Comment on Shahraz et al. Do Patient Characteristics Impact Decisions by Clinicians on Hemoglobin A_1c Targets? Diabetes Care 2016;38:e145–e146

We have read with great interest the analysis by Shahraz et al. (1) regarding the impact of patient characteristics on HbA_1c targets.

In recent years, it has been recognized that ameliorating hyperglycemia does not lead to reduced mortality and improved cardiovascular outcomes in the short term. Therefore, the means whereby glucose is controlled and the “cost” of control (i.e., hypoglycemia and medication side effects) must be considered (2). Accordingly, the importance of individualizing glycemic targets and setting safe targets in frail populations has been emphasized in diabetes position statements in recent years (3).

An interesting observation in the analysis by Shahraz et al. (1) is an exceptionally low rate of patients (4%) reporting that their physicians thought they should strive for a glycemic target of >7%. It has been estimated that approximately half of the U.S. adults with diabetes should have a glycemic target set at ≥7% (4). Individualization of glycemic control, though unanimously supported, is open to subjective interpretation that may result in discrepant goals proposed by different physicians with respect to the same patient. Additionally, setting a glycemic target while considering multiple clinical aspects of the individual patient is demanding and time consuming. The primary care physician may lack the time or experience to set and discuss the desirable goal with the patient. Furthermore, with the lack of a uniform target, quality control has become challenging; physicians may still aim to improve their patients’ glycemic status so as not to have many “poorly controlled” patients under their care, who in reality have an HbA_1c that more properly matches their frail state.

We have recently proposed an algorithm, which has been developed into a mobile phone app, designed to assist the physician in determining a glycemic target for the individual patient based on clinical characteristics. The algorithm is intended to be a decision support tool and is based on the opinion of more than 150 expert worldwide leading diabetologists and validated by more than 50 additional physicians. The use of the algorithm may promote further thought and consideration to the different decision elements used to determine the most suitable target (5).

The algorithm proposed may be used by both the physician and the patient, highlighting the importance of patient involvement in the clinical decision-making process. Shahraz et al. (1) noted that only 54% of the 1,782 responders to the question “What does [your doctor] say [your] ‘A1C’ level should be?” reported a target, whereas the others responded they did not know or that no target was set. Interestingly, the rate of patients unaware of their glycemic target consistently declined from 2005 to 2013 from 30% to 10%. This emphasizes a welcome trend of increased patient empowerment in recent years.

The shift of glycemic targets from a uniform target to individualized targets is an ongoing process that will take time to fully appreciate and implement. The use of decision support programs, such as the one we proposed, may promote this transition and improve the care of our patients.