Figure 1.

Change from baseline in HbA_1c at week 26 by subgroup in (a) PIONEER 9 and (b) PIONEER 10. Baseline values are for all treatment arms combined for each subgroup in each trial. For all analyses, missing values were imputed by a pattern mixture model using multiple imputation. The pattern was defined by randomized treatment arm and treatment status (premature trial product discontinuation or initiation of rescue medication, or both), and imputation was carried out within groups defined by trial product and treatment status. For the subgroup analyses, the estimated changes from baseline were analyzed using an ancova model with treatment, strata, subgroup, and interaction between treatment and subgroup as categorical fixed effects, and baseline HbA_1c as a covariate. The statistical analyses were not controlled for multiplicity. α‐GI, alpha‐glucosidase inhibitor; ancova, analysis of covariance; BMI, body mass index; HbA_1c, glycated hemoglobin; N, number of patients contributing to the analysis; SGLT2i, sodium‐glucose cotransporter‐2 inhibitor; SU, sulfonylurea; TZD, thiazolidinedione.