The American Diabetes Association and the European Association for the Study of Diabetes recently produced a position statement and update (1,2) and the American Diabetes Association produced a standards of medical care for the management of type 2 diabetes (3). These documents place shared decision making at their core, and as clinicians, we applaud this sentiment.
A recent systematic review found that when evaluating treatments and tests “the majority of participants overestimated intervention benefit and underestimated harm” (4). Therefore, front-line clinicians need the following information to undertake shared decision making:
When it comes to glucose control, we need to provide patients with ballpark estimates of harms over a specific time period associated with specific glucose levels as compared with “normal” glucose (e.g., cardiovascular disease, renal disease, vision changes, and symptoms).
We also need to provide, based on the best available evidence, a reasonable idea of the impact, or lack thereof, of glucose-lowering treatments on these end points and the magnitude and nature of treatment harms.
To that end, we searched through the American Diabetes Association and European Association for the Study of Diabetes documents (1,2) (∼20 pages) and American Diabetes Association Standards of Medical Care in Diabetes (3) (∼93 pages).
Risk Estimation for Clinical End Points Without Treatment
There was no mention or discussion of the magnitude, in relative or absolute terms, of any adverse clinical end points associated with elevated glucose. In addition, there was no discussion about risk tools or reviews to help clinicians to make estimates. Although they have limitations, there are tools (5) and articles (6) that offer methods for estimating risks associated with levels of glucose.
Impact of Treatment for Hyperglycemia on Important Clinical End Points
There was no mention of the magnitude (real or potential) in relative or absolute terms with regard to retinopathy, kidney disease, or neuropathies.
All cardiovascular disease (CVD) benefits were presented with relative numbers: a nonstatistically significant “16% reduction in CVD events” (3) and “reductions in [myocardial infarction]” (15% sulfonylurea/insulin, 33% metformin) and “in all-cause mortality (13% and 27%, respectively)” (3) from the UK Prospective Diabetes Study (UKPDS) 10-year follow-up. In addition, there were two statements based on meta-analyses that “every HbA1c reduction of ∼1% may be associated with a 15% relative risk reduction in nonfatal myocardial infarction, but without benefits on stroke or all-cause mortality” (1) and a 9% “reduction in major CVD outcomes” (3).
There were two statements that intensive therapy produced an increase in mortality “22% increase” with a “hazard ratio of 1.22” and in absolute terms “1.41% vs. 1.14% per year” (3). No other quantitative glucose-lowering information was proffered. While there is reference to “the impact of glucose control on cardiovascular complications” (2) remaining uncertain, the two main figures in these articles do not mention cardiovascular, ophthalmic, or renal end points.
This relative lack of information is an unfortunate oversight given the considerable debate surrounding the impact of treatments on clinical end points (7).
Potential Harms From Treatments
Approximately 50 disadvantages/harms are usefully listed in tables for 12 classes of medications (1,2). However, nowhere in the tables, and only twice in the documents, are absolute numbers provided (sodium–glucose cotransporter 2 inhibitors/mycotic infections and dipeptidyl peptidase-4/heart failure).
Conclusions
The information presented in these documents (1–3) is glucose-centric and not organized or presented in a way that could be construed as supporting shared decision making. We believe position statements and standards that call for shared decision making should include information necessary to achieve this laudable goal.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
