Making Sense of the ACCORD Study: Doctors Should Treat People, Not Numbers
Much to the dismay of scientists, policy makers, and health care administrators, good medical decision making is not always black and white. I understand and sympathize with our desire to distill complex disease management issues into specific, easily measured variables. But unfortunately, the human body is exceedingly complex, and willfully resists reductionist thinking.
The recent ACCORD trial (which was designed to quantify the value of aggressive glucose management in a diabetic population) actually demonstrated a higher mortality rate in the intensive treatment group. What? That’s right, people were more likely to die if they had been randomized to the group that used all means necessary to keep blood sugars in a near normal range.
Now, this does NOT mean that it’s a bad thing for diabetics to keep tight control of their blood sugars, but it MAY mean that if they have to take high doses of multiple drugs to get them to that aggressive goal, the negative drug side effects may collectively outweigh their benefits.
I spoke with Dr. Zachary Bloomgarden, a renowned diabetes expert, to discuss his interpretation of the trial results. Here is a snippet from our interview:
My feeling is that this study shows that there is an art to medicine, and that patients can’t be managed via cookbook methods to treat their disease. If a person can control their blood sugar to an A1c of 6.0 without using too many medications, then that might be a good goal for him or her, but if you have to take high doses of several pills to get to that same goal (and therefore experience all the unfavorable additional side effects from taking them like weight gain, fluid retention, and potential arrhythmias) then it might not be appropriate in that case.
Ultimately, it takes a personalized approach by an experienced physician to determine the best treatment plan for an individual patient. One size doesn’t fit all - that’s part of my
take away from this study. We still
certainly want all people with diabetes to do as well as they can with blood
sugar as well as blood pressure, cholesterol, and the myriad other markers of
control of the disease.
And so my plea is that in our race to ensure “quality care for all” in this country, we take a moment to consider that real quality may not be about getting every patient to the same blood test target, but to get every patient to a primary care physician who can apply evidence based recommendations in a personally relevant way. Cookbook medicine is no substitute for good clinical judgment. Let’s invest in our primary care base, and make it financially viable for them to spend the time necessary to ensure that their patients are on individually appropriate therapeutic plans. I hope our next President will appreciate the critical role of primary care in a healthy medical system.
Addendum: a like-minded fellow blogger weighs in on the study
.This post originally appeared on Dr. Val’s blog at RevolutionHealth.com.
I think there are two other dimensions to this study. One is technical and the other has to do with who is treating who.
The difference in mortality appeared to be on an “intention to treat” basis. In other words, being targeted
to an A1c < 6% (with aggressive treatment) led to the increase in mortality over four years. That’s not the same as actually having an A1c of 6% or less.
As a physician I appreciate that fact that “doctors treat patients,” but when it comes to diabetes mellitus it’s really the patients who treat patients. Thanks to ACCORD, patients have a key piece of information to use in reconciling their preferences and values in their own diabetes management.
More available at:
http://diseasemanagementcareblog.blogspot.com/2008/02/disease-management-blog-d oesnt-presume.html