How To Get Physicians To Use The Same Science-Based Playbook

Pretty much everyone agrees that we need to improve the quality of healthcare delivered to patients in the US. We’ve all heard the frightening statistics from the Institute of Medicine about medical error rates – that as many as 98,000 patients die each year as a result of them – and we also know that the US spends about 33% more than most industrialized country on healthcare, without substantial improvements in outcomes.

However, a large number of quality improvement initiatives rely on additional rules, regulations, and penalties to inspire change (for example, decreasing Medicare payments to hospitals with higher readmission rates, and decreasing provider compensation based on quality indicators). Not only am I skeptical about this stick vs. carrot strategy, but I think it will further demoralize providers, pit key stakeholders against one another, and cause people to spend their energy figuring out how to game the system than do the right thing for patients.

There is a carrot approach that could theoretically result in a $757 billion savings/year that has not been fully explored – and I suggest that we take a look at it before we “release the hounds” on hospitals and providers in an attempt to improve healthcare quality.

I attended the Senate Finance Committee’s hearing on budget options for health care reform on February 25th. One of the potential areas of substantial cost savings identified by the Congressional Budget Office (CBO) is non evidence-based variations in practice patterns. In fact, at the recent Medicare Policy Summit, CBO staff identified this problem as one of the top three causes of rising healthcare costs. Just take a look at this map of variations of healthcare spending to get a feel for the local practice cultures that influence treatment choices and prices for those treatments. There seems to be no organizing principle at all.

Senator Baucus (Chairman of the Senate Finance Committee) appeared genuinely distressed about this situation and was unclear about the best way to incentivize (or penalize) doctors to make their care decisions more uniformly evidence-based. In my opinion, a “top down” approach will likely be received with mistrust and disgruntlement on the part of physicians. What the Senator needs to know is that there is a bottom up approach already in place that could provide a real win-win here.

Some 340 thousand physicians have access to a fully peer-reviewed, regularly updated decision-support tool (called “UpToDate“) online and on their PDAs. This virtual treatment guide has ~3900 contributing authors and editors, and 120 million page views per year. The goal of the tool is to make specific recommendations for patient care based on the best available evidence. The content is monetized 100% through subscriptions – meaning there is no industry influence in the guidelines adopted. Science is carefully analyzed by the very top leaders in their respective fields, and care consensuses are reached – and updated as frequently as new evidence requires it.

Not only has this tool developed “cult status” among physicians – but some confess to being addicted to it, unwilling to practice medicine without it at their side for reference purposes. The brand is universally recognized for its quality and clinical excellence and is subscribed to by 88% of academic medical centers.

In addition, arecent study published in the International Journal of Medical Informatics found that there was a “dose response” relationship between use of the decision support tool and quality indicators, meaning that the more pages of the database that were accessed by physicians at participating hospitals, the better the patient outcomes (lower complication rates and better safety compliance), and shorter the lengths of stay. (Caveat: Additional studies are needed to clarify whether use of UpToDate is a marker for the better performance, an independent cause of it, or a synergistic part of other quality improvement characteristics at better-performing hospitals.)

So, we already have an online, evidence-based treatment support guide that many physicians know and respect. If improved quality measures are our goal, why not incentivize hospitals and providers to use UpToDate more regularly? A public-private partnership like this (where the government subsidizes subscriptions for hospitals, channels comparative clinical effectiveness research findings to UpToDate staff, and perhaps offers Medicare bonuses to hospitals and providers for UpToDate page views) could single handedly ensure that all clinicians are operating out of the same playbook (one that was created by a team of unbiased scientists in reviewing all available research).

I believe that this might be the easiest, most palatable way to target the problem of inconsistent practice styles on a national level. And as Senator Baucus has noted – the potential savings associated with having all providers on the same practice “page” is on the order of $757 billion. And that’s real money.

Posted in: Politics and Regulation, Public Health, Science and Medicine

Leave a Comment (21) ↓

21 thoughts on “How To Get Physicians To Use The Same Science-Based Playbook

  1. unBeguiled says:

    UpToDate rocks!

    For the most part, I agree with your skepticism about a top down carrot/stick approach. I did experience an anecdotal case of top-down regulation that quickly lead to better outcomes.

    I was working at a hospital where we had many older physicians who had just not kept up. We instituted guidelines and regulation of initial orders and discharge orders for Community Acquired Pneumonia and CHF. Things got better quickly.

    I know this kind of institutional initiative is not what is at issue here, but merely serves as an example of a top-down approach that worked.

    (Do you have a financial interest in UpToDate? Sorry, my conflict-of-interest detector is hypertrophied from years of exposure to Drug Reps bearing gifts.)

  2. weing says:

    As a subscriber of UpToDate, I rely on its impartiality. If the government subsidizes subscriptions or gives bonuses based on use, I have no problem with it. I am not sure if I trust the impartiality of government sponsored comparative clinical effectiveness research findings. I see the potential for conflicts of interest there.

  3. Adrian says:

    Even though you emphasized that the US is an outlier in spending, the gap is much worse than you say.

    we also know that the US spends about 33% more than most industrialized country on healthcare

    Actually, the study you linked to shows the US spends $6,037 per capita and the second most expensive nation, Switzerland, spent $4,045. This means the US spends 50% (not 33%) more than the most expensive European nation and this includes a huge number of perks that are unheard of in the US. The average industrial country is actually much less than this.

    without substantial improvements in outcomes.

    Saying that the US doesn’t get a substantial improvement leaves the impression that the outcomes in the US are matched by countries like Switzerland (the European country with the highest health care spending, against which you’d compare per-capita spending). Not true. In fact, the US has one of the lowest life expectancies of any industrialized country. Far from not having any substantial improvements, it would be more accurate to say that despite the extra money, the outcomes are significantly worse.

    I tried to find a good source which compares life expectancy with per-capita health-care spending and found:

    Note that the US life expectancy exceeds only Cuba, Portugal, Ireland and Cyprus. In fact, the US spends over 10 times more per capita than Cuba (possibly much more) and has a comparable life expectancy. It spends 300% more than any of the other three and here is truly “without substantial improvements in outcomes”.

    I think your suggestions are interesting. I question whether the large disparities between the spending and outcome of the US and, say, Canada or even Cuba can be bridged by PDAs. Canada and Europe are hardly technological backwaters but I doubt that gaps in spending and patient outcome are due to better availability of technology. If there are simple changes the US can make to cut costs those same changes will be adopted by other countries and the gap will remain.

  4. weing says:

    Comparisons with other countries are very unreliable. When I look at the data, which is sparse, I find the conclusions are suspect. For example, a simple thing like life expectancy at birth should be easy to determine. Is the definition of birth the same in all the countries? I checked with the data the CDC uses and I couldn’t find any information on this. They simply assume our definitions are the same. Are what we call very low birth weight neonates that end up in NICUs considered nonviable in other countries and considered as spontaneous abortions and not counted? When you consider that the difference between the US and the country with the highest life expectancy is about 5-6 years, that information may be useful in explaining both the costs and longevity.

  5. Adrian says:


    If you look at the data, it isn’t chunky or tiered as we’d expect if the data was due to these hypothesized collection differences.

    I think the costs aren’t as hard to explain as you imagine. You’ve got an enormous bureaucracy which must be funded. HMOs, health care plans, tiered government programs and layers and layers of administration, all of which add to the costs of everything and which no other industrial country has. This probably has follow-on effects such as increased drug costs but this is icing on a bloated red-tape cake. As well, there are large disparities in access and affordability to health care in the US which no other industrial country has which can easily explain the lower life expectancy.

  6. Adrian says:

    This means the US spends 50% (not 33%) more than the most expensive European nation and this includes a huge number of perks that are unheard of in the US. The average industrial country is actually much less than this.

    I meant to say “The average industrial country spends much less than this”, which means the US actually spends up to 100% more than other industrialized nations.

  7. daedalus2u says:

    Adrian, it is my understanding that a large part of the additional costs (unique to the US) imposed by health insurance companies include profit, and excluding high risk patients from their subscribers. Neither of these costs do anything to deliver health care, only to deliver profits to shareholders of health insurance companies.

    When insurance companies are large enough to negotiate fixed prices for procedures, hospitals with less uncompensated care are at a tremendous advantage.

  8. Adrian says:

    daedalus2u – interesting. I wonder how you could go about quantifying that. According to a page on the “National Coalition For Health Care”‘s site:

    A recent study by Harvard University researchers found that the average out-of-pocket medical debt for those who filed for bankruptcy was $12,000. The study noted that 68 percent of those who filed for bankruptcy had health insurance. In addition, the study found that 50 percent of all bankruptcy filings were partly the result of medical expenses.9 (Source:

    That says that, at the very minimum, 36% of people who file for bankruptcy due to medical expenses had health insurance (18% of all bankruptcy claims). supports this:

    Surprisingly, most of those bankrupted by illness had health insurance. More than three-quarters were insured at the start of the bankrupting illness. However, 38 percent had lost coverage at least temporarily by the time they filed for bankruptcy. (Source:

    Based on the details, it looks like some people are denied treatment outright but more subtly, others have very high deductibles and co-payments so that they are stuck with a large chunk of the bill. The piece ends with a quote:

    “The paradox is that the costliest health system in the world performs so poorly. We waste one-third of every health care dollar on insurance bureaucracy and profits while two million people go bankrupt annually and we leave 45 million uninsured” said Dr. Quentin Young, national coordinator of Physicians for a National Health Program.

    If that’s true and only one third is wasted on bureaucracy, there are still a lot of dollars that are wasted elsewhere.

  9. pmoran says:

    UptoDate sounds like a brilliant idea for making medical practice more evidence-based, but I am not at all sure that this would save money overall, or how the authors of the relevant reports could know that.

    It looks to me as thought the areas of the US with the highest medical services are simply those likely to have the highest population of doctors and specialists, a factor well known to be a major driver of health care costs (typically without any measurable effect upon crude outcome indices).

    This is why countries with public-funded systems such as Australia and Canada periodically try to restrict doctor numbers.

  10. Val Jones says:

    Actually pmoran, the explanation for practice pattern variation isn’t as simple as “physician concentration.” The CBO has scratched their heads for an extended period of time on this – trying to identify the underlying cause. In the end they attribute it to micro-cultures (the old “we do this because we always have” in a given local area of influence.)

    The potential savings is related to efficiencies that are not immediately apparent, including things like administrative simplification:

    1. Health Plans might agree to forgo “pre-auth” requirements for tests/procedures recommended by UpToDate. This could eliminate admin burdens and allow docs more time to see patients.
    2. One could imagine incorporating UpToDate with EMR systems that would reduce test redundancy, the “gunshot approach” to diagnosis, and allow docs to order the right test the first time (e.g. not having to order an X-ray just to get an MRI when MRI is indicated first)
    3. If UpToDate were combined with Comparative Clinical Effectiveness Research, physicians and patients could have more intelligent shared-decision making discussions at the point of care. I do think this would lead towards a decline in over-treatment.

    And to Adrian – thanks for checking my “maf.” :)

  11. pmoran says:

    Oh, I don’t doubt that money could be saved in some areas by better use of an evidence base. But to have any overall effect on health costs you would have to be very selective as to which areas of EBM you chose to enforce or encourage.

    In my area, surgery, you will find that many patients still do not receive all the components of optimum evidence-based care of breast or bowel cancer, either through geography or practitioner oversight. Will the CBO pay the additional expenses involved in bringing rural patients into major centers for radiotherapy or better oncological care? Or will we need another layer of bureaucracy watching what doctors do?

  12. Jules says:

    I don’t think getting government sponsorship for a peer-reviewed tool would necessarily be the wisest course, considering that the government is anti-abortion, pro-creationism, and all for prayer.

    Well, okay–the government isn’t, but Murphy’s Law says that the ones who are will be the ones overseeing the sponsorship.

    It already happens enough–South Dakota, anybody?–that I’m surprised anybody would think that the answer to government interference is more government.

  13. marilynmann says:

    David Rind, one of the authors of that study and an editor of UpToDate, is a friend of mine. As you know, I’m not a physician, but David has sent me topics from UpToDate on a couple of occasions and I was favorably impressed with it.

  14. DevereuxBob says:

    This is the kernel of a very powerful approach, combining two important ideas, an “open source” approach and positive behavioral supports. I love it. I am very interested in the commitment in the stimulus package to creating a standardized electronic health record, and have been thinking about the benefits of an open source approach, rather than a top down approach. Although the behavioral health organization I work for ( is committed to positive behavioral approaches, rather than negative reinforcement, I had not thought about the power of combining the two (open source and positive reinforcement). I will be in DC early next week talking to legislators, if anyone has additional thoughts about this approach, I would very much like to hear them.

  15. Dr Benway says:

    “…committed to positive behavioral approaches, rather than negative reinforcement”

    Nitpick: “Negative reinforcement” is like turning off an annoying alarm clock – the removal of a bad thing, which is good, from your perspective.

    I think you mean something like, “bad dog!” which can be categorized as “punishment.”

  16. Dr Benway says:

    Oh, fergot* what I wanted to say to de legislators…

    Look at reality before trying to fix it.

    Before imposing some carrot-stick game on doctors to straighten them out, let’s find out what they’re actually doing. Let’s figure out a way to aggregate data regarding who is being seen for what, what’s being done to help, and how that’s working out.

    Please read my little rant about stupid data tracking.


    *fell on ice, bonked head x2, has concussion

  17. Dr Benway says:

    Hmm. UpToDate may be crap for psychiatry, unless the neurology section is expansive.

    $500 is not a casual purchase.

  18. DevereuxBob says:

    # Dr Benway on 02 Mar 2009 at 11:26 pm
    “…committed to positive behavioral approaches, rather than negative reinforcement”

    Nitpick: “Negative reinforcement” is like turning off an annoying alarm clock – the removal of a bad thing, which is good, from your perspective.

    I think you mean something like, “bad dog!” which can be categorized as “punishment.”

    You are right, of course. I enjoyed your rant on data tracking. Thanks.

  19. Dr Benway says:

    Everyone makes that “negative reinforcement” slip. It’s because “punishment” is a dirty word. Few can bring themselves to use it.

Comments are closed.