-
placebo controlled: The subjects on the drug and the controls get identical pills…
-
randomized: subjects are assigned to groups by a random number generator…
-
double blind: Neither the subjects nor staff know who’s on what until it’s over…
-
protocol: Analytic methods and outcome parameters are defined before the trial begins…
And so to, "there’s something backwards about Randomized Clinical Trials"
This was a low standard. Approval was never intended to tell doctors what to do with the drugs. The RTCs required for approval weren’t the end of anything. They were only a rough beginning – where things started. It was assumed that Medicine itself would take over after approval and figure out how to use the drugs in practice through its collective clinical experience. That’s why my introduction to the last post [the why of it…] was so awkward. I didn’t know how to say where I learn about new drugs, because it is from everywhere – bedside, clinic, colleagues, mentors, conferences, case reports, reps, the PDR, meetings, textbooks, journals, CME, my patients, your patients, even RCTs, etc., and it never really stops. It’s just part of the ongoing stuff of doctoring.
But then along came Evidence Based Medicine [wince]. At first glance, that term ought to be sacrosanct. Who wants a doctor who shoots from the hip? makes wild guesses? Of course, a doctor should base decisions on evidence. But that’s not what Evidence Based Medicine has come to mean. It now seems to mean «backed up by a positive clinical trial». And that’s what I mean by "there’s something backwards about Randomized Clinical Trials." The weakest link, the one most vulnerable to manipulation and distortion, has been elevated to the position of being the gold standard. How that happened is a long story that I don’t know as well as I’d like. But the why and the consequences are painfully obvious.
Since the pharmaceutical and device industries pay for and essentially write-up the RCTs, it gives them an enormous control over what physicians do. Their jury-rigged articles are further distorted in the direct-to-consumer and other advertisements. And even a trial that is honestly analyzed and published is only a crude approximation of the clinical reality of widespread usage. This kind of Evidence Based Medicine based on RCTs gives Managed Care a tool to deny medications or other treatments [and lots of other things]. It reduces the needs of an individual patient to membership of a proxy grouping. It implies a precision and generalizability to medical science never realized [or realizable]. It trivializes and marginalizes the other very important resources that we physicians traditionally use to learn about and prescribe drugs ourselves. And it takes the responsibility away from a doctor to assess in an ongoing and diligent process all the evidence in his/her clinical decision making.
So that’s what’s backwards about Randomized Clinical Trials"
I totally get your aversion to evidence-based psychiatry, but we do need some sort of evidence-based medicine to answer questions like, “on average how often should a diabetic be seen by a doctor?”
Confucius say: Evidence based medicine need medicine based evidence.
Congress said: FDA is not permitted to regulate the practice of medicine.
East,
Absolutely. That’s what EBM is supposed to mean…
Bernard,
“Congress said: FDA is not permitted to regulate the practice of medicine.”
Absolutely! But by the elevation of the Clinical Trial to the sine qua non, that’s exactly what has happened – even though it was hardly the FDA’s intent…
It happened because the road to hell is paved with good intentions. Meanwhile, the professional medical societies fumbled the ball with managed care.
Another aspect of what you describe here is the ascendancy of meta-analyses in lieu of insightful narrative reviews of a therapeutic field. The former are often performed by junior researchers just starting out in their careers, and they are subject to the GIGO principle – of which the young analysts are sometimes naively unaware. But number crunching is perceived to be a higher form of analysis than informed reflection and clinical wisdom.
From the patient point of view, what’s happening is that doctors are ignoring patient reports of adverse reactions because the doctors haven’t seen any papers mentioning them.
(Contributing to this is the likelihood the doctor hasn’t kept up with his or her reading, or doesn’t read widely in the literature anyway.)
Ergo, such adverse reactions are impossible. Something doesn’t exist until it’s studied, and if it isn’t studied, it doesn’t exist.
However, widely advertised claims of efficacy and safety, supposedly based on RCTs, are internalized by doctors.
Have you ever come across the book How to Practice Evidence-Based Psychiatry? I’d really love to hear your thoughts on it, since I was taught EBM in a fashion similar to what that book outlines. I feel that it does a decent job of providing a more nuanced view of how to practice EBM besides just focusing on RCTs.
Dr Nardo
I think this is one of the areas where you (we) need to take ownership of our lack of understanding of the drug approval process. The rules for drug approval are not secret and are well known and have been since 1962. Surprisingly when I went to med school and at the med school that I am on faculty, there is no single lecture on the drug approval process. I give such a lecture to the psych residents and they are usually shocked to learn how the process works. A new drug is not a better drug, its just a new drug, shown in 2 pivotal trials to be better than nothing (placebo). Its a demonstration of efficacy NOT effectiveness. We as physicians should know that. It is our responsibility to be as knowledgeable about the medications we prescribe including the approval process. But do not blame the drug companies for following the rules. They are doing what the law mandates to get an indication. It is our responsibility as physicians to understand that medication and its actual clinical utility. Evidence based medicine is not the RCT, but certainly includes the RCT. WE must read them (or at least the package inserts that have the RCT summarized) to understand the basis of the indication as a starting point to begin to utilize that medication in a real life setting. (Depakote indication for 21 days as an example, who has had pt on longer than 21 days raise your hands). RCTs are one type of evidence and tell us something important, but not everything we need to know. I do agree that all of the pivotal trials attempted to get the 2 positives should be published as fully as the 2 positive trials, that would also be critical evidence in the practice of evidence based medicine. The one failing in the system is that independent funding agencies (read NIH) are reluctant to fund clinical investigations of drugs so that we might have better understanding of the effectiveness as opposed to efficacy. That was the goal by the way of CATIE and STARD and the other large trials you tend to rail against; to try and develop some independent effectiveness data on the use of those medications. Certainly the implementation left much to be desired, but if we as phyisicans want effectiveness and comparative efficacy data we need to find the funding sources to do the types of clinical investigations necessary to generate that data.
Steve,
Thanks for the comment. As you know, I used to teach in that same university. But I didn’t teach the residents this, didn’t even know it was something to teach. So good for you. I’d just clarify that I don’t rale about CATIE so much. I save my raling for CAFE [and, of course, STAR-D]…
Yep, psychiatry really did peak in the late ’80s/early 90s. Then the real greed kicked in, and the shams played out mercilessly. Better living through chemistry, yeah, right…
Its has been clear to me that:
1. Real life is far from a double blind – placebo controlled trial.
2. Approval/efficacy trials with small Ns are essentially meaningless.
3. The FDA is an agency that is interested in conditional rather than unconditional safety.
4. The FDA is an agency that is interested in conditional rather than unconditional efficacy.
5. The political nature of the agency can result in it’s scientific advisory panels being ignored or overruled. The best example was the FDA overruling the scientific committee’s recommendation to not approve high dose hydrocodone in the middle of an epidemic of accidental opioid overdoses.
6. The FDA routinely ignores historical precedence, especially when approving addictive drugs over the approval of their scientific committees. The best recent example is the approval of lisdexamfetamine for binge eating and ignoring what happened when amphetamines were widely available for obesity and neurosis.
7. In the absence of absolutes when it comes to safety or efficacy, the goal of the FDA is clearly to get drugs to market, monitor what happens with widespread use and react accordingly.
As teachers we need to provide this information to residents and medical students and make sure we teach them how to overcome these deficiencies in this agency.
I see the swell job they are doing freaking everyone out about Lunesta and 23andme.com and I have zero faith in the competence of the FDA.