We assessed the efficacy of eight classes of diabetes medications used in current clinical practice [metformin, sulphonylureas, α-glucosidase inhibitors, thiazolidinediones, glinides, dipeptidyl peptidase-4 inhibitors, glucagon-like peptide-1 (GLP-1) analogues and insulin analogues] to reach the HbA1c target <7% in type 2 diabetes.Methods:
MEDLINE, EMBASE and the Cochrane CENTRAL were searched from inception through April 2011 for randomized controlled trials (RCTs) involving antidiabetic drugs. RCTs had to report the effect of any diabetes medication on the HbA1c levels, to include at least 30 subjects in every arm of the study, and to report the effect of therapy after a minimum of 12 weeks. Data were summarized across studies using random-effects meta-regression.Results:
A total of 218 RCTs (339 arms and 77 950 patients) met the inclusion criteria. The proportion of patients who achieved the HbA1c goal ranged from 25.9% (95% CI 18.5–34.9) with α-glucosidase inhibitors to 63.2% (54.1–71.5) with the long-acting GLP-1 analogue. There was a progressive decrease of the proportion of patients at target for each 0.5% increase in baseline HbA1c, ranging from 57.8% for HbA1c ≤7.5% to 20.8% for HbA1c ≥10% (p for trend <0.0001), with some difference between insulin and non-insulin drugs: for insulin, the proportion of patients at goal reached a plateau for basal HbA1c value >9.0% with no further decrease, whereas for non-insulin drugs the relationship was continuous without any evidence of plateau.Conclusions:
There is a considerable variability with regard to attainment of HbA1c goal of <7% among the different classes of diabetes medications; baseline HbA1c is an important determinant of observed efficacy.