Projecting the incidence and costs of major cardiovascular and kidney complications of type 2 diabetes with widespread SGLT2i and GLP-1 RA use: a cost-effectiveness analysis

Figures

• 

  Fig. 1. Schematic of the model. The model begins on 1 January 2020 with the entire population with type 2 diabetes in Australia. Each cycle (1 year), people with type 2 diabetes are at risk for MI, stroke, HF, ESKD and death, with the number experiencing each event in the cycle tracked. If an individual experiences an event they have not had before, they then transition to the relevant health state (straight black arrows). If an individual does not experience an event, or experiences only the event of the type they have previously had, they remain in the same health state for the next cycle (circular arrows). There are 17 possible health states, represented as either absence of all of MI, stroke, HF and ESKD (i.e. people with type 2 diabetes not contained within the Venn diagram), any combination of these four outcomes (the 15 spaces within the Venn diagram), or death. Additionally, at each cycle, a population with incident type 2 diabetes is added, who enter the model in their respective health states

• 

  Fig. 2. Results from scenario analyses in common cost-effectiveness planes. Incremental costs and benefits are as compared with the current use under each scenario. Not shown are results from the US medication price scenario, as the costs were too large to fit in a common plane (all results are shown in ESM Table 8). The pink line represents the AU$28,000 per QALY willingness-to-pay threshold. All costs are in 2020 Australian dollars. All health economic outcomes have been subject to 5% annual discounting unless otherwise indicated. (a) SGLT2i/total population, (b) GLP-1 RA/total population, (c) SGLT2i/secondary prevention population, (d) GLP-1 RA/secondary prevention population

• 

  Fig. 3. Tornado diagrams showing results from one-way sensitivity analyses. Tornado diagrams display the uncertainty in the model primary outcome (ICER) associated with variation in the input variables. Lower-limit ICERs are the ICERs under the lower limit of uncertainty associated with the variable, and upper-limit ICERs are the ICERs under the upper limit of uncertainty. For example, the effect of GLP-1 RAs on all-cause mortality is estimated using the HR of 0.88 (95% CI 0.82, 0.94); thus, the lower limit for this variable is 0.82 and the upper limit is 0.94, and the ICERs displayed in the tornado diagram are the ICERs when the model is run using these values, with the middle value representing the ICER using the point estimate of 0.88. All health economic outcomes have been subject to 5% annual discounting. (a) SGLT2i/total population, (b) GLP-1 RA/total population, (c) SGLT2i/secondary prevention population, (d) GLP-1/RA secondary prevention population

• 

  Fig. 4. Results from probabilistic sensitivity analyses presented in a cost-effectiveness plane. Incremental costs and benefits are as compared with the current use for each scenario. The pink line represents the $AU28,000 per QALY willingness-to-pay threshold. All costs are in 2020 Australian dollars. All health economic outcomes have been subject to 5% annual discounting. (a) SGLT2i/total population, (b) GLP-1 RA/total population, (c) SGLT2i/secondary prevention population, (d) GLP-1 RA/secondary prevention population