Is Evidence Based Medicine Based Medicine The Holy Grail?
Stanley Feld M.D.,FACP, MACE
“Reanalysis of an ACCORD study refutes link between intensive therapy, higher death risk from Cardiovascular Disease”
Most of my readers know I am a retired Clinical Endocrinologist. I practiced Clinical Endocrinology for 30 years. I do not have a billfold interest in the healthcare system. My goal is to influence policy makers to develop a healthcare system that will work and decrease costs.
During my practice of Clinical Endocrinology I became involved in Clinical Research. My involvement forced me to understand the nuances of clinical statistics. I gained an appreciation of the defects of statistical analysis in clinical research.
When I was Vice President, President elect and then President of the American Association of Clinical Endocrinologists I was Chairman of AACE’s Diabetes Guidelines. I also co-chaired the guidelines for Thyroid Disease, Thyroid Nodules, and Thyroid Cancer.
The process of writing the guidelines increased my appreciation for both excellent and shabby clinical research.
The AACE Diabetes Guidelines are a “System of Intensive Diabetes Self-Management for Type 2 Diabetes Mellitus.” Clinical endocrinologist knew that in order to avoid the complications of diabetes mellitus, patients had to be taught how to manage their disease. It was a team approach. Patients had to be taught by physicians and their healthcare teams to be responsible for their day to day blood sugar control.
The AACE Diabetes guidelines were an outline of a process of care. Over the last 15 years many new oral drugs have been introduced to control blood sugar in Type 2 Diabetes. The concept of patients’ responsibility for the self management of their disease must not be de-emphasized.
Patient responsibility has not been included in President Obama’s healthcare reform bill.
In the last few years there have been studies that have discredited the concept that normalizing the blood sugar would decrease diabetes complications. One such study was a study call ACCORD “The Action to Control Cardiovascular Risk in Diabetes”. The study found patients in the intensive control of blood sugar group had a higher mortality from Coronary Artery Disease than patients in the control group.
This conclusion was counterintuitive to previous evidence, evidence I believed to be statistically significant. However, I understood the defects in Accord’s statistical analysis at publication. The journal referees did not pick up the defects
The media publicized the findings undermining the significance of the progress in the culture change toward tight blood sugar control in Type 2 Diabetes Mellitus. Patients were confused. They decreased their effort to normalize their blood sugar.
I wrote extensively about the defects in the statistical analysis of the Women’s Health Initiative (WHI). My critique fell on deaf ears. None of the finding in the WHI were statistically significant.
The conclusion that estrogen therapy was dangerous to women’s health was misguided. The costly NIH study was poorly designed and did not prove there is an increased incidence of complications statistically.
It might be that estrogen therapy is dangerous to the health of women. This study did not prove it in a statistically significant way.
The study did change the course of women’s health in this country forever. None of the academic leaders spoke out to refute the study statistically.
The WHI is the Holy Grail for evidence based medicine in women’s health. I think the WHI has been a setback in treatment for post menopausal women 50 years and older.
The Accord study has been a setback for the concept of intensive self management and tight blood sugar control for Type 2 Diabetes Mellitus.
President Obama’s healthcare reform bill is in the process of creating bureaucratic review boards that determine the standards of evidence based medicine.
Some clinical studies are excellent and should set the standard for evidence based medicine. It should be the way physicians practice. President Obama’s boards should be educational boards for physicians. Instead they will be punitive boards focused on reducing reimbursement.
Some evidence based medicine is shabby, poorly designed, and not statistically significant. Those studies conclusions should not be forced on the practicing medical community.
“The Action to Control Cardiovascular Risk in Diabetes (ACCORD) study was abruptly halted in 2008 when researchers noticed an increase in deaths in the group of Type 2 diabetics being intensively treated to bring their blood sugar levels down to near-normal levels.”
Termination of the study received wide coverage by the traditional media. No one critiqued the design of the study or the reasons for its statistical outcome.
Two years later the investigators have realized the errors in the study.
“ A new analysis of data from that study has concluded that a rapid lowering of blood sugar (glucose) levels was not the cause of the increased risk of mortality.
However, even though the researchers were able to exclude intensive diabetes management from their list of suspects, they still aren’t clear what factors may have played a role.”
I applaud study author Dr. Matthew Riddle, a professor of medicine at Oregon Health and Science University in Portland, for further analyzing the data.
"The original question underlying the study was,
would rapid lowering of blood glucose be an explanation for the excess mortality rates?"
"The answer was no.
People who rapidly lowered their A1c didn’t have excess deaths," Riddle said. In fact, "it was the ones who couldn’t bring their A1c’s down that had increased mortality."
This is the way errors in clinical research findings have been discovered in the past.
President Obama’s quest to reduce government spending will result in decisions about evidence based medicine made by bureaucratic panels before the data is tested and retested. My fear is the standards of care will be changed to reflect new evidence too swiftly. The change will be to the detriment of advances in patient care.
The opinions expressed in the blog “Repairing The Healthcare System” are, mine and mine alone.