Post-authorisation Safety Study in Patients With Type 2 Diabetes to Assess the Risk of Liver Injury, Kidney Injury, Urinary Tract and Genital Infections, and Diabetic Ketoacidosis in Patients Treated With Empagliflozin, Compared to DPP-4 Inhibitors

Incidence rates (IRs) of acute liver injury (ALI) in patients with no predisposing conditions (ALI1) in propensity score-trimmed study cohorts for ALI1 are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% confidence interval (CIs) were generated using a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of acute kidney injury (AKI) in propensity score-trimmed cohort for AKI are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of diabetic ketoacidosis (DKA) in propensity score-trimmed cohort for DKA are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of severe complications of urinary tract infections (UTIs) in propensity score-trimmed cohort for UTI among CPRD participants are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of genital infections in males (GIM) in propensity score-trimmed cohorts for GIM among CPRD participants are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of genital infections in females (GIF) in propensity score-trimmed cohort for GIF among CPRD participants are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of acute liver injury (ALI) in patients with or without predisposing conditions (ALI2) in propensity score-trimmed cohort for ALI2 are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of chronic kidney disease (CKD) in propensity score-trimmed cohort for CKD among CPRD participants are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of severe genital infections in males (GIMH) in propensity score-trimmed cohort for GIM among CPRD participants are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.

Incidence rates (IRs) of severe genital infections in females (GIFH) in propensity score-trimmed cohort for GIF among CPRD participants are reported.

Unadjusted IRs were computed as the number of events divided by the total person-years at risk in the empagliflozin and DPP-4 inhibitors cohorts overall and for subgroups with and without insulin use at the index date. IRs adjusted for propensity score decile were generated for empagliflozin and DPP-4 inhibitor cohorts overall (not stratified by any other variable). These estimates with corresponding 95% CIs were generated through the application of a Poisson regression model where the outcome was modelled as a function of treatment cohort (empagliflozin or DPP-4 inhibitors) and propensity score decile (specified as a categorical variable) with the log of time of exposure (in years) as the offset. If any variable remained unbalanced after trimming, it was added as an independent variable in the Poisson regression model.