What would be better than a RCT? Answer- a RCT that makes a comparison between realistic alternatives. This issue was the subject of a insightful editorial in the April 12, 2006 issue of JAMA.(Recent Trials in Hypertension-compelling science or commercial speech;JAMA, vol 295, no. 14, p1704-1706)
Three researchers well known for their expertise in this area authored the piece,B.M. Psaty,N.S. Weiss and C. D Furberg.
Full text requires subscription. Here are some highlights.
We are all becoming increasingly aware that a RCT comparing two drugs does not mean squat if the comparator drug is given in too low a dose or is one that no one (or hardly anyone) would choose anyway.
Is one treatment superior to another? Lets do a RCT. O.K. but " a fundamental principle of active treatment-controlled trials is the scientific obligation to use the best available treatment in the control group."
The editorialists were addressing the issue of blood pressure active treatment trials (an active treatment trial compares one drug with another- not with a placebo) and point out that in the ASCOT trial, low dose diuretics were not used in 45% of the atenolol group.Further,the ASCOT investigators acknowledged that possibly atenolol was not the appropriate comparator drug
for contemporary BP treatment. (Atenolol has fallen into some disrepute in that regard)
One point is that one could argue that low dose diuretics would have been a more appropriate comparator drug and atenolol is not the best beta-blocker to use as well.
Comparing drugs or sequences of drugs in hypertension trials is not simple.Many- if not most- patients require two or more medications to achieve control and the protocols in active comparisons trials may not reflect the way BP is really treated in the non-RCT world in which real physicians live. The highly hyped and highly criticized ALLHAT trial has also been accused of that fault.
The authors point out that although atenolol does not have a great track record in decreasing cardiovascular disease mortality in treated hypertensive patients, 3 large recent industry-sponsored trials used atenolol as the comparator drug.
The editorial quotes a recommendation of the National Heart,Lung and Blood Institute Working Group proposing a comparison of several drugs in patients who were receiving low dose diuretics, i.e. adding on an ARB, or an ACEi, or a CCB or a beta-blocker (but not atenolol) and doing a head-to-head comparison.
1 comment:
Excellent post. I love your discussions of EBM.
One problem, if I understand the FDA phase III trial regulations correctly, is that Big Pharma is not required to test a new product against a good drug already on the market - it just needs to test against a placebo. Because most clinical trials in the world are from the US, and because most of these that are large and of good quality are Big Pharma-sponsored, for most new drugs we're only ever going to see a comparison with a placebo. Once that trial is wrapped up and the drug approved, the work of convincing doctors to prescribe the new expensive med over the older cheaper one is accomplished via marketing. A trial comparing the candidate drug to good controls is only ever risked in situations where the older drug has such a large market share and is so widely prescribed that the company believes it will need actual data to sway doctors to make a change, in addition to their notorious marketing tricks. And when such a trial is embarked upon, it is always designed in such a way that all kinds of "possible" benefits of the new drug and/or harms of the old drug may be "discovered", usually via underpowered subgroup analysis. And so they can come out saying:
"Our new MiracleConcoctionThatYouJustHaftaPrescribe is *as effective* as SimpleOldCheapDrug, but without the XYZ side effects of SimpleOldCheapDrug!!!"
Never mind how dubious the evidence for the s/e is, or what s/e the new drug might have, or how unimportant the s/e might be in the subpopulation you're treating, etc. etc etc...
Post a Comment