nos-trum. pronunciation: \nos'-trum\. noun. Etymology: Latin, neuter of noster our, ours.
1. a medicine of secret composition recommended by its preparer but usually without scientific proof of its effectiveness.
2. a usually questionable remedy or scheme.
See here for more discussion.

Wednesday, July 14, 2010

What Makes A Medicine Too Risky

The emphasis is on "too" risky.  Every single drug has some risk. But do you want to do without them?

The issue has been much discussed recently, based on the FDA's review this week of the diabetes drug Avandia.  The reports seem clear that the drug increases the risk of heart disease (in this case, heart attacks).  It's been around since 1999, when it was first approved.  A previous review by the FDA left it on the market, but with increased warnings.

Diabetics are at increased risk of heart disease, and the drug appears to be effective in controlling blood sugar.  Researchers had to tease out whether the drug itself was causing an increase in risk beyond that caused by diabetes alone.  That was done, and in at least one study (NEJM, June 14, 2007), combining the results of many individual studies, Avandia (or rosiglitazone, the chemical name) DID contribute to additional risk.  The additional risk is not large, but it's real (so we think now).

As with most drugs, getting at the "why" of that increased risk has proven elusive.  Almost every drug on the market has effects on the body that go beyond what they were designed to do.  Avandia raises the bad cholesterol somewhat (LDL, 18%), it lowers hemoglobin levels slightly, it may increase blood volume.  If you have someone who is susceptible to heart disease, any of these--or all--could precipitate a heart attack or heart failure.  We don't really have a way to predict accurately whether any given patient is susceptible or not (given that they don't already have a history of heart attacks or heart failure).

Are there alternatives to Avandia?  Yes, but there's some question--and difference of opinion--whether the primary alternative, Actos, has the same risk...and if it does, is the risk greater or less?  Unknown.  Finally, the history of oral agents for diabetes is littered with drugs that have come and gone for one reason or another...sometimes due to  adverse reactions.

All of this--risk, cause-and-effect, effectiveness, precautions, alternatives--creates a complex decision matrix.  Adding to the complication, it's hard to say how many diabetes are alive BECAUSE of Avandia.  It's unethical to do research that gives treatment (Avandia) to one group of diabetics, and nothing to another, to compare the benefit of that specific treatment.  Untreated diabetes is probably the riskiest situation of all.

In the article referenced above, the authors point out
"The FDA considers demonstration of a sustained reduction in blood glucose levels with an acceptable safety profile adequate for approval of antidiabetic agents. However, the ultimate value of antidiabetic therapy is the reduction of the complications of diabetes, not improvement in a laboratory measure of glycemic control. Although reductions in blood glucose levels have been shown to reliably reduce microvascular complications of diabetes, the effect on macrovascular complications has proved to be unpredictable....the use of blood glucose measurements as a surrogate end point in regulatory approval must be carefully reexamined." [emphasis mine]
Putting the question of Avandia aside for now, is it possible to take a step back from all this and develop guidelines about risk that help us make complex decisions?  Doing that would force us to look at the risks of drugs in general.  Would doing that clear the air on the issues that pull us in different directions?

Going back to the beginning of the article, how much risk are we willing to accept to achieve benefit.  I hope it's clear from the discussion so far, that it's not just a simple matter of "do no harm."  Hypothetically, if you need treatment, discarding a drug that harms one person while thousands of others would have benefited is clearly not optimal, for you or for everyone else.  When you can't predict who that one person harmed will be, do we just forget about them and press on with the "greatest good for the greatest number?"

For those of us who think decisions in life are never yes-no, black-or-white...going down this utilitarian road of "greatest good"  leads to madness.  Philosophers and moralists have been arguing about this since the dawn of time.  If you're willing to accept one harmed, how about ten?  If the one harmed is "great" harm, and the benefit to the thousand is "small?"  Blah, blah, blah....

Rather than join in thousands of years of  argument back and forth, try this:

If you followed the discussion above about Avandia, then you are fully capable of making a decision about whether you would want to take Avandia for your diabetes or not.  Discuss all the information, look at alternatives, see what we don't know for sure yet, etc.  You decide.  Not your doctor, not the FDA, not ObamaCare...although all of these can be sources of advice and information. You should have a doctor you trust, who can work with you to lay it all out.

And just like when you get in the car every day--and risk having an auto accident--you take responsibility for the decision you make  (Actually, driving is a lot riskier than Avandia).

Ultimately, then, what makes a medicine too risky is your assessment, as the owner of your body, that something isn't right for you.  You exercise the freedom to choose.  I've written about what's called "patient-directed healthcare" before.  We need the research, the FDA, your doctor and all those things, plus absolute transparency of information...but when it comes down to it, you should have ownership and direction of what you do and don't do for your health. 

By contrast, what's been called "patient-centered healthcare," a concept championed by the new director of Medicare--a recess appointment by the President--Dr. Donald Berwick, puts the patient at the center of a healthcare system that does a lot of the deciding for the patient, offering to provide what is in the patient's best interest.  All they ask (as de Tocqueville said) is that you not question the "best interest" they offer.  See my post linked above.

Critics of patient-directed care say that patients can't ever absorb fully, or understand completely, in order to decide.   What do you think?  You read the situation above on Avandia; was it too hard for you? that's the assumption of "patient-centered care," "comparative effectiveness," and all the other programs that sound supportive, but assume that you can't decide intelligently.  Someone else--experts--needs to step in and take over for your lack of aptitude.

I don't know what the FDA will decide about Avandia.  Some people may die using it.  But some people die every year from Tylenol.  Maybe they can come up with some precautions regarding patients with heart risk that removes some of those patients that might be harmed.  But nothing is going to eliminate risk entirely, no matter how good a medicine is.

As one senior physician told me, "If you want to avoid risk, dig a grave and get in it."

Doc D

No comments:

Post a Comment


What I'm Reading - Updated 3 May

Blog Archive