Table 4.

Results from the main analysis: 4-arm model with 56 patients in the basal–bolus arm

(a) Estimated rates of confirmed hypoglycemia per treatment arm, based on a negative binomial model
	IDegLira (N = 199)		GLP-1RA add-on to basal insulin (N = 225)		Basal–bolus with insulin glargine as basal component (N = 56)		Basal-only (up-titrated insulin glargine) (N = 329)
	Events/100 PYE	95% CI	Events/100 PYE	95% CI	Events/100 PYE	95% CI	Events/100 PYE	95% CI
Overall confirmed hypoglycemia	122.8	(90.71; 166.14)	124.4	(89.49; 172.84)	1060.8**	(680.17; 1654.37)	286.1**	(231.10; 354.08)
Severe hypoglycemia	0.4	(0.04; 4.68)	NA	NA	2.4	(0.38; 14.63)	2.6	(0.86; 7.94)
Non-severe hypoglycemia	121.9	(89.90; 165.23)	124.1	(89.16; 172.68)	1056.3**	(675.44; 1651.96)	282.8**	(228.06; 350.58)

(b) Estimated rate ratios for confirmed hypoglycemia, based on a negative binomial model
	IDegLira versus GLP-1RA add-on to basal insulin			IDegLira versus basal–bolus with insulin glargine as basal component			IDegLira versus basal-only (up-titrated insulin glargine)
	Rate ratio	95% CI	P value	Rate ratio	95% CI	P value	Rate ratio	95% CI	P value
Overall confirmed hypoglycemia	0.99	(0.63; 1.54)	0.95	0.12	(0.07; 0.20)	<0.0001	0.43	(0.30; 0.62)	<0.0001
Severe hypoglycemia	NA	NA	NA	0.18	(0.01; 2.77)	0.22	0.16	(0.01; 1.81)	0.14
Non-severe hypoglycemia	0.98	(0.63; 1.54)	0.94	0.12	(0.07; 0.20)	<0.0001	0.43	(0.30; 0.63)	<0.0001

IDegLira significantly different: ** P < 0.01

NA Not applicable: could not be estimated as no severe hypoglycemic events were observed with GLP-1RA add-on to basal insulin

Confirmed hypoglycemia was defined as the occurrence of severe episodes (i.e., requiring assistance), or episodes in which plasma glucose concentration (confirmed by self-monitored blood glucose) was less than 56 mg/dL (3.1 mmol/L), irrespective of symptoms

CI confidence interval, GLP-1RA glucagon-like peptide-1 receptor agonist, IDegLira insulin degludec/liraglutide, PYE patient-year of exposure