Medical Management of Hyperglycemia in Type 2 Diabetes: A Consensus Algorithm for the Initiation and Adjustment of Therapy: A Consensus Statement of the American Diabetes Association and the European Association for the Study of Diabetes

Response to Woo and Eizirik

We appreciate Dr. Woo's response to our study (1) and his comparison (2) of the Canadian Diabetes Association (CDA) clinical practice guidelines and the American Diabetes Association (ADA)/European Association for the Study of Diabetes (EASD) consensus algorithm and welcome the opportunity to address the differences he has identified.

The original ADA/EASD consensus algorithm and the subsequent revisions were developed independently by the consensus group and were then presented to the two major diabetes organizations for their review and approval of the process. Dr. Woo notes the CDA's process of data review with a “standardized evidence-based approach” and the participation of “over 90 authors and a steering committee of 18” and concludes that the process removed “as much bias as possible,” implying that this approach was superior to our consensus algorithm process. He further notes that the CDA committee “believes that patients and practitioners deserve more choices” and states, with no evidence provided, that “when basal insulin is used with an intermediate-acting insulin, it is very likely that more than one injection will be required.” These statements belie the objective methodologies cited by Dr. Woo.

More important, in our view, is the “equal weight” that CDA gave to all pharmaceutical agents available in Canada. This even-handed approach, presumably to provide the flexibility that CDA believes is important, does not aid practitioners and patients. There are demonstrable differences between these medications in their effectiveness, side-effect profiles, tolerability, ease of use, and costs that were used in the development of our algorithm and that should influence clinicians. The example cited by Dr. Woo, in which he takes the algorithm to task for not recommending glyburide, is illustrative. Our consensus algorithm specifically recommends against using glyburide, despite its glucose-lowering effectiveness and low cost, because it has a higher risk for hypoglycemia than other similarly effective and low-cost sulfonylureas (3,4). In our opinion, this type of information is highly relevant and should inform health care providers and their patients as to which agents should be used. Our goal was to provide useful guidelines for the choice of medications from the many that are available.

Finally, even with the careful evidence-based review carried out by CDA, we must disagree with their conclusion that the Action to Control Cardiovascular Risk in Diabetes (ACCORD), Veterans Affairs Diabetes Trial (VADT), and Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycaemia in Diabetes (RECORD) trials (5–7) have exonerated rosiglitazone of the safety concern raised in previous meta-analyses. The trials were not “specifically designed to help address this question,” despite the CDA's claim to the contrary. Only the RECORD study was designed to address the effects of rosiglitazone on cardiovascular disease outcome, and its interim analysis suggested a trend, albeit nonsignificant, for worse outcomes with rosiglitazone (7,8).

We agree with Dr. Eizirik's concern (9) regarding the potential for dualities of interest to influence the development of algorithms such as ours. This issue potentially affects all of academic medicine. In the current research environment, where academic investigators participate in company-supported trials, it is difficult to find individuals with the requisite expertise and experience who do not have dualities of interest. During the selection of the consensus group members and our deliberations, all of our dualities were considered and discussed openly. Although the results of the process do not directly address Dr. Eizirik's concern, we note with some irony that we have received far more complaints regarding our recommendations from the pharmaceutical industry than from individuals concerned with our potential conflicts. We hope that the stated rationale for our choices convinces readers that the algorithm was developed without the intrusion of any bias other than the shared bias of the consensus group to provide the best care for patients with type 2 diabetes.