The Real Lessons from Accord

Reprinted with permission of Amy Tenderich at DiabetesMine.com

Rich has a very impressive new title at Joslin Diabetes Center in Boston; he's Director of Medical Affairs, Healthcare Services, and Strategic Initiatives, which would be totally intimidating if he weren't such an easygoing and approachable person. We became friends writing a book together last year, and I am delighted to be able to tap his deep knowledge on diabetes drugs and treatments.

A Guest Post by Richard Jackson, MD

We keep hearing about the recently published ACCORD study, and how tight blood glucose control might not be that important - or even risky.  What should your typical Type 2 patient really be taking away from these studies?

Although the reported results from the ACCORD study may have been unexpected, the overall messages are fairly clear, and they are good news.  For those of you with pressing time commitments who wish to cut to the quick, there are two main takeaway messages:

*  First, the incidence of heart disease in patients with Type 2 diabetes continues to decrease.

* Second, if you are older, with longer duration of diabetes, with multiple risk factors for heart disease, and have a high A1C, you should not be using multiple glucose lowering therapies to achieve an A1C of lower than 7.0.

To elaborate: The main message demonstrated by the ACCORD study is that the incidence of cardiovascular events in people with Type 2 diabetes is steadily decreasing, and decreasing faster than the experts have expected.  This isn't the message that appeared in the media, nor was it the focus of the expert discussions at the American Diabetes Association meeting in June.  However, this message has the most impact for people with Type 2 diabetes.

The ACCORD study addressed the question of whether aggressively pushing for an A1C target of under 6.0 would reduce serious heart disease and deaths from heart disease, when compared to a control group treated to the currently accepted A1C target of 7.0 or less.  The primary result was unexpected; the study was stopped early because the people in the more intensively treated group had significantly more deaths, including deaths from heart disease.  This result received the most press, but I find it the second most important message, and here's why: The differences in deaths was relatively small, 14 deaths/1,000 patient years in the intensive group vs. 11 deaths/1,000 in the standard treatment group.  This difference was statistically significant, however, so the study was halted by the safety monitoring group that oversees every big clinical study.

However, the most impressive number of all was not directly quoted in the paper.  When planning a study such as this, calculating the ‘expected event rate' is one of the most important steps.  In the ACCORD study, this step involved estimating, as accurately as possible, using the best data available at the time, the expected incidence of heart attacks and deaths from heart attacks in this particular high-risk patient population. This was necessary so that researchers could plan how many people with diabetes they would need to follow, and for how long, to determine whether there was an advantage to tighter control.  Experienced biostatisticians, using the best data available, calculated the expected event rate to be 55 deaths/1,000 patient years.  AMAZING! Both the intensive treatment (14/1,000 patient years) and the standard treatment group (11/1,000 patient years) had a much lower incidence of severe heart events than predicted.