Figure 1: Adjusted mean (95% CI) change from baseline in HbA1c at week 24. Number of patients by treatment group and HbA1c outcomes. Data collected before rescue treatment. Change from baseline in HbA1c was evaluated using an ANCOVA model with last observation carried forward to account for missing data. At week 24, saxagliptin 2.5 mg and saxagliptin 5 mg significantly decreased HbA1c from baseline compared with placebo (P<0.0001 for each saxagliptin dose). ANCOVA=analysis of covariance; HbA1c=glycated hemoglobin. *Number of randomized patients with baseline and week 24 values.