Treatment guidelines pros and cons

I was puzzled by a Health Affairs article showing the public finds arguments against treatment guidelines a lot more compelling than arguments in favor. But after reading the technical appendix, which contains the full text of the survey, I think the problem is that the researchers framed the question poorly. In particular, the researchers portrayed guidelines as unrealistically rigid.

In A National Survey Reveals Public Skepticism About Research-Based Treatment Guidelines, Gerber et al. asked respondents to rate their level of agreement with arguments for and against the use of guidelines.

Arguments for guidelines (in order of percentage of respondents finding them convincing) include:

  • Doctors have economic incentives to provide inappropriate care
  • Following guidelines will improve care for most patients
  • Doctors don’t keep up with the literature
  • Doctors are unaware of better approaches followed elsewhere

Con arguments (also ordered by percentage of respondents finding them convincing) are:

  • No outside group should come between doctors and patients
  • Doctors will be unable to tailor care to needs of individual patients
  • Guidelines are vulnerable to abuse and corruption
  • Payers will use guidelines to control costs and ration care
  • Guidelines can’t keep up with pace of medical innovation

About half or more of respondents found the average “pro” argument convincing while close to 80 percent felt that way about the “con” arguments. That’s discouraging to people like me who think guidelines are useful. But what caught me by surprise is that even though most people were concerned that doctors provided inappropriate care due to economic incentives and that guidelines would improve care for most people, they came out so strongly in favor of preventing an “outside group” from issuing guidelines and were so concerned that guidelines would prevent tailoring.

However, once I read in the technical appendix how the question was posed I could understand the results. In fact I would have answered the “con” questions similarly to the typical respondent. Here’s the wording:

Some people have proposed establishing an outside group to develop national treatment guidelines based on the latest scientific evidence. Doctors would be required to follow these guidelines when they treat patients. The government and insurance companies would refuse to pay for any treatments not supported by the guidelines even if a doctor thinks this treatment is best for their patient.

The researchers themselves included the emphasis on the words “even if a doctor thinks this treatment is best for their patient.” Instead, I think the researchers should have replaced the italicized words with, “unless a physician documents why this treatment is best for their patient.” Payers do tend to base payments on adherence to guidelines, but there is generally a lot more wiggle room than the authors imply. The authors’ wording certainly explains why people would be convinced that doctors would not be able to tailor treatment!

What the researchers call “guidelines” are more like mandated treatment pathways.

More appropriate wording about guidelines would probably shift results toward the “for” arguments.

David E. Williams is co-founder of MedPharma Partners and blogs at the Health Business Blog.

Submit a guest post and be heard on social media’s leading physician voice.

Comments are moderated before they are published. Please read the comment policy.

  • IVF-MD

    Many medical decisions have no cookbook answer. Option A may have a 80% chance of success vs B’s 40% of success, but option A costs twice as much, results in more pain and has a 3% risk of a certain side effect that B doesn’t have. Patients are individuals and a lot of the management is based on small individual differences, their varying personalities, their tolerance of risk, their fear of pain, etc. Isn’t having rigid guidelines like forcing all football coaches to call plays based on guidelines? OK If it’s third down with fewer than 2 yards to go and the scoring deficit is less than 7 points and there are more than four minutes left in the game and there is more than a 6 mph wind blowing against the offense and your QB rating is in the lower 25%ile and your RB averages over 3.0 yards per carry, then play A should be called.

    I think the folks at IBM who are deservedly cheering Watson’s Jeopardy victory are exploring sending Watson to medical school so that they can replace human thinking with AI and having rigid guidelines would be a crucial part of that process. Would this be a bad thing? Well, healthcare costs would go down a lot if most decision-making were reduced to algorithms and guidelines. I guess I should be grateful to be in a surgical specialty as it’ll buy me some more time before robotics can advance enough to replace me :)

  • ninguem

    “Payers do tend to base payments on adherence to guidelines, but there is generally a lot more wiggle room than the authors imply.”

    Not in my experience.

    Any practicing physicians who actually agree with that?

  • SarahW

    Guidelines tend always to become more rigid than they ought to be. There is a real danger, and the public, who lives in the real world and has experience with such systems, is right to be skeptical.

  • primaryMD


    sure, you’re told you can deviate from the guidelines when appropriate. But those who more thougthfully apply the guidelines will have lower “quality scores” than those who blindly apply the guidelines.

    So if you want to get paid less, and be called a lower quality doctor, then you can use your wiggle room on the guidelines.

    some choice, eh?

  • Zoe Brain

    A particularly egregious example is the “Standards of Care” for Gender Identity Disorder, version 6. An annotated copy is at

    Note that it doesn’t give doses, or even call out medications, there’s been no funding for research on this issue – too controversial. All treatment is based on “accumulated wisdom”, spurred on by the distressing mortality rate when treatment is withheld.

    Note also the “medical” requirement for a “Real Life Experience” – that the patient must somehow manage for a full year to present 24/7 in the target gender (illegal in some states), get documentation changed (impossible in some states), and hold down a job while doing it.

    The real-life experience tests the person’s resolve, the capacity to function in the preferred gender, and the adequacy of social, economic, and psychological supports.

    It’s very difficult to see how a medical decision regarding authorisation of treatment should depend on unemployment rate and local ordnances, unless the object is to protect the doctor from societal opprobrium, not treat the patient.

  • Michael Kirsch, M.D.

    the word guidelines, suggest an optional template fo follow. These ‘guidelines’ however assume the role of a mandate, which was never their intent. They are to be consult in the context of an individual patient’s circumstance, not applied as a ‘one-size-fits-all’ solution. Violating a guideline may be excellent medical care.

  • gzuckier

    A long time ago, a doctor in a managed care organization pointed out to me that his industry operated in the void created by current standards of medical education. The vast variation in treatment modalities which are impressed into impressionable young minds can’t all be correct, and the need to cling verbatim to what they were taught is emphasized much more than any value of “evidence based” skeptical evaluation (possibly with good reason).

    You don’t see “managed engineering” organizations springing up to mediate between purchasers and providers of bridge construction, for instance; because engineers are educated with a pretty rigid slate of evidence based standards. Yes, of course, medicine is inherently far more variable than engineering, but it’s really two points on the same scale, not in opposition; after all, every large building, every bridge, every spacecraft is an individual entity with its own specific characteristics. And in fact, sometimes there are failures, despite said rigorous standards, when some novel factor is not correctly addressed. But there is still innovation in the engineering world, despite the lack of a “let a thousand flowers bloom” attitude.
    Aside for the need for a license, medicine appears to be closer allied to “software engineering” in practice and in education of its practitioners. Anybody who doesn’t remember the joke from a few years back re “If airplanes were designed by Microsoft” or similar ought to Google it up.

Most Popular