Cookbook Medicine

There are two fundamentally different ways of thinking about complex social systems: the economic approach and the engineering approach.

The social engineer sees society as disorganized, unplanned and inefficient. Wherever he looks, he sees underperforming people in flawed organizations producing imperfect goods and services. The solution? Let experts study the problem, discover what should be produced and how to produce it, and then follow their advice.

In medicine, the engineering approach results in what many call “cookbook medicine.” This means that in treating patients with various symptoms, doctors must follow guidelines or protocols and they must record each step along the way. Cookbook medicine is the opposite of personalized medicine — an approach which aims to tailor the therapy to the characteristics of the patient, including her genetic makeup.

Before criticizing it, let me say something good about the “cookbook” approach. I think a doctor would be foolish to ignore protocols. Being aware of how other professionals have treated conditions and what outcomes they have experienced is part of being on top of what is happening in the medical profession in general and the doctor’s specialty in particular.

In MinuteClinics and in other walk-in primary care retail clinics around the country, nurses are doing a superb job of following computerized protocols. In fact they seem to follow best practices better than traditional primary care doctors.  They also seem to pretty good at recognizing when a patient’s condition is outside their area of expertise and referring that patient to a specialist or to an emergency room for more complex treatment.

But things will go wrong if the cookbook becomes a master rather than a servant; if it becomes a book of orders rather than a book of suggestions; and if complying with endless checklists takes valuable time away from patient care.  Yet that is exactly what is happening in American medicine.

I’m gonna cry,
Cry, cry, cry.

Remember, the cookbook that MinuteClinic nurses follow is a cookbook created in the marketplace for the purpose of meeting the needs of cash paying customers. MinuteClinic has an incentive to weigh costs against benefits in doing what it does. If a nurse has to type too much low-value information into her computer terminal, she will be able to see fewer patients and earn less revenue for the clinic. The cost of information overload will be judged not worth the benefit.

Contrast that with what is happening to doctors dealing with impersonal bureaucracies, which do not bear any of the costs they impose on doctors and their patients. Dr. Virginia McIvor, a pediatric physician at Harvard Medical School explains the problem as follows:

…[T]he quality police demand that for any child who comes in for a physical whose body-mass index is above the 85th percentile, I must comply with certain measures — what we call box checking. I first need to check a box stating that the child is overweight. Then I must acknowledge that I entered “overweight” in his problem list. Next, I need to check a box stating that diet and exercise counseling were provided. Finally, I need to be sure that the counseling is documented in the patient note. If this patient has asthma, I need to check more boxes for an asthma action plan, use of an asthma-controller medication, and flu-shot compliance.

When a healthy child visits, I must complete these tasks while reviewing more than 300 other preventative care measures such as safe storage of a gun, domestic violence, child-proofing the home, nutrition, exercise, school performance, safe sex, bullying, smoking, drinking, drugs, behavior problems, family health issues, sleep, development and whatever else is on a patient’s or parent’s mind. While primary-care providers are good at prioritizing and staying on time — patient satisfaction scores are another quality metric — the endless box checking and scoring takes precious time away from doctor-patient communication. Not one of my patients has lost a pound from my box checking.

In Priceless, I made the following observation:

Over all, health care is a field that can be described as a sea of mediocrity punctuated by islands of excellence. The islands always spring from the bottom up, never from the top down; they tend to be distributed randomly; they are invariably the result of the enthusiasm, leadership and entrepreneurial skills of a small number of people; and they are almost always penalized by the payment system.

Now if you think like an economist, you will say, “Why don’t we reward, instead of punish, the islands of excellence and maybe we will get more of them?” But if you think like an engineer you will reject that idea as completely unacceptable. Instead you will want to 1) find out how medicine should be practiced, 2) find out what type of organization is needed for doctors to practice that way, so 3) you can then go tell everybody what to do.

Atul Gawande is the author of The Checkbook Manifesto, in which he argues that doctors can improve the quality of medicine by following a checklist similar to the ones that have reduced airline accidents. Here is his explanation of how medicine should be practiced:

This can no longer be a profession of craftsmen individually brewing plans for whatever patient comes through the door. We have to be more like engineers building a mechanism whose parts actually fit together, whose workings are ever more finely tuned and tweaked forever better performance in providing aid and comfort to human beings.

Here is Karen Davis, explaining (in the context of health reform) how medical care should be organized:

The legislation also includes physician payment reforms that encourage physicians, hospitals and other providers to join together to form accountable care organizations [ACOs] to gain efficiencies and improve quality of care. Those that meet quality-of-care targets and reduce costs relative to a spending benchmark can share in the savings they generate for Medicare.

Finally, check out this post by Bill Gardner at the Incidental Economist, who often approaches health care from a different perspective. He writes:

As I’ve written previously, time spent documenting care may be crowding out time spent interacting with patients. A recent study of emergency room physicians found that they spend more time interacting with screens than patients, clicking the mouse 4000 times in a 10-hour shift.

The Affordable Care Act (ObamaCare) was heavily influenced by the engineering model. Who, but a social engineer, would think you can control health care costs by running “pilot programs”? What’s the purpose of a pilot program if not to find something that appears to work so that you can then order everybody else go copy it? Pilot programs are a prime example of the social engineer’s fool’s errand.

And by the way, there is no evidence whatever that pay-for-performance schemes improve quality or reduce costs — either in this country or abroad — either in health care or in education.

Social engineers invariably believe that a plan devised by people at the top can work, even though everyone at the bottom has a self interest in defeating it. Implicitly, they assume that incentives don’t matter. Or, if they do matter, they don’t matter very much.

To the economist, by contrast, incentives are everything. Complex social systems display unpredictable spontaneous order, with all kinds of unintended consequences of purposeful action. To have the best chance of good social outcomes, people at the bottom must find that when they pursue their own interests they are meeting the needs of others. Perverse incentives almost always lead to perverse outcomes.

In the 20th century, country after country and regime after regime tried to impose an engineering model on society as a whole. Most of those experiments have thankfully come to a close. By the century’s end, the vast majority of the world understood that the economic model, not the engineering model, is where our hopes should lie. Yet health care is still completely dominated by people who steadfastly resist the economic way of thinking.

Comments (34)

Trackback URL | Comments RSS Feed

  1. Larry says:

    And then there is real personalized medicine…crowd sourcing…when we create real connections with patients experiencing the same or similar conditions. Allowing these patients to interact with each other. Doctors and other providers focus on many conditions, but patients with a specific disease focus on that disease and collect data, information and knowledge about that condition and sharing that with other similarly situated patients is invaluable knowledge.

  2. Keith says:

    I kept thinking that you were going to move into a discussion of the ICD-10, which will, once again, impose high and unnecessary costs on providers to suit the desires of bureaucrats who will not pay the extra costs. Going from 17,000 billing codes to 155,000 codes this Fall will add to the Chaos of implementing the Affordable Care Act–all in the name of being able to record service for death from burning water skis and suicide by jellyfish , or (more likely) to deny payment based on incorrect coding.

  3. Devon Herrick says:

    One of them flawed aspects of organized medicine is the patient. Patients themselves causes more than half of chronic conditions with poor lifestyle choices. Medical non-adherence is cited as a problem, although it’s not clear how much damage is actually done. Patients stop taking their medications nearly half of the time.

  4. Perry says:

    The government and large institutions love to “standardize” treatment. I have no problem in theory with guidelines, but the physician must have the ability to practice within his or her judgement and intuition to some degree, otherwise a monkey could practice medicine.

  5. Perry says:

    Oh, and there’s a lot of debate on the Patient-Centered Medical Home approach. Again, this is a protocol-type solution, which may in fact have promise for some patient populations, but I don’t think it should be mandated for all primary care practices.

    • Chancy says:

      I agree. No single approach can be dominant.

    • Roger Waters says:

      LOL, in many studies (most not published as they are proprietary) the “patient centered medical home” (especially those that meet the NCQA checklist) do not improve quality or reduce costs at all, in fact in one study we did they actually cost more because of the up front costs, and certainly they don’t do any better than the typical physician practice – unless the metric is increasing the physician income. Nuff said, back to work in the real world.

  6. John R. Graham says:

    Can you imagine if we had waited for the Federal Communications Commission to launch pilot programs to determine whether people wanted phones in their cars or hands?

  7. Centrist says:

    In a nut shell, isn’t P4P simply requiring participating providers to warranty their work?

  8. Martin L. says:

    Cookbooks will not improve efficiency in the healthcare sector. Compare it to a recipe you find in an actual cookbook. It tells you exactly what ingredients to use and in what quantities. Because you don’t know there is no option but to follow it according to the instructions. The problem arises if you are missing one of the ingredients. Without the ingredients, many will decide not to cook that recipe at all. If in medicine we start sticking to a cookbook there is a chance that the quality of the healthcare system will significantly decrease.

    • Peter A says:

      But as you said people who are cooking from a recipe book don’t know what they are doing. They have no knowledge on which other ingredients can replace the one they are missing. With healthcare is the opposite. Doctors have advance knowledge on the topic, so they are capable of treating the patient with different procedures than the ones stated in the cookbook. They have sufficient knowledge to perform as they believe is most effective, in a way that they consider the patient will be benefited the most.

      • Osama says:

        Then that will not be a cookbook approach.

        • Peter A says:

          It will to a certain degree. The doctor will be following the set of procedures that the law requires him to perform, but using his medical knowledge, the physician will be able to change those procedures to ones that better suit the needs of the patient. Having a cookbook is not a bad thing, especially, as the post mentions, when it is used as a mere guideline. The problem is when these guidelines are set in stone and are the only procedures that the physician can perform.

        • SteveofCaley says:

          What Peter offers is the canonical definition of guideline. What is coming is the Orwellian definition of guideline. We should change the term “Criminal Code” to “Behavioral Guidelines,” out of deference to Newspeak. As one person said to me “Your clinical choice is whether to be obedient, or disobedient – that’s the new clinical choice!” (He forgot to say – “Comrade!” at the end of the sentence.)

    • Wilt says:

      In fact, a world-class chef never got stuck on any cookbook. Quality comes from individualization.

  9. Chancy says:

    “The social engineer sees society as disorganized, unplanned and inefficient. Wherever he looks, he sees underperforming people in flawed organizations producing imperfect goods and services.” This sounds quite cynical. Doctors must be capable of judging which method to use based on the characteristics of the patients. A standardized treatment does not actually solve the problem because such protocol is created and concluded by the doctors. It is very easy to manipulate it.

  10. Ken says:

    Very good post.

  11. Dale says:

    “the author of the Checkbook Manifesto”…maybe freudian slip, but I think you meant “the Checklist Manifesto”.

    Here’s another reference, my wife actually worked with Dr. Pronovost at Hopkins, spoke highly of him:
    http://www.amazon.com/Safe-Patients-Smart-Hospitals-Checklist/dp/0452296862

  12. SteveofCaley says:

    American Retail Medicine is rapidly going along the way towards American Retail Education – a horrible, unwieldy, ineffective and unsatisfying machine that produces poor-quality rubbish – the Bulgarian cardboard shoes of the old Warsaw Pact nations.

    I have yet to hear an intelligent argument for twenty years on the matter, other than “We should do better – so watch my trick pony perform!” Whether iPhone ultrasound apps or pay-for-performance or medical-homes, nothing will work until medicine-as-bureaucracy ultimately fails. The misery may be breathtaking; but we have chosen it.

  13. SteveofCaley says:

    The concepts put forth by Pronovost and Dr. Atul Gawande – shame on you two! This is only the IOM reform put forward under President Clinton, for goodness sakes. We’re trolling the intellectual midden pile for ideas of 15 years ago? Yes, they work, and work very well. They require more personnel, and more expense. They might in fact be brilliant. But in today’s culture, where the actual delivery of medicine is secondary to the cost, they won’t play well in Peoria.

    • Dale says:

      I don’t claim this stuff is intellectually ground-breaking…to me, it’s simply the adaptation of basic disciplines which are common to the delivery of any quality product or service, and applying such disciplines to medicine.
      Now, I’m not personally in medicine, but I see it through my wife’s experience as an anesthetist. It certainly appears to me that the sloppy execution of basic tasks and algorithms is a major problem. The shocking thing to me is that such basic practices (e.g. hand-washing, labeling drug dosages) need to be championed by individuals in the field. I don’t see this as an issue of personnel or cost, but rather accountability. You can get the most advanced care in the world at Hopkins, for example–but the housekeeping staff is unionized, and unaccountable. That means you have a *much* higher chance of getting an infection there than other comparable care centers. So the quality of care of these brilliant diagnosticians is offset by the failure to apply basic practices. For me (as a consumer) that’s an unacceptable situation–so I think the work of people like Dr. Pronovost has value. (I haven’t read the Gawande book, can’t speak to it).

      • SteveofCaley says:

        I agree, and wish to slow down and scrutinize. How is is possible that such rudimentary steps are failing? Semmelweis brought handwashing to medicine in the 1860′s. The elemental structures have been unimproved since the publication of “To Err is Human,” published in 1999, from which Dr. Gawande has unconsciously filched the ideas of redundancy in testing from. These ideas have largely been untouched.
        Why, why, why? That is the hard question in systems analysis in medicine. Intel does not have a problem with people wearing cotton t-shirts or bringing cats into the clean rooms where they make chips. Not labeling things, and not washing hands, are at a level comparable to bringing a cat into a clean room.
        Why? In spite of our tendency to reach for the ad hominem argument, it’s not that people are bad, stupid or nasty. It’s not that nobody has thought of the problem or how to fix it before. The problem is its immovable.
        I’ll bet Intel doesn’t have a “NO CATS!” sign outside their clean rooms. Don’t have to.

  14. Roger Waters says:

    Great post. Puts into perspective the typical policy wonk, as s/he perhaps should be called “social engineer” instead?

    The real question for me, having been one of the former, is how to instill economics into a discussion overwhelmingly occupied at the top (Congress/ Executive Branch) by social engineer “do gooders,” whose mission is to find an answer from their perspective – which is top-down?

  15. Morris Bryant, MD says:

    In reality, I believe a mixture of the models serves best. As a teacher of medicine, cookbook remedies have a place in teaching groups of doctors how to approach common conditions, especially when you are faced with large numbers of the same condition. However, at the bedside, the illness becomes much more personal. Sick people are not unlike cars in accidents. No two are bent in the same way. That’s when the cookbook should serve no more than as a guide. This is also where you find opportunities to individualize and find the encouraging incentives that help and engage a given patient to take part in getting well and staying well.

    • SteveofCaley says:

      As a physician, the problem in medical pedagogy, as in many other fields, is that fields have to be taught under the pretense of order and formalism, as though the topic is crisply defined.
      The reality is that diagnosis and disease taxonomy is to a large degree nominalism – there is a vast difference between one person’s manifestation of disease, and another’s.
      It reminds me of an old joke when two guys were planting telephone poles. The first guy got twenty done in one day. The second guy got four done.
      The boss asked the second guy why he couldn’t keep up with the first guy. He said, “Look how far apart his are, and how much he left sticking out of the ground!”
      In real medicine, there is a lot of interpolation in the space between the comfortable indications of guidelines. It’s the space in between that’s tricky and undefined.

  16. charlie bond says:

    Good morning, John:

    Once again, you are wandering into territory that is all too familiar. Cookbook medicine, per se, will work–just as soon as they standardize the human body and all the diseases and injuries that attack it.

    Having said that, you seem to be discouraging “pilot projects” as “engineering” simply because they are planned. In fact, one of the great (perhaps unintended ) consequences of the current upheaval in health care is the encouragement of innovation and pilot projects.

    By casting off regulatory shackles, a few ACO’s are coming up with effective alternatives to fee for service–systems that blend quality and cost incentives. The best systems use active case management predicated on protocols, while leaving to the doctor just how best to implement care for the optimal outcome at the most efficient cost.

    The most direct example is Accountable Care Associates based in central Mass. They have been very effective in creating tools and incentives to both improve care and lower cost. Through gain-sharing contracts their providers and payors have done well. Hopefully, the next iteration will use some of the gainsharing to incentivize patient behaviors and choices as well.

    Likewise, continuum-of-care organizations directed at the 5% of the population that drives 90% of our health care costs also seem to be working. Again, they rely heavily on case management. While case management, in turn, relies on best practices protocols, it is not cookbook medicine. To the contrary, it is active intervention in the patient’s care to remind everyone to THINK and be mindful of outcomes and cost. Funny what a little mindfulness can do. And if we save a million here, a million there, after a while it adds up to real money.

    So before we condemn “cookbook” medicine, let’s look in the cookbooks to see if there are recipes for success. in the end, health care is a one-on-one personal services industry. It is, therefore, only as good as each practitioner’s skill (and knowledge of protocols) and only as successful as each patient’s motivation. Without individual engagement and participation by all parties in addressing the patient’s unique problems, health care will continue to be a unnecessarily expensive and increasingly impersonal exercise. To the extent we can bring this mindfulness to all parties through case management and incentivized quality metrics, we can improve care. We just have to be sure to check the patient, as well as checking the box.
    Cheers,
    Charlie Bond

  17. Greg Scandlen says:

    This is related to two other bumper sticker “solutions” — standardized care and population health. Both are monstrous. Standardized care is the polar opposite of personalized care. And population health implies that the health of the individual is unimportant, especially if it detracts from the health of the population. One way to improve the health of the overall population would be to kill sick individuals.

    It is all Marxist thinking. The same thinking that drives union contracts — pay everyone the same regardless of how good they might be at doing their jobs.

  18. Sherif Khattab, M.D. says:

    It is truly amazing that so many bright, intellectual and experienced minds keep missing the “ONE THING”, as in City Slickers, that Healthcare is for. It was, remains and will always be the PATIENT.
    So, is missing the one thing and keep trying to find an alternative, is it due to : A- Lack of awareness?
    Or
    B- Intentional?
    I keep asking this question to my friends and colleagues and when I get an answer it is usually an interesting one.

    • SteveofCaley says:

      In healthcare – and healthcare exists everywhere that humans live – the most immediate and relevant controlling connection is that between the patient and doctor. The patient seeks not solitary control but problem relief; the doctor seeks not authoritative obedience but empathic partnership.

      The decisions reached at a result of this discussion, in some environments, be seen as controlling and without contravention.
      People have been striving to figure how to bust into this connection for years. I pick (B) – intentional.

      The Emperor’s Clothes of the healthcare argument assumes that deferring to the primacy of the doctor-patient relationship leads to poor care and exploding costs. All I asks is – is it so?

      Many other decisions are amenable to authoritative intrusion from others – why do we not permit that in our environment? Individual housing market purchases, unregulated, led to a cost explosion on healthcare. Shouldn’t the government establish a presence between real estate buyers and sellers, or is that not as absurd as it sounds? If absurd in real purchases, why is it not in healthcare?

  19. Ron says:

    Government pilot programs should be contrasted with entrepreneurial proof of concepts. Entrepreneurs react quickly to consumer interests, demands, and purchases. They develop improvements to products and services. Government pilot programs are rarely expanded, rather they are checked off a bureaucrats to-do list and the follow up is another pilot that leads to no where.

  20. Larry Wedekind says:

    Hi John, et al. I have been way too busy, but finally have had a moment to catch up on your insightful and well written healthcare Blogs. I had to have a minor surgical procedure and am recovering at home and am forced to slow down and smell the roses. John, your Blog is a very fine and is a fragrant rose. Thank you for your thoughtful contributions that force us to think about the what, why, and how that we manage and operate within the healthcare system.

    My comments: Cookbook medicine, as you put it, is an essential innovation of the 20th century that is saving lives every day in every part of the country. Cookbook Medicine contributed to my procedural success in my surgery as well as the art that my surgeon employed to produce a favorable outcome. When Cookbook Medicine (standardized clinical diagnoses and approaches, tests, and procedures) is required by government in all cases, then recognition of patient (human) differences suffers and this must NEVER be allowed. I totally agree with Charlie’s statement that Pilot Projects that reward innovation and results are worthwhile and needed in the healthcare arena.

    I do have the general answer as to why most Pilot Projects have failed to produce the results that were sought by the government. These Pilot Projects do not typically involve any risk for the recipients of the grants that enable these Pilots. Note that when my company is enabled to embark upon a new venture, we typically assume significant financial risk and are rewarded if we succeed in accomplishing the stated goals. This is highly motivational.

    My company (IntegraNet) operates an ACO and we assumed substantial financial risk in developing and managing this ACO. In fact, it has cost us a small fortune. However, our first year resulted in an 11% Shared Savings! We saved the taxpayer $9.8 MM in our first 12 months. We actually exceeded the maximum allowed Shared Savings by over $300,000! It appears that we were the top performing ACO in the country our first year.

    Why? We employed tried and true cookbook gain sharing and Care Coordination methods to change physician and patient behavior in order to improve population health and save money as a result. We have been doing this Care Coordination thing for quite awhile now and so we are better than most – so far. Note that we also encourage our physicians to be innovative within their own offices and we encourage our physicians to learn from each other in the process. We do not require our physicians to only practice a certain way and our physicians employ many different methods to comply with our outcome related quality measures. Quality measures like “A1c below 8″ in diabetics. We don’t care too much about how the physician actually gets their diabetic patient to this desired outcome. We simply reward the physicians for accomplishing and maintaining this very difficult goal for their diabetics and we penalize those who don’t care enough to get their patients to this desired outcome.

    Simple enough? The devil is in the details of course. But we are motivated by financial risk and reward to get it done! This is the Economic model that John talks about – and we at IntegraNet love it!

  21. George says:

    Plato contemplated a version of this duality in his last book, The Laws. In a brief and remarkably prescient chapter titled Two Categories of Doctor, he contrasted “Slave Doctors”, who care for “invalids of the state” dispensing provisions from a tray and without dialogue, with “Physicians Befitting Free Men,” where there is an accommodation and individualization of care based on a relationship. So now we are institutionalizing our versions of “engineered” (? slave) “health care,” and systematically suppressing “economic” (? free wo/men) medicine. By the way, the engineers appear to be winning the battle; as to the war, maybe we will get the outcome we deserve. Plus ca change…

  22. George Sack says:

    “Cookbook” medicine is the antithesis of whatever is meant by “personalized” medicine. Growing recognition of human heterogeneity at the DNA level should ultimately inform and serve as the basis for the latter. Unfortunately, the complexity is far greater than anticipated – a concern is that the “cookbook” folks so overwhelm practice that true “personalized” care cannot arise.