Now get the fish out of my office before you ruin my score.”

      • Are we able to say that the medical care available at the Broadmoor is better than the medical care available on South Nevada? (Make the necessary substitutions appropriate for your neck of the woods if you’d like.)

        It seems to me the fact that South Nevadans not having medical care as good as those who live near the Broadmoor have was one of the things that inspired the PPACA.

        I mean, surely what we mean when we say that the Broadmoor is better than South Nevada vis a vis medical care is something measurable…

        • And surely the same is true about schools, but when attempting to measure it largely results in teaching nothing but the answers to the standardized test that does so, we’ve lost more than we’ve gained.

          • That depends on how useful the answers to the questions on the standardized test are.

          • Well, I personally use “C” and “none of the above” all the time.

          • None of the above, rather. Money too, I guess, if I could make that make sense.

          • @Will I’m guessing less so than learning how to write a paragraph or approach solving story problems in general.

            That ain’t teaching, that’s the way you do it.
            You get the test out and it’s one two three
            That ain’t teaching, that’s the way you do it.
            Money of the above and the letter “C”

  3. ” One standard holds emergency room doctors responsible for reducing the time between deciding that a patient needs hospitalization and sending the patient out of the emergency room.”

    That big noise you just heard? That was the sonic boom that EM just made sprinting down the list of my possible specialties. I can’t think of a WORSE metric to gauge a doc on in the ED. (Yes I can, a customer service survey on someone who either was properly denied pain meds, or had to wait in triage for 4 hours for whatever legitimate-to-everybody-but-them reason)

    • a customer service survey on someone who either was properly denied pain meds

      Or antibiotics. Or an unnecessary test. Or a referral to a quack specialist. Etc.

    • One suspects and hopes that the metrics will be more complex than this alone, but the point is a good one.

      The problems here are twofold:

      1) Patients are not in a particularly good position to judge whether they’ve received good care or not. They can judge outcomes, but bad outcomes result from good care in a statistically significant number of cases. Good outcomes result from bad care in a nontrivial number of incidents, too. They can judge quality of customer service, too, but as you point out this is hardly a reliable indicator of the performance of the medical provider complex as a whole nor of the service delivered by an atomized provider (an individual doctor, nurse, or tech). Customer satisfaction can at most only be one factor among many — but it is the factor that is driven by the people who ultimately pay for it all. So it’s not obvious how heavily to weigh it.

      2) Optimal decisions in a given situation may not be optimal in other, even very similar, situations, based on a wide universe of potential variables. In particular, the cost-control dimension of the metrics Doc describes in the OP suggest that insurance is looking for up-front costs to be cut, although again I suspect that there will be some actuarial influence for likely outcomes and cost-control will be taken from that perspective rather than a more risibly short-term approach. But even cost control is not quality optimization, and Doc has written extensively how identical or near-identical procedures can be coded, and thus billed, differently based on all manner of things. Even deciding whether a decision was well-made fully in context is one that intelligent professionals of good faith will disagree on, perhaps sharply, when there are multiple reasonable options to choose from. How, then, to pick the “best” resolution to an issue and financially incentivize it?

      • I’ve written at some point in the past about a different model, which Children’s Hospital is trying to get in front of. It’s the “global payment” model, in which you get a certain sum from the third-party entity based upon some calculation of subscriber utilization, level of expertise of faculty, etc and then whatever cost controls and other efficiencies your institution wants to enact to come in under budget is up to them.

        I can see other perils with this model, but at least it seems more proximately-related to some measurable outcome, as opposed to a vague notion that only lightly glances against actual patient care.

        • there are uses for press ganey even in an ED context, but those uses are really, really, really, really, reaaaaaally broad. really broad.

          i am skeptical about pay for performance, but we’re going to see an awful lot more of it regardless.

    • Down the LinkyFriday pipeline is a post on a study suggesting that hospital returns (where they come back to the hospital after being sent home) are not usually the product of doctors giving patients the brush off but something called post-hospitalization syndrome where they feel new or different symptoms. This cuts at the heart of one of the metrics that was supposed to be used to measure hospitals. The assumption is that hospitals are costing the system money by punting sick patients who return.

  4. Do you see any obvious things that would improve care and reduce costs at the same time, Doc?

    • Sure! Good medicine, that’s what. But good medicine can be:

      1) Time-consuming. It can take a lot longer to explain to a patient’s satisfaction that the antibiotic they promise they’re not seeking (but are) is really not indicated, that the blood test really won’t show anything useful, that referral to an expensive consultant will yield no additional insight, than to scribble out a scrip or order or referral.

      2) Unrelated (as noted) at Actual Feelings of Satisfaction. Despite spending many, many hours of my life that I will never get back Explaining Things, many patients leave my office thinking I am incompetent, callous or both because I refuse to scribble as per their desires.

      • FYI: When I was looking for a pediatrician for my children, I purposely sought out one who would not prescribe medications to the child to make the parent feel better.

        I would have loved you!

          • So when you “play God”, you go more OT-style then.

            On another type of OT, another icon has fallen, Doc: photo evidence of “Weird” Al Yankovic caught with a dead hooker has hit the newswires.

          • I’m trying to work out if a) that’s a pun, b) it’s Millicent’s judgement, though I doubt she’s ever use such language, or c) you actually hear this from your patient’s parents on a regular basis, and so have come to believe it true.

        • I did mention I work in the field, didn’t it? It’s what a lot of our “performance metrics” are based on.
          For now, at any rate, they don’t seem totally nefarious. Skeptical me is cynical.

  5. What’s your take on outfits like JCAHO? Having done some work for them, long ago, I don’t know that my opinion matters much.

    • Honestly, I really don’t know enough about the nuts and bolts of JCAHO to comment with any particular authority. I DO know that, during my residency, we had this plastic tub where we’d toss the (non-sharp or -biohazard, obviously) detritus from shots and blood draws and the like. It was more convenient than the garbage can down the hall, and when it got full someone would empty it out.

      It disappeared when we knew JCAHO was coming and reappeared after they left. FWIW.

      • Things might have changed somewhat in the interval, but back then, JCAHO was the equivalent of IG Inspection in the military. Lots of superficial fixes, attempts to obscure more fundamental shortcomings, painting rocks and suchlike — but little in the way of actual quality control.

        But the IG guys were pretty clever. They knew the usual tricks and sometimes would hit paydirt. Then some changes for the better would be made.

  6. Holding employees responsible for things they can’t control is a big fad in government circles these days. In BC, deputy ministers in the provincial government have part of their salary held back, and receive part of it based on whether morale/satisfaction in their particular ministry are up (which makes at least a little, even if it’s not something they have complete control over, and encourages them to treat their staff well) and another part based on whether morale/satisfaction within the entire public service are up (which is not something they can control, and is ridiculous).

  7. Russell – How would “number of nosocomial infections per year” work as a metric for evaluating quality? It seems like something that should involve the actions of all staff at some level (or at least, all staff who are interacting with patients in any way). And it’s an important issue.

    • I think that’s a perfectly fine quality measure. Indeed, I’m sure there are a great many quality measures that would comport with actual quality of care delivered. I just think it’s going to be some time before an especially good system is truly put into play.

Comments are closed.