Rationing through research

In a September 2 article out of Reuters, HMO executives lament that they are being asked to pay for medical services that have never been verified in carefully conducted clinical trials.  The lack of clinical trials translates, they imply, into millions of dollars wasted on unproven procedures and treatments, and is driving up the cost of care.  By seizing the scientific high ground - demanding well-documented proof before agreeing to pay for a service - they say they can improve medical care and drive down spending.

Doctors, of course, always the obstructionists, tend to object to such an approach on the grounds that medicine is largely an art, and not a pure science.

That's just the problem, the HMO executives counter - "physician autonomy is one major barrier" to taking a more reasoned, scientific approach to reimbursement decisions.

Rich comments:

I am going to ignore the implication that insurance executives are still worried about physician autonomy, when they have long since solved that "problem," and will address instead the rationale of insisting on randomized trials before paying for health care services.

Insisting on randomized clinical trials that prove the efficacy of a new medical test or treatment, while on the surface being a scientifically unassailable position, is in practice a powerful method of delaying or even denying coverage of useful medical services.  Randomized trials are extremely complex and expensive animals.  They take years to design and conduct, and are of such a nature that seemingly minor details in their design can utterly and unpredictably change the outcome once the trial is completed.  Because of this latter feature, if someone (a professional society, say, or an insurance carrier) does not like the outcome of a randomized trial, it is always a trivial thing to point to some controversial design feature of the trial and say, "There - that's the reason for this crazy result.  We can't believe this data - we need a new trial!"  This way, inconvenient results can be ignored altogether, or at the very least delayed for years and years.

Example: DrRich spent a large chunk of his career as an electrophysiologist arguing that randomized trials were not needed to prove the efficacy of the implantable defibrillator.  He (apparently misguidedly) believed that the efficacy of this device was so dramatic (e.g., a patient walking down the street has a cardiac arrest, collapses and loses consciousness, then 10 seconds later automatically receives a shock from his implantable defibrillator, gets up, and continues on his way) that a randomized trial was not only unnecessary but would be unethical (since it would require identifying a large group of patients who were at high risk for sudden death, then systematically withholding the defibrillator from half of them.)  Because such arguments were "unscientific" and because insurance companies insisted on randomized trials or they wouldn't pay for this device, a series of randomized trials with the defibrillator was indeed conducted over the last decade.  Not surprisingly, all of them showed that the implantable defibrillator saves lives. 

So are insurance companies now paying for these devices in patients who have a high risk of sudden death?  Not really.  They reluctantly pay for defibrillators in a few small subsets of patients, but because they "don't like" the design of some of the randomized trials, they are holding out for "better data" before agreeing to pay for them in the majority of patients who probably ought to be receiving them.  And even if by fiat they were required, beginning today, to pay for these devices for all high risk patients, by insisting on clinical trials they have still successfully avoided having to spend their money for more than a dozen years.

Another example: The Reuters article laments the fact that, due to a lack of clinical trials, nobody really knows the best way to treat prostate cancer, and thus the insurance companies don't have a scientific basis for agreeing to pay for various treatments.  The fact is, prostate cancer is being attacked simultaneously on multiple fronts - various surgical approaches, various radiation approaches, various chemotherapy regimens, and sundry other methods such as cryoablation (freezing the tumor).  Each of these approaches is evolving rapidly based on what doctors are learning and on advances in pharmacology and technology.  The fact is, nobody could design a meaningful randomized trial today for the "optimal" treatment of prostate cancer - the treatment options simply aren't "ripe" enough to do so.  Even if one or two of the current approaches were clearly "in the lead" and were chosen for testing in a randomized trial, during the 5 - 6 years it takes to conduct such a trial these approaches would certainly become obsolete due to continued technological advances.  The results of such a trial would be clinically meaningless.

But doing such a trial would give insurance companies plenty of cover for withholding any treatments they deemed too expensive in the meantime.

And that, one suspects, is the whole point.

September, 2002

Back to news and commentary

YourDoctorintheFamily.com Home Page