Skip to main content

Table 3 Results of cost-effectiveness analysis of GLP-1RAs versus insulin (base-case and sensitivity analyses)

From: Cost-effectiveness of GLP-1 receptor agonists versus insulin for the treatment of type 2 diabetes: a real-world study and systematic review

 

ΔEffectivenessa

ΔCostb

Cost per case of event prevented (ICER)

NNTc

Third-party payer perspective

Healthcare sector perspective

Third-party payer perspective

Healthcare sector perspective

Base-case analysis (ITT scenario for effectiveness estimation + cost measured from the index date until the end of observation)

 All-cause death

56.63

969

− 342

54,851

− 19,391

 Hospitalized hypoglycemia

30.06

969

− 342

29,115

− 10,293

1st sensitivity analysis (ITT scenario for effectiveness estimation + cost measured from the index date until the occurrence of study event)

 All-cause death

56.63

971

− 325

55,002

− 18,425

 Hospitalized hypoglycemia

30.06

1128

47

33,901

1399

2nd sensitivity analysis (AT scenario for effectiveness estimation + cost measured from the index date until the end of observation)

 All-cause death

100.80

969

− 342

97,633

− 34,515

 Hospitalized hypoglycemia

29.79

969

− 342

28,856

− 10,201

3rd sensitivity analysis (AT scenario for effectiveness estimation + cost measured from the index date until the occurrence of study event)

 All-cause death

100.80

971

− 325

97,901

− 32,796

 Hospitalized hypoglycemia

29.79

1128

47

33,600

1387

  1. ΔCost: difference in costs per subject between GLP-1RA and insulin users over a mean follow-up period; GLP-1RAs: glucagon-like peptide-1 receptor agonists; ICER: incremental cost-effectiveness ratio; ITT: intention-to-treat; AT: as-treated
  2. aIn the primary and 1st sensitivity analyses, the follow-up period of effectiveness estimation was measured from the first prescription of the stable use of a GLP-1RA or insulin (the index date) to withdrawal from Taiwan’s National Health Insurance program, event occurred, death, or the end of 2016, whichever came first (i.e., ITT analyses). In the 2nd and 3rd sensitivity analyses, the follow-up period of effectiveness estimation was measured from the first prescription of the stable use of a GLP-1RA or insulin (the index date) to the occurrence of a study event of interest (i.e., all-cause death or hospitalized hypoglycemia), withdrawal from Taiwan’s National Health Insurance program, death, discontinuation of the current treatment (i.e., GLP-1RA or insulin), or the end of 2016, whichever cam first (i.e., AT analyses)
  3. bIn the primary and 2nd sensitivity analyses, costs were measured for the follow-up period from the first prescription of the stable use of a GLP-1RA or insulin (the index date) to withdrawal from Taiwan’s National Health Insurance program, death, or the end of 2016, whichever came first. In the 1st and 3rd sensitivity analyses, costs were measured for the follow-up period from the first prescription of the stable use of a GLP-1RA or insulin (the index date) to the occurrence of a study event of interest (i.e., all-cause death or hospitalized hypoglycemia), withdrawal from Taiwan’s National Health Insurance program, death, or the end of 2016, whichever came first
  4. cThe NNT estimate was calculated based on the survival probabilities from the Cox proportional hazard model with the adjustment of the imbalanced baseline patient characteristics shown in Additional file 1: Table S2
  5. All costs are presented in 2019 US dollars and rounded to the nearest integer