Table 1

Patient Baseline and Treatment Characteristics and Change in HbA1c

RW (n = 221)RCTsa (n = 2,600)RW (n = 652)RCTsa (n = 1,889)
Baseline patient characteristics
 Age (years), mean (SE)57 (0.71)56 (0.20)63 (0.46)56 (0.22)
 Male sex, % (n)58 (127)53 (1,378)58 (378)53 (1,001)
 White race, % (n)82 (181)68 (1,768)81 (525)77 (1,455)
 HbA1c (%), mean (SE)8.34 (0.06)8.41 (0.02)8.15 (0.03)7.81 (0.02)
 HbA1c (mmol/mol), mean (SE)67 (0.7)68 (0.25)66 (0.3)62 (0.35)
 Any diabetes complications,b % (n)62 (136)15 (390)c62 (401)15 (283)c
 Use of advanced type 2 diabetes therapy before index,d % (n)83 (184)50 (1,300)71 (460)7 (132)
Treatment characteristics during follow-up
 Adherent to index drug, % (n)29 (64)e95 (2,470)e37 (242)95 (1,795)f
 Discontinued index drug, % (n)45 (100)g,h0 (0)40 (263)g,h0 (0)
  Switched to another class of diabetes drug, % of patients who discontinued (n)44 (44)42 (110)
 Did not discontinue index drug, % (n)55 (121)g,h100 (2,600)60 (389)g,h100 (1,889)
  Added new diabetes medication and continued index therapy (e.g., rescue therapy), % of patients who did not discontinue (n)21 (26)29 (114)
 Addition of other type 2 diabetes drug(s) after GLP-1 RA/DPP-4 initiation, % (n)32 (70)h0 (0)i34 (224)h0 (0)i
Change in HbA1c level ∼1 year after drug initiation
 Level difference (%), mean (95% CI) [range]−0.52 (−0.66, −0.38)−1.30 [−1.60, −0.84]j−0.51 (−0.59, −0.43)−0.68 [−0.90, −0.47]j
 Level difference (mmol/mol), mean (95% CI) [range]−6 (−16.3, −27.6)−14 [−18, −9]j−6 (−17.1, −28.2)−8 [−10, −5]j
  • aInformation from seven GLP-1 RA and four DPP-4 trials was extracted. One GLP-1 RA trial was a head-to-head comparison against a DPP-4 agent (19). Data for both treatment arms were extracted.

  • bPatients were considered to have diabetes complications if their DCSI score was greater than 1 (22) and included the following seven categories of complications: cardiovascular disease, nephropathy, retinopathy, peripheral vascular disease, stroke, neuropathy, and metabolic.

  • cPatients in GLP-1 RA and DPP-4 trials were assumed to have a similar percentage of patients with any diabetes complications as to the percentage reported in one of the selected GLP-1 RA trials (25).

  • dUse of advanced diabetes medications was defined as equal to 1 if the patient was treated with any monotherapy or combination of type 2 diabetes drugs beyond metformin monotherapy and defined as equal to 0 if the patient was treated with metformin monotherapy during the baseline year.

  • eFor patients with multiple treatment measurements due to multiple weight measurements eligible for the study, PDC was weighted by the number of observations per patient.

  • fMedication adherence was not widely reported in RCTs and was assumed to be 95% based on information reported in the included RCTs (see Supplementary Table 1).

  • gDiscontinuation was defined by the absence of the index drug on hand for at least 30 days and until the second HbA1c measurement date.

  • hFor patients with multiple treatment measurements due to multiple HbA1c measurements eligible for the study, discontinuation and treatment augmentation associated with the latest observation of each patient were reported.

  • iBecause data after patients receive additional diabetes drugs (e.g., rescue therapy) are excluded from analyses of trial data, the addition of type 2 diabetes drugs post index in trial settings was assumed to be zero for the purpose of predicting change in HbA1c under typical trial conditions.

  • jOwing to the lack of individual patient-level data of RCTs, ranges of change in HbA1c were reported for GLP-1 RA and DPP-4 studies, respectively.