The analysis took a cost of control modeling approach, in line with a previous US analysis of once-weekly semaglutide versus dulaglutide, and previous analyses of liraglutide (a once-daily GLP-1 analog) versus sitagliptin (a dipeptidyl peptidase-4 inhibitor) [12, 16]. Based on three common endpoints used in the treatment of T2D, this approach allows the short-term cost-effectiveness of interventions to be evaluated in a clinically relevant manner that is both straightforward and transparent .
Clinical data for the analysis were taken from the SUSTAIN trial program, specifically the SUSTAIN 3 and SUSTAIN 7 trials, comparing once-weekly semaglutide 1.0 mg with exenatide ER and once-weekly semaglutide 0.5 mg and 1.0 mg with dulaglutide 0.75 mg and 1.5 mg, respectively (Table 1) [15, 16]. No head-to-head trial has compared once-weekly semaglutide 0.5 mg with exenatide ER, and clinical data were therefore unavailable for this comparison. The present cost of control analysis was based on the proportion of patients achieving target endpoints with each intervention. The statistical significance of differences between treatments was assessed using a logistic regression model. In line with the US FDA label for once-weekly semaglutide, all outcomes were based on the intent-to-treat population in which missing data were imputed using multiple imputation based on retrieved dropout analysis .
A model was developed in Microsoft Excel (Microsoft Corporation, Redmond, WA) to assess the relative costs of bringing a single patient to each of three pre-specified single or composite endpoints in the two SUSTAIN trials. The endpoints covered glycemic targets, body weight, and hypoglycemia outcomes: (1) HbA1c < 7.0%, (2) HbA1c < 7.0% without hypoglycemia and without weight gain, (3) ≥ 1.0% HbA1c reduction and ≥ 5.0% weight loss (Table 1).
Cost Data, Resource Use, Time Horizon, and Perspective of the Analysis
Only wholesale acquisition cost (WAC) prices were included in the analysis, and these were sourced from the Price-Rx database (Table 2) . Needle costs were not captured, as needles are included in the once-weekly GLP-1 RA packs, while no costs relating to self-monitoring of blood glucose (SMBG) testing were applied, as testing was assumed to be the same regardless of the treatment regimen. Adherence to all modeled regimens was assumed to be 100%.
All medications were dosed at the weekly doses recommended in the product labels: 2.0 mg per week for exenatide ER, 0.75 mg and 1.5 mg for dulaglutide 0.75 mg and 1.5 mg respectively, and 0.5 mg and 1.0 mg for once-weekly semaglutide 0.5 mg and 1.0 mg, respectively.
The analysis took the perspective of a US private healthcare payer, with outcomes projected over a 1-year time horizon.
Relative Cost of Control Calculations
The relative efficacy for each comparator was calculated by dividing the proportions of patients achieving each of the three endpoints with the comparator by the proportions of patients achieving the corresponding endpoint with once-weekly semaglutide from the relevant SUSTAIN trial. The relative cost of each comparator was calculated by dividing the annual cost with the comparator by the annual cost of once-weekly semaglutide. The relative efficacy and relative cost of each comparator were then plotted on a cost–efficacy plane as the abscissa and ordinate, respectively (Figs. 1, 2, 3, 4, 5, 6). The reference of once-weekly semaglutide 0.5 mg or 1.0 mg thereby formed the identity line or line of equality (i.e., x = y), with comparators falling above the line having a worse efficacy-to-cost ratio (incurring higher costs for the same efficacy or lower efficacy for the same cost), and comparators falling below the line having a better efficacy-to-cost ratio (incurring lower costs for the same efficacy or higher efficacy for the same cost), based on an assumption of a linear relationship between cost and efficacy. This is a novel methodology, allowing easy, visual interpretation of results that has, to date, only been published in abstract form in an analysis for the Canadian setting .
The model results were presented in terms of relative cost of control outcomes only (i.e., cost relative to once-weekly semaglutide 0.5 mg or 1 mg and efficacy relative to once-weekly semaglutide 0.5 mg or 1 mg). A constant ratio of treatment costs was assumed between the treatment arms over time to avoid differences occurring from the different trial durations, as efficacy outcomes from SUSTAIN 3 and SUSTAIN 7 were reported over 56 weeks and 40 weeks of follow-up, respectively (Table 1). While the use of cost and efficacy ratios over these time periods mitigated the impact of the different trial durations, the duration of follow-up in each trial should be considered when interpreting the results.
Compliance with Ethics Guidelines
The present modeling analysis is based on previously conducted clinical studies and does not reference any studies with human participants or animals conducted by the authors.