Voodoo Medical Science

Where to begin, where to begin.  I get busy with end-of-semester things and head out of town for a few weeks and poof! Bloody chaos.

Women’s reproductive rights suddenly back on the table in the U.S. and the legality of abortion tabled in the House of Commons here as a private bill.  Good grief. Was that plane I took the one in that Twilight Zone episode; the one that goes through the clouds and goes back in time?  More idiocy in the Commons, with this ludicrous Omnibus bill as they’re calling it.  Long guns taken out of the registry which means that automatic weapons can more readily be sold in Canada.  And of course zombie killers. (OK, that last one was ghoulishly interesting, I have to confess.)

And in health care news, as always some bright lights insisting they know what’s best – most recently a report from researchers at McGill (the term researchers usually being code for statisticians) expressing shock, shock I tell you, that drugs are used off-label when this lacks “scientific support”.

Um, OK. So what scientific support would that be? Drug company funded clinical trials – given that all other funding has been cut to the bone? Or do they actually mean data which, I would remind you, does not equal knowledge and can be massaged, manipulated and moulded to fit the theory-du-jour.

One class of drugs these experts took exception to was the use of anti-psychotics in situations where no clinical trials had been done. Years ago a physician friend of mine discovered that one of the anti-psychotics, quitiepine I think, seemed to help a patient with Huntington’s with some of her more onerous symptoms. But of course Pharmacare wouldn’t pay for it because – yup, you guessed it – there was no “evidence” that it worked for Huntington’s.  And naturally we all know that everyone, especially drug companies, are lining up to do an expensive drug trial with a teensy subgroup of patients with a rare, fatal, genetic disease ….

Needless to say, there’s never going to be “scientific support” for this. A point these McGill researchers who’ve clearly never had to deal with an actual patient don’t appear to have twigged to.

Research, clinical trials are expensive, time consuming and difficult to do. Who in their right mind is going to fund one for an old drug that’s no longer on patent that’s been around forever – but that still helps a lot of people? Not going to happen.

The pendulum has so swung so far, moreover, in favour of the stats and the algorithms and the “evidence” that everybody from Obama to your pharmacist to that nice young doctor in the clinic down the road honestly believe that medicine is a science and if we could just figure out the right questions to ask and do the right research (which  angels – taking time out from their dancing-on-a-pin thing – would fund) then All Would Be Revealed and we would all live happily and healthily ever after. As if.

What few people realize, alas, is that the bedrock of “scientific” medicine, the clinical trial, is very recent –though to hear people ramble on about  it you’d swear the dratted thing was on one of those tablets Moses brought down with him.

1948. That’s when the first official clinical trial was conducted: by the first medical statistician on record, Bradford Hill, who gave one group of patients with TB streptomycin (then a very new drug) and another group nothing. The idea took off and before his death in the 1990’s Hill’s book on medical stats (Principles of Medical Statistics) was in its 12th printing.

Hill was no dummy though and realized he’d created a monster. He backtracked. Where once he’d exhorted statisticians to “rise from their humble place” to help medicine become more scientific through the clever application of numbers he suggested we should “relax and reflect”; that such single-mindedness could easily lead to poor patient care: “cookbook medicine”.  It would be better, Hill wrote, if clinical trials were designed to “promote rather than hinder the traditional method in medicine of acute observation … by the clinician at the bedside”.  (All quotes from Richard Horton, the editor of The Lancet writing in 2000 in the journal Statistics in Medicine, “Common Sense and Figures: the rhetoric of validity in medicine” Vol. 19, pp 3149-64)

Probably what Hill had not appreciated in the early fifties as he began his little crusade was the extent to which post-war enthusiasms, technological advances and various social, political and institutional changes – ranging from the ascendancy of the United States to the shifts in finance, corporate influence and law – would transform his notions into a paint-by-numbers fiasco.  Biomarkers and surrogate end points (blood sugar, cholesterol, blood pressure, bone density) would reign supreme and you could feel perfectly fine but be told you had minutes to live.

Then it was EBM guru, David Sackett who picked up where Hill left off. Ably assisted by the new profession of health economics whose sole purpose it was to assist payers (like HMO’s in the U.S. and governments of countries with public health care) cut costs (and realized this statistical scientific rhetoric could aid their cause), the newly minted evidence-based medicine or EBM took off like wildfire, leading to the proliferation of guidelines and Hill’s cookbook medicine.

Sackett also backtracked, emphasizing that “the practice of evidence based medicine means integrating individual clinical expertise with the best available clinical evidence” everybody pretty much ignored him. After all, who cared if patients were different and physiology, difficult; as long as you had your bullet form guidelines and decision trees.

Meanwhile, everybody forgets that evidence has serious limitations, not the least of which is human error, external validity (in other words the people in the trial are not representative of the people in the community who actually take the drug or use the treatment) and conflicts of interest. At best even the best designed of trials tend to encapsulate a narrow slice of life which is not the reality of medical care which tends to be centred around the elderly and those with chronic conditions. (duh)

The old and the sick, precisely the people who use medicine, are excluded from clinical trials; in fact as Bradford Hill pointed out, the clinical trial “at best shows what can be accomplished with a medicine under careful observation and certain restricted conditions”.  I won’t even mention the exclusion of women from trials until the NIH stepped in in the ‘90’s to enforce its own regulations because the top of my head would blow off and that would create such a mess.

Ironically, where scientific medicine and epidemiology do excel is at giving us clues as to what doesn’t work, e.g., in common preventive measures such as mammography and PSA testing. But we don’t like those recommendations so we ignore them.

Turns out the “science” of medicine is like the Sasquatch. Often sighted and excitedly talked about but not actually real.