Wednesday, January 05, 2005

Drugs, risks, benefits, and honesty

A review in the 4 December BMJ (BMJ 2004;329:1317) finds that when we get right down to it, NSAIDs (both traditional and COX-2's) have minimal effect over and above placebo for osteoarthritic pain. Aside from being humbling (but not overly surprising, really), it caused me to think about the role of hope and expectation in health care and also the role of alternative medicine.

As others have recently pointed out on this blog, all the pain meds can be bad for us. As it now seems, they aren't particularly good for us either, leaving one to wonder whether the balance does in fact lie on the positive side of the scale. Perhaps the answer is to be a bit less rosy-eyed in our own tendency to overestimate the benefits and underestimate the risks of what we prescribe, and to inform patients honestly about the magnitude of benefit and risk so they can make their own decisions. (That's really what got Merck into trouble over Vioxx: not that it caused harm, but that they knew it and concealed it while simultaneously inflating claims of benefit.) But then if we tell patients "this stuff can eat your stomach, might give you a heart attack, and doesn't really do any more for your knee than those M&M's over there", how will patients respond? People come to us for doctoring, not just technology. The pills we prescribe represent hope: the doctor prescribed them, and they are powerful "medicine" (in the sense that my Chippewa ancestors meant it, not in the sense pharmacists mean it).

If we're addressing people's strongly felt (and I believe absolutely legitimate) need for doctoring by prescribing something that can hurt them, costs a lot, and doesn't really do much, and we're not leveling with them about that, are we committing the quackery we accuse alternative providers of? Hmm. I'm not about to start defending the nutbars of pseudoscience, but if we're scientific we have to apply our standards ourselves too, eh?

I recommend glucosamine a lot lately. It doesn't work any less well than NSAIDs or COX-2s, it doesn't plug arteries or eat gastric mucosa, and it's pretty cheap. At bottom, I have to respect what the patient came to me for. If it's almost a placebo... let's not knock 'em, just pick safe ones.

2 comments:

Egan said...

Well stated.

In my experience most of what happens in the exam room more deals with perceptions and not realities. The power of placebo is great indeed. Just putting my steth on a patient has value to the patient. When trading notes with colleagues in primary care it has been my personal experience and the experience of others that we see 80% depression 20% physical pathology when you really come right down to it.

The "painful shoulder" is not the "real" issue. It is her abusive mother-in-law, dysfunctional spousal relationship, inability to cope with an overbearing boss, you name it. We don't see a "normal", in terms of coping skills, cross-section of the population in adult internal medicine primary care practice. It is has been my feeling that if physicians spent more time listening (the opposite occurs in studies I recall, how long does it take the doctor to interupt the patient with her first question?) less drugs would be prescribed, less studies would be ordered, and hence, less would be spent on the high ticket items. Ah, but that would mean paying more for cognitive care providers (to pay for longer visits)...ain't going to happen.

Egan

InformaticsMD said...

Agreed. Drugs that we know work to a significant degree (narcotics, antibiotics, antihypertensives, anxiolytics, to name a few) are fine for the appropriate situations [and the appropriate safeguards]. However, drugs whose effects are far less apparent should not be dispensed like candy.

Earler in my career, in occupational medicine in a regional public transit agency, I dispensed NSAIDS often for actual injuries one encounters in such a setting - e.g., back and limb sprains and injuries from heavy railroad track work and mechanical maintenance and repair of buses and trains and signal repair, degenerative problems, fractures, etc.

I never heard anyone jumping for joy about the effectiveness of the NSAID medicines over something like short-term percocet or similar - which I rarely prescribed. However, gastric upset and GI bleeds I did see, which really didn't help teh patients at all.