"However, the sheer abundance crowds out an important—in fact, the only—skill that matters in treating a patient: how to critically appraise published clinical trials. Few doctors ever read them."

I don't know how far removed from medical school you are, but as a current medical student, I can say this comment is absolutely insane. The skill of critically analyzing clinical trials is taught extensively both at the undergraduate and medical school level...

I think that may have changed lately and may also depend on the medical school you attend (some emphasize "evidence based medicine").

I can also tell you that some physicians are simply not interested in putting up with stats, study design and critical thinking, regardless of whether they may have learned about it or not.

On the one hand, as a physician, I remember being taught how to read studies and use statistical analysis. On the other hand, after you finish the clinical years, and do another 3-10 years of residency (5 in my case), where you are just trying to master the basics, your ability to keep up with all the studies (let alone determine the good from the bad) becomes more limited. The best analogy I have is calculus. I assume, as a med student, you took calculus in high school and again in college. But can you still solve the problems you last saw 4-5 years ago? But you know at one time, you had that skill set mastered...

Not only do doctors read clinical trials; scientists read them too.

I'd imagine that the problem with clinical trials (if it exists) is noncompliance. Its extra work to read papers, and unlike scientists, who are out of a job if they're not up to date, there's little incentive for doctors to be fully informed.

bobbywg:

"However, the sheer abundance crowds out an important—in fact, the only—skill that matters in treating a patient: how to critically appraise published clinical trials. Few doctors ever read them."

I don't know how far removed from medical school you are, but as a current medical student, I can say this comment is absolutely insane. The skill of critically analyzing clinical trials is taught extensively both at the undergraduate and medical school level...

Now, now, let's not be careless here, Student Doctor Bobbywg!

Dr. Sanghavi's comment is not "absolutely insane". While critically analyzing medical studies is taught in medical school, that does not mean it is taught well, nor does it mean that it is learned by the students. In any event, Dr. Sanghavi's primary point (you really should read what you quote!) was that few doctors read the published clinical trials in the first place. Further, his point is that over the whole course of medical training, there is a much much stronger emphasis on diagnosis, and all this time and attention crowds out the more important training in skills needed to manage a patient's illness or condition. You may disagree with his conclusion, but that does not make in an "insane" line of reasoning.

Out in the "real world", I see evidence that supports Dr. Sanghavi's assertions. Many, many doctors focus on simply prescribing medications rather than formulating a real treatment plan. Many, many doctors jump on the latest drug to come out without critically assessing whether there is any valid evidence the new medication is any better or safer than the current medications we use. One example: I see many of the local cardiologists starting patients on Crestor first for lipid lowering, even though there aren't any studies that show evidence of clinically relevant benefits--lowered risk of heart attack, stroke, death, etc. None. Nada. Zippo. They ignore medications (e.g., Pravachol, Zocor) with multiple studies showing clinically relevant benefits and a proven track record of safety when treating high-risk patients with high cholesterol and just skip to the "latest and greatest". Now that is what I call "absolutely insane"!

NNT post ACS to prevent one death:

Atorvastatin, simvastatin, pravastatin - 95 (two years)

Co-Q10 - 4 (one year)

How many physicians know that?

Have you read those studies and contemplated them and decided whether the Co-Q10 article results are worthwhile emulating?

The ONLY way a physician can keep up with the literature is to rely on summary services of the literature, co-practice with a clinical pharmacist (who in my experience does not keep up either), use a very adept and customizable EMR (Praxis) and/or rely on an organization like the VA to cover the team/nurse/EMR/review contingencies. And look in some VISN's how the VA insanely restricts ARB's based on a very outdated set of practice guidelines, another example of a whole system unable to keep up like this great cardiologist author expects each physician to keep up.

So, there are 160,000 private practices in primary care that are not able to do the above unless they have a super-customized version of Praxis EMR (voted #1 by AAFP) - I have NO financial interest. Boston mega-practices are represented by THREE people on the national group charged to develop the NHII standards applicable to EMR's. Small private practice physicians have NONE. How about that representation for primary care physicians in the trenches?

cbday:

The ONLY way a physician can keep up with the literature is to rely on summary services of the literature, co-practice with a clinical pharmacist (who in my experience does not keep up either), use a very adept and customizable EMR (Praxis) and/or rely on an organization like the VA to cover the team/nurse/EMR/review contingencies. And look in some VISN's how the VA insanely restricts ARB's based on a very outdated set of practice guidelines, another example of a whole system unable to keep up like this great cardiologist author expects each physician to keep up.

I think you are making a good point, although you should rather write: The ONLY way a clinician can keep up with the literature relevant for his practice (unless he is an absolute subspecialist) and come to correct conclusions is to rely on summary services of the literature.

Most physicians don't like to hear it, but we need more cookbook medicine and easy guidelines (there are enough exceptions and complicated situations where you can't apply them anyway) ... and these guidelines should be written by professional organizations that are supported by membership fees only.

It is self evident the the rate of adoption of new recommendations from recent medical literature is slow but that it not necessarily a bad thing. Medical science is often influenced by fashion, trending, funding chasing, and political motivation as much as anything else. Just because a particular study is the newest doesn't mean that its conclusions are any more valid than those of the preceding literature. In medicine it is not uncommon for what is obviously true one day to become patently untrue the next day and then once again become newly discovered in years to come. Look at mastectomies, stents vs. coronary artery bypasses, and the treatment of ear infections. Changing your practice just because of the most recent study means that you are wrong most of the time.

Now, now, Maryland MD, let's not make ourselves seem more patronizing and arrogant than we have to - it demeans the profession (and diminishes your credibility). If your argument is sound, you should be able to make it without resorting to the rank-pulling and snarky asides.

That issue aside, I agree that this is a problematic article - some of the basic points are quite valid (familiarity with the primary literature is important, review articles should be unbiased, treatment should be thorough and evidence-based) , but the Dr. Sanghavi goes overboard, both in tone and substance, in what he demands of primary care physicians.

He says that they should be familiar with the most up-to-date evidence on all treatments they provide, but then disparages (with vague intimations of the evil inherent in being a "private company") some of the sources that best provide that information. And then says that they should be able to discuss the evidence base of their treatment, (relying purely on their own analysis of the primary literature) whenever a patient asks for it. Apparently he feels that every primary care doctor should be able to quote on demand the specific studies that specifically justify each medication he or she prescribes. Not merely quote them, but give a dissertation on their relative strengths and weaknesses, how they compare to studies that showed different outcomes, and so forth. In other words, be able to write their own "Up-to-date" article on every condition they run across. It'd be wonderful if everyone could do that, but it's just not possible.

It may be possible for an academic sub-specialist such as Dr. Sanghavi to give such dissertations on every condition that he sees in his practice. But a) he sees a very limited number of conditions, and b) has dedicated, paid time for keeping up with the literature in his very narrowly focused field. But I would challenge him (and you, Maryland MD) to give a well-informed discussion, right now, about the optimal medication regimen for a 64 year old female type II diabetic with chronic renal insufficiency, hypertension, glaucoma and Graves disease who presents with her third case of community acquired pneumonia this year. Discuss the various drugs available, their dosages, side effects, NNT, and how each drug would interact with each of her conditions. Include in your discussion citations of all relevant primary literature (along with your own analysis of the strengths and limitations in each paper), a discussion of current clinical controversies, and a brief analysis of the quality-adjusted life years per dollar spent on treatment that this patient can expect to receive from her medication regimen.

I bet Dr. Sanghavi could pretty much do just that for a 12 year old male with a VSD, Factor-5 Leiden deficiency and long QT-syndrome. And evidently you could do that for the statins. Maybe you could do it for a couple other drug classes as well. But how many? 10? 20? 100? There's a whole lot of medical conditions out there, and a whole lot of possible treatments. And a huge amount of literature, some of it good, some of it not, coming out every week.

How does Dr. Sanghavi expect anyone to sort through all that literature, for every disease they might treat, and remember it all? It's not possible. And I don't see how Dr. Sanghavi, from his rarefied position, can presume to tell other doctors that they ought to be experts in every condition that comes through their doors.

In fact, what they really ought to be able to do is diagnose (or develop a pretty good sense of the direction of the diagnosis) the diseases that a patient walks in with. And that is the hard part, the part that a computer program can't do, that requires the years of training and medical school. As for treatment? Well, if it's something common to their practice (and I will certainly grant you hypercholesterolemia in primary care or cardiology) then yes, they should be familiar, through their own analysis of the literature, with what works and what doesn't. But for everything else, why not use the resources that are available in this country, in this day and age? For simple stuff, I don't see why that can't include turning to a reliable literature review, including up-to-date. For complex stuff, it means referring to a specialist like Dr. Sanghavi. The critical part, I would argue, is making that diagnosis in the first place. Once that's done, getting to the correct treatment is really not that challenging (heck, most patients can look it up themselves).

I would also disagree with Dr. Sanghavi, admittedly based on my own anecdotal experience (but that seems to be the basis of his assertion as well) that medical education is skewed wildly in favor of diagnosis at the expense of treatment. Medical school itself does focus largely on diagnosis, pathophysiology, and so forth - this is appropriate, as understanding these things forms the basis for understanding and applying treatment (and is part of what allows one to critically review the literature in the first place). But residency education, at least in my experience, focuses very much on treatment - that, indeed, is probably the main educational focus of residency. Which treatments to use, in what circumstances, what the risks and benefits are, what the evidence base is, and so forth - that's the majority of what gets focused on, both clinically and didactically.

In summary, I think Dr. Sanghavi makes some reasonable basic point, but he goes much to far with it. He sets unreasonable (impossible, actually) standards for practice; maligns an the entire primary care community with isolated, anecdotal examples (too many tonsillectomies in a Vermont town), and implies that any medical opinions not published in a peer-reviewed journal must be the work of incompetents and pharm-industry hacks. He and you, Maryland MD, may have legitimate points to make, but first you both need to lose the holier-than-thou posturing.