Evaluation of the long-term cost-effectiveness of IDegLira versus liraglutide added to basal insulin for patients with type 2 diabetes failing to achieve glycemic control on basal insulin in the USA

Abstract

Background and aims: IDegLira, a fixed ratio combination of insulin degludec and glucagon-like peptide-1 receptor agonist liraglutide, utilizes the complementary mechanisms of action of these two agents to improve glycemic control with low risk of hypoglycemia and avoidance of weight gain. The aim of the present analysis was to assess the long-term cost-effectiveness of IDegLira vs liraglutide added to basal insulin, for patients with type 2 diabetes not achieving glycemic control on basal insulin in the US setting.

Methods: Projections of lifetime costs and clinical outcomes were made using the IMS CORE Diabetes Model. Treatment effect data for patients receiving IDegLira and liraglutide added to basal insulin were modeled based on the outcomes of a published indirect comparison, as no head-to-head clinical trial data is currently available. Costs were accounted in 2015 US dollars ($) from a healthcare payer perspective.

Results: IDegLira was associated with small improvements in quality-adjusted life expectancy compared with liraglutide added to basal insulin (8.94 vs 8.91 discounted quality-adjusted life years [QALYs]). The key driver of improved clinical outcomes was the greater reduction in glycated hemoglobin associated with IDegLira. IDegLira was associated with mean costs savings of $17,687 over patient lifetimes vs liraglutide added to basal insulin, resulting from lower treatment costs and cost savings as a result of complications avoided.

Conclusions: The present long-term modeling analysis found that IDegLira was dominant vs liraglutide added to basal insulin for patients with type 2 diabetes failing to achieve glycemic control on basal insulin in the US, improving clinical outcomes and reducing direct costs.

Keywords:

• Cost
• cost-effectiveness
• IDegLira
• liraglutide
• type 2 diabetes mellitus
• USA

Introduction

In 2012 it was estimated that 29.1 million people in the US were living with diabetes mellitus (including diagnosed and undiagnosed patients), a prevalence of 9.3%¹. This is projected to increase to over 35 million by 2040². The disease represents the seventh leading cause of death in the US, listed as the underlying cause of death on 69,071 death certificates, and mentioned as a cause of death in a total of 234,051 death certificates in 2010¹. However, studies suggest that this may be an under-estimate, as only 35–40% of patients with diagnosed diabetes who have died have the condition mentioned on their death certificate, and patients with diabetes may not be diagnosed at all¹. Diabetes is also associated with a large morbidity burden, with increased risk of cardiovascular disease, renal disease, blindness, and amputation compared to the general population. Data from long-term studies in the US has shown that the event rates for diabetes-related complications have fallen from 1990 to 2010 (myocardial infarction by 67.8%, stroke by 52.7%, amputation by 51.4%, and end-stage renal disease by 28.3%)³. Over this 20-year time period, better management of risk factors, such as serum lipid levels (using statins), blood pressure (using renin-angiotensin-system-acting agents), and smoking status are likely to have resulted in reduced incidence of cardiovascular disease, end-stage renal disease, and amputation. Improvements in the management of patient blood sugar levels may have led to reductions in the incidence of end-stage renal disease, amputation, and, to a lesser extent, cardiovascular disease.

Lifestyle intervention and oral anti-diabetic therapies are recommended as initial treatments following diagnosis of type 2 diabetes, but many patients require additional treatment (such as glucagon-like peptide-1 [GLP-1] receptor agonists) and later initiate insulin therapy due to the progressive nature of the disease⁴. Basal insulin alone is often used as the first injectable treatment regimen, and doses are titrated with the aim of bringing patients to a recommended glycated hemoglobin target of 7%⁵. However, data suggest that 67% of patients with type 2 diabetes receiving basal insulin in the US are not achieving this treatment target⁶. Patients failing to achieve glycemic targets on basal insulin are likely to receive additional interventions as part of treatment intensification. Current injectable options for treatment intensification include addition of up to three times daily prandial insulin (such as insulin aspart) or addition of a GLP-1 receptor agonist (such as liraglutide) to basal insulin⁴. IDegLira represents an alternative treatment option for patients with type 2 diabetes not achieving treatment targets on basal insulin.

IDegLira is a fixed ratio combination of insulin degludec and GLP-1 receptor agonist liraglutide in a single injection device. The fixed-ratio combination was developed to take advantage of the combined effects of a basal insulin and a GLP-1 receptor agonist on glycemic control through their complementary mechanisms of action. The combination of a basal insulin and a GLP-1 receptor agonist may reduce the increased risk of hypoglycemia and weight gain often associated with insulin therapy, and slower titration of the GLP-1 receptor agonist as part of the fixed ratio combination may reduce nausea rates^7,8. The safety and efficacy of IDegLira has been assessed in the DUAL trial program^9–13.

The DUAL VII trial comparing IDegLira with basal-bolus insulin therapy has completed and will be reported in mid-2017. However, to date, no head-to-head clinical trials comparing IDegLira with liraglutide added to basal insulin have been conducted. The aim of the present analysis was to assess the long-term cost-effectiveness of IDegLira vs liraglutide added to basal insulin, for patients with type 2 diabetes not achieving glycemic control on basal insulin in the US setting, based on a published indirect comparison.

Methods

Model description

The analysis was performed using the IMS CORE Diabetes Model (IMS Health, Basel, Switzerland), the architecture, assumptions, features, and capabilities of which have been previously published¹⁴. The model is a validated, non-product specific diabetes policy analysis tool, and is based on a series of inter-dependent sub-models that simulate the complications of diabetes (angina, myocardial infarction, congestive heart failure, stroke, peripheral vascular disease, diabetic retinopathy, macular edema, cataract, hypoglycemia, ketoacidosis, lactic acidosis, nephropathy including end-stage renal disease, neuropathy, foot ulcer and amputation, and non-specific mortality). Each sub-model has a semi-Markov structure and uses time, state, time-in-state, and diabetes type-dependent probabilities derived from published sources. Monte Carlo simulation using tracker variables overcomes the memory-less properties of the standard Markov model, and allows interconnectivity and interaction between individual complication sub-models. Long-term outcomes projected by the model have been validated against real life data in 2004 and, more recently, in 2014^15,16.

In the present analysis, a simulated cohort of 1,000 patients was run through the model 1,000 times for each simulation (base case and sensitivity analyses). Mean values and standard deviations were generated for long-term outcomes (life expectancy, quality-adjusted life expectancy, cumulative incidence of diabetes-related complications, time to onset of diabetes-related complications, direct medical costs). The time horizon was set to patient lifetimes in the base case (50 years) to capture all relevant long-term complications and associated costs, thereby assessing their impact on life expectancy and quality-adjusted life expectancy. This approach is in line with guidance on the assessment of the cost-effectiveness of diabetes interventions¹⁷. Future costs and clinical benefits were discounted symmetrically by 3% per annum, in line with published health economic guidance for the US¹⁸.

Clinical inputs

To date, no head-to-head clinical trial comparing IDegLira with liraglutide added to basal insulin has been conducted. In light of this, an indirect comparison was conducted, with patient-level statistically modeled to account for differences in baseline characteristics across the trial populations¹⁹. The indirect comparison was performed using a robust methodology (utilizing patient-level data and, therefore, providing a high degree of reliability) aligned with the European Network for Health Technology Assessment guidelines on how to conduct indirect analyses, and has been used previously for indirect comparisons of trials of interventions for diabetes^20,21. Treatment effects applied in the first year of the analysis are shown in Table 1.

Table 1. Treatment effects applied in the first year of the analysis.

	IdegLira (mean (SD))	Liraglutide added to basalinsulin (mean (SD))
HbA1c (%)	−1.68 (0.94)	−1.33 (0.94)^a
Systolic blood pressure (mmHg)	−6.84 (12.95)	−4.68 (12.95)
Total cholesterol (mg/dL)	−10.44 (29.67)	−13.26 (29.67)
HDL cholesterol (mg/dL)	+0.47 (6.74)	−0.74 (6.74)
LDL cholesterol (mg/dL)	−7.56 (24.07)	−9.86 (24.07)
Triglycerides (mg/dL)	−18.61 (78.66)	−16.56 (78.66)
Body mass index (kg/m^2)	−1.02 (1.21)	−1.27 (1.21)
Severe hypoglycemic event rate (events per 100 patient years)	0.42	0.00
Non-severe hypoglycemic event rate (events per 100 patient years)	121.88	124.08

HbA1c: glycated hemoglobin; HDL: high density lipoprotein; LDL: low density lipoprotein.

^a Statistically significant difference at 95% confidence level.

Data to populate the IDegLira arm of the pooled analysis conducted by Freemantle et al.¹⁹ were taken from the DUAL II study (NCT01392573)²². DUAL II was a 26-week, randomized, double-blinded trial comparing IDegLira with insulin degludec (capped at 50 international units [IU]) in people with type 2 diabetes uncontrolled on basal insulin in combination with metformin ± sulfonylurea/glinides. Sulfonylureas and/or glinides were discontinued at randomization. Data to inform the pooled analysis and, therefore, present health economic analysis was taken from all patients in the IDegLira arm of the study.

The liraglutide added to basal insulin arm was based on the LIRA-ADD2BASAL study (NCT01617434)²³. In this 26-week, double-blinded, randomized trial, patients previously receiving basal insulin were randomly allocated to receive either liraglutide 1.8 mg or placebo. In this study, 66.7% of patients were receiving insulin glargine, and 33.3% of patients were receiving insulin detemir, and the present health economic analysis takes this into account.

The baseline cohort was based on the characteristics of patients receiving IDegLira in the DUAL II study²². Mean age, duration of diabetes, glycated hemoglobin (HbA1c), and body mass index (BMI) were 56.8 years, 10.3 years, 8.7%, and 33.6 kg/m², respectively.

Treatment intensification and long-term parameter progression

Patients receiving IDegLira or liraglutide added to basal insulin were assumed to receive that treatment for 5 years before intensifying treatment to basal-bolus insulin therapy. This assumption recognizes that intensification to basal-bolus insulin therapy will likely be required to maintain glycemic control over the long-term due to the progressive nature of the disease (with continued decline of beta-cell function and increased insulin resistance over time). Following application of the treatment effects based on the data from the indirect comparison in the first year of the analysis, systolic blood pressure, and serum lipids were assumed to follow the natural progression algorithms built into the IMS CORE Diabetes Model, based on the UK Prospective Diabetes Study (UKPDS) or Framingham data (as described by Palmer et al.¹⁴). Benefits in terms of HbA1c were assumed to persist for the 5 years patients received initial treatments and were abolished on treatment switching. This approach was chosen as there is growing evidence from long-term studies that, in patients with type 2 diabetes mellitus, particularly those receiving some form of insulin therapy, HbA1c remains stable and does not increase over time^24–28. Studies suggest that HbA1c can remain stable for 5 years in both conventional and intensive treatment arms, and therefore assumptions around HbA1c used in the present analysis are likely to reflect clinical practice. These assumptions were varied in sensitivity analyses.

Costs and utilities

Diabetes medication resource use was based on the trial data from which the treatment effects were taken for the two interventions, and was adjusted in the indirect comparison¹⁹. In the IDegLira arm patients were assumed to use one needle per day, with patients in the liraglutide added to basal insulin arm using two needles per day. Costs of self-monitoring of blood glucose (SMBG) testing were based on an analysis of insurance claims in the US²⁹. Annual costs were calculated by multiplying the daily cost by 365.25 (Table 2).

Table 2. Annual treatment costs.

	IDegLira ($)	Liraglutide added tobasal insulin ($)
Daily IDegLira cost	24.02	–
Daily basal insulin cost	–	9.60
Daily liraglutide cost	–	24.92
Daily needle costs	0.46	0.88
Daily SMBG cost	1.63	1.63
Total daily cost	26.11	37.04
Total annual cost	9,537.52	13,527.53

$: 2015 US dollars; SMBG: self-monitoring of blood glucose.

The modeling analysis captured the cost of treating diabetes-related complications in the year of the event and in subsequent years. Costs of treating diabetes-related complications were based on a literature review^30–34. Costs were inflated to 2015 values if necessary. As diabetes progresses, patients develop complications that influence their overall health-related quality-of-life. Utilities associated with each of the modeled complications were taken from published sources^14,35–38.

Sensitivity analyses

The extrapolation of clinical results by modeling the long-term consequences is associated with uncertainty. Sensitivity analyses were performed on key parameters in the model to assess the robustness of the base case findings. The influence of time horizon on the outcomes projected by the model was investigated by running analyses over 10 and 20 years. It should be noted that a time horizon of 50 years was required for all modeled patients to have died, and therefore shorter time horizons did not capture all potential late-stage complications and costs. To examine the effect of discounting on cost-effectiveness outcomes, simulations were performed with (symmetric) discount rates of 0% and 6% annually. Five simulations were run to assess the key drivers of clinical benefits associated with IDegLira. In the IDegLira arm, changes in HbA1c, systolic blood pressure, serum lipids, BMI, and hypoglycemia were set to the value in the liraglutide added to basal insulin arm in turn. This allowed the contribution of individual clinical effects to long-term health economic outcomes to be assessed. A further analysis was prepared with only statistically significant differences between the treatment arms applied (Table 1).

Two alternative approaches to HbA1c progression were explored. In the first, no HbA1c changes were applied following the treatment effects applied in the first year of the analysis. This attempts to capture the legacy effect, where an early improvement in HbA1c has a benefit in the later years of life, even if the HbA1c difference no longer persists. In the second, the UKPDS HbA1c progression equation was applied in both arms of the simulation. HbA1c increases over time in both arms of the analysis, with the HbA1c benefit in the IDegLira arm gradually reduced. Analyses were run with the upper and lower 95% confidence interval of the HbA1c change seen in the IDegLira arm of the pooled analysis applied, with all other parameters in the IDegLira and liraglutide added to basal insulin arm remaining unchanged. The base case analyses assumed that the BMI difference between the treatment arms was abolished on treatment switching, and an alternative to this was explored in a sensitivity analysis, with the difference maintained for the duration of the analysis.

To investigate the effect of the timing of treatment switching on cost-effectiveness, simulations were performed with the year of treatment switch to basal-bolus therapy brought forward to the end of year 3 in the IDegLira and liraglutide added to basal insulin arms, pushed back to the end of year 7 in the IDegLira and liraglutide added to basal insulin arms, and no treatment switching.

The effect of over- or under-estimating the direct cost of treating diabetes-related complications was investigated in two scenarios. In the first, the cost of treating complications was increased by 10%, and in the second the cost was reduced by 10%. The impact of applying alternative disutilities for severe and non-severe hypoglycemic events was assessed by using the values published by Currie et al.³⁹ (–0.0118 per severe hypoglycemic event and –0.0035 per non-severe hypoglycemic event). Additionally, a scenario was investigated in which 28% of patients receiving liraglutide added to basal insulin were assumed to require twice daily insulin, incurring the cost of a further needle for subcutaneous injection, based on a 5-year parallel group study of insulin glargine vs neutral protamine Hagedorn (NPH) insulin⁴⁰. Insulin doses in the comparator arm were conservatively assumed to remain unchanged in this analysis, To further investigate the impact of consumables on cost-effectiveness outcomes, scenarios were evaluated with costs of needles and SMBG testing excluded. The effect of the cost of basal insulin was investigated in analyses with the cost of the comparator basal insulin replaced with the cost of NPH insulin.

In February 2014, an update to the IMS CORE Diabetes Model incorporating data from the UKPDS 82 was released, and an analysis using this version of the model has been conducted. While a validation study of the revised model has been published, the model proprietors suggest that the update is used in a sensitivity analysis, with the previous version being used in the base case¹⁶. Similarly, version 9.0 of the IMS CORE Diabetes Model was released in summer 2015, and this was used in a sensitivity analysis. To date, no validation studies or user guide for the updated version of the model have been released, and, therefore, it was not considered appropriate to use version 9.0 of the model for the base case analysis.

Probabilistic sensitivity analysis (PSA) was performed using the pre-defined function in the IMS CORE Diabetes Model. Cohort characteristics, treatment effects, and complication costs and utilities were sampled from distributions, and the simulation was run using a second order Monte Carlo approach. Cohorts of 1,000 patients were run through the model 1,000 times for the PSA, as results were not subject to random statistical variation with these settings.

Results

Base case analysis

In the base case analysis, IDegLira was associated with improvements in discounted life expectancy of 0.02 years and discounted quality-adjusted life expectancy of 0.03 quality-adjusted life years (QALYs) vs liraglutide added to basal insulin (Table 3). These small but statistically significant improvements (95% confidence intervals for the differences in life expectancy and quality-adjusted life expectancy, 0.01–0.04 years and 0.02–0.05 QALYs, respectively) were due to minor reductions in the incidence of diabetes-related complications, including myocardial infarction, stroke, amputation, end-stage renal disease, and blindness. IDegLira was also associated with a small increase in the time to onset of diabetes-related complications, with mean time free of all complications increased by 0.2 years.

Table 3. Long-term cost-effectiveness outcomes.

	IDegLira (Mean (SD))	Liraglutide added tobasal insulin (Mean (SD))	Difference
Discounted life expectancy (years)	14.04 (0.20)	14.02 (0.21)	+0.02
Discounted quality-adjusted life expectancy (QALYs)	8.94 (0.13)	8.91 (0.14)	+0.03
Discounted direct costs ($)	206,385 (2,924)	223,072 (3,055)	–17,687
ICER (life expectancy)	IDegLira dominant
ICER (quality-adjusted life expectancy)	IDegLira dominant

$: 2015 US dollars; ICER: incremental cost-effectiveness ratio; QALYs: quality-adjusted life years; SD: standard deviation.

IDegLira was associated with mean cost savings of $17,687 per patient vs liraglutide added to basal insulin over patient lifetimes (Figure 1). The cost saving was driven predominantly by the lower acquisition cost of IDegLira compared with liraglutide added to basal insulin over the first 5 years of the analysis (annual cost savings of $3,990 per patient). Lower acquisition costs resulted from the lower dose of liraglutide received as part of IDegLira (37.80 dose steps of IDegLira contains 1.36 mg of liraglutide, vs 1.8 mg used in the LIEA-ADD2BASAL trial). Additional cost savings were made due to diabetes-related complications avoided, most notably ulcer and neuropathy complications (mean cost savings of $209 per patient) and cardiovascular complications (mean cost saving of $138 per patient). IDegLira was associated with a small improvement in clinical outcomes and reduced costs compared with liraglutide added to basal insulin, and, therefore, was considered dominant, with no calculation of an incremental cost-effectiveness ratio required (Table 3).

Figure 1. Mean direct costs over patient lifetimes. $: 2015 US dollars.

Sensitivity analyses

IDegLira remained dominant or associated with equivalent efficacy, but lower costs compared with liraglutide added to basal insulin in all sensitivity analyses prepared (Table 4). Shortening the time horizon of the analysis had a notable impact on the calculated outcomes, with the clinical benefit reduced at shorter time horizons. Surprisingly, cost savings were greater at 10 years than at 50 years, which is likely to be due to the survival paradox, where greater survival in the IDegLira arm results in greater costs in later years. Altering the discount rates reflected the long-term benefits associated with IDegLira, with clinical benefits and cost savings increasing when a discount rate of 0% was used, and decreasing when a discount rate of 6% was used.

Table 4. IDegLira vs liraglutide added to basal insulin: Sensitivity analyses.

	Discounted quality-adjusted life expectancy (QALYs)			Discounted costs ($)			ICER ($ per QALY gained)
	IDegLira	Liraglutide addedto basal insulin	Difference	IDegLira	Liraglutide addedto basal insulin	Difference
Base case	8.94	8.91	+0.03	206,385	224,072	−17,687	IDegLira dominant
20 year time horizon	7.76	7.71	+0.05	168,359	186,026	−17,666	IDegLira dominant
10 year time horizon	5.21	5.19	+0.02	99,781	117,567	−17,786	IDegLira dominant
0% discount rates	12.62	12.57	+0.06	309,213	328,422	−19,209	IDegLira dominant
6% discount rates	6.76	6.74	+0.02	148,088	164,402	−16,314	IDegLira dominant
HbA1c difference abolished	8.91	8.91	0.00	206,520	224,072	−17,552	IDegLira equally effective, but less costly
Blood pressure difference abolished	8.93	8.91	+0.02	206,055	224,072	−18,018	IDegLira dominant
Lipid difference abolished	8.94	8.91	+0.03	206,141	224,072	−17,932	IDegLira dominant
BMI difference abolished	8.95	8.91	+0.04	206,560	224,072	−17,512	IDegLira dominant
Hypoglycemia difference abolished	8.96	8.91	+0.05	206,330	224,072	−17,742	IDegLira dominant
Statistically significant differences only	8.96	8.91	+0.05	206,447	224,072	−17,625	IDegLira dominant
HbA1c benefit maintained	9.03	8.91	+0.12	206,468	224,072	−17,605	IDegLira dominant
UKPDS HbA1c creep	8.45	8.40	+0.05	205,189	222,613	−17,425	IDegLira dominant
Upper 95% CI of HbA1c change in IDegLira arm	8.97	8.91	+0.06	206,596	224,072	−17,476	IDegLira dominant
Lower 95% CI of HbA1c change in IDegLira arm	8.91	8.91	0.00	205,926	224,072	−18,146	IDegLira equally effective, but less costly
BMI benefit maintained after treatment switch	8.93	8.91	+0.02	206,415	224,072	−17,658	IDegLira dominant
Treatment switching at 3 years in both arms	8.86	8.82	+0.04	213,815	224,628	−10,813	IDegLira dominant
Treatment switching at 7 years in both arms	9.05	8.99	+0.06	199,974	223,415	−23,441	IDegLira dominant
No switching	9.52	9.37	+0.15	161,077	218,816	−57,739	IDegLira dominant
Costs of complications +10%	8.94	8.91	+0.03	208,451	226,172	−17,721	IDegLira dominant
Costs of complications −10%	8.94	8.91	+0.03	204,370	222,023	−17,653	IDegLira dominant
Currie et al.^39 hypo utilities applied	9.04	9.01	+0.04	206,385	224,072	−17,687	IDegLira dominant
28% of basal insulin patients twice daily	8.94	8.91	+0.03	206,802	224,678	−17,876	IDegLira dominant
Needle and SMBG costs not included	8.94	8.91	+0.03	185,111	202,158	−17,048	IDegLira dominant
NPH cost applied in liraglutide added to basal insulin arm	8.94	8.91	+0.03	180,724	191,109	−10,384	IDegLira dominant
UKPDS 82 equations applied	9.44	9.41	+0.04	221,237	239,114	−17,877	IDegLira dominant
Version 9.0 of the IMS CORE Diabetes Model	8.88	8.81	+0.07	209,962	227,071	−17,109	IDegLira dominant
PSA	8.62	8.56	+0.06	200,115	216,963	−16,848	IDegLira dominant

$: 2015 United States dollars; BMI: body mass index; CI: confidence interval; HbA1c: glycated hemoglobin; NPH: neutral protamine Hagedorn; PSA: probabilistic sensitivity analysis; QALY: quality-adjusted life year; SMBG: self-monitoring of blood glucose; UKPDS: United Kingdom Prospective Diabetes Study.

Abolishing each of the changes in physiological parameters associated with IDegLira identified that the key driver of improved clinical outcome was the greater reduction in HbA1c. Abolishing this difference between the treatment arms resulted in equivalent quality-adjusted life expectancy (although IDegLira remained less costly). Applying only the statistically significant differences did not change the conclusion of the analysis that IDegLira was dominant over liraglutide added to basal insulin. Applying an alternative HbA1c progression with no increases applied at any stage of the analysis in both arms (attempting to replicate the legacy effect) resulted in an increased clinical benefit and increased cost savings in the IDegLira arm, while application of the UKPDS HbA1c progression equation resulted in only small changes in the incremental quality-adjusted life expectancy benefit and cost savings, with IDegLira remaining dominant over basal insulin plus liraglutide.

When the upper 95% confidence interval of the HbA1c change was applied in the IDegLira arm, clinical benefits and cost-savings with IDegLira were found to increase. Conversely, when the lower 95% confidence interval was applied, IDegLira and basal insulin plus liraglutide were associated with equivalent efficacy, but IDegLira was associated with lower costs. Maintaining the BMI difference between the arms (i.e. a small benefit in the basal insulin plus liraglutide arm) resulted in a small reduction in the quality-adjusted life expectancy benefit associated with IDegLira, with only a small change in the difference in costs.

Changing the assumptions around treatment switching had a notable impact on the calculated health economic outcomes. Maintaining patients on IDegLira and basal insulin plus liraglutide for longer increased the incremental clinical benefit and cost saving associated with IDegLira. Assuming that patients remained on IDegLira or basal insulin plus liraglutide for patient lifetimes led to a quality-adjusted life expectancy benefit of 0.15 QALYs and a cost saving of $57,739. Use of alternative hypoglycemia disutilities resulted in increased quality-adjusted life expectancy in both arms of the analysis, but incremental differences showed only small changes due to the similar hypoglycemia rates in both arms. The conclusion that IDegLira was dominant over basal insulin plus liraglutide did not change.

Increasing the cost of complications resulted in increased cost savings with IDegLira, while reducing the cost of complications had the converse effect. Assuming that 28% of patients receiving basal insulin plus liraglutide required twice daily basal insulin resulted in increased cost savings with IDegLira, due to the increased needle costs in the basal insulin plus liraglutide arm. Conversely, removing the costs of needles and SMBG testing resulted in smaller cost savings with IDegLira. In the conservative scenario, with the cost of NPH insulin applied in the basal insulin plus liraglutide arm, IDegLira remained dominant, although cost savings were reduced.

Variation in the risk equations applied and the version of the IMS CORE Diabetes Model used resulted in increased quality-adjusted life expectancy and costs in both arms, but incremental differences showed only minor changes, with IDegLira remaining dominant over basal insulin plus liraglutide in both analyses. PSA suggested that, at a willingness-to-pay threshold of $100,000 per QALY gained, there was a 95.8% probability that IDegLira would be considered cost-effective vs liraglutide added to basal insulin.

Discussion

Based on clinical efficacy data from the DUAL II and LIRA-ADD2BASAL trials, the present study found that, in the US, IDegLira was associated with small improvements in clinical outcomes compared with liraglutide added to basal insulin. Sensitivity analyses identified that the key drivers of superior clinical outcomes with IDegLira was the greater improvement in HbA1c. IDegLira was associated with cost savings as a result of lower daily acquisition costs of treatment and reduced costs of treating diabetes-related complications. Therefore, IDegLira was considered dominant over liraglutide added to basal insulin.

IDegLira may have a number of advantages compared with alternative treatment intensification strategies for patients with type 2 diabetes not achieving glycemic control on basal insulin. It is administered in a single injection device and requires only a single daily injection, compared to up to four daily injections with a basal-bolus regimen and two daily injections with liraglutide added to basal insulin. This lower injection burden is likely to be attractive to patients with type 2 diabetes, as studies show a preference for reduced frequency of injection⁴¹. The single device may also be attractive to general practitioners treating patients with diabetes, as it allows two agents to be titrated simultaneously. This avoids the complexity of titration of basal and prandial insulin doses, aiming to manage overall glycemic control (as measured by HbA1c), post-prandial glycemic excursions, and hypoglycemic events (both nocturnal and diurnal).

To date, no head-to-head trials have been performed comparing IDegLira directly with liraglutide added to basal insulin. In the absence of such head-to-head data, a statistical indirect pooled analysis was performed using robust methodology in line with the European Network for Health Technology Assessment guidelines on how to conduct indirect analyses^19,20. A meta-analysis using equivalent methodology to assess the efficacy of diabetes interventions has been previously published⁴². That the comparison was not based on head-to-head studies could be considered a shortcoming of the present analysis. However, selection of the most appropriate comparator was the first priority in the analysis and the use of evidence synthesis, using robust methodologies, is becoming increasingly important and accepted for health technology assessment globally⁴³.

The present analysis used wholesale acquisition costs to calculate annual treatment costs associated with IDegLira and liraglutide added to basal insulin, and this may represent a limitation of the study. Costs of treatment may have been over-estimated, as potential contracting and rebate structures between the healthcare plans and manufacturers were not captured. Negotiations around these reductions in costs are almost always confidential and, as such, are difficult to capture in cost-effectiveness analyses. The present analysis conducted extensive sensitivity analyses around the pack prices and doses used to calculate annual treatment costs, with IDegLira remaining dominant over liraglutide added to basal insulin in all scenarios. Therefore, the conclusions of the analysis are likely to hold true when rebates are applied, although the magnitude of cost savings may vary.

Another limitation, common to many health economic analyses, particularly those of diabetes interventions, was the reliance on relatively short-term clinical trial data to make long-term projections. Every effort was made in the present analysis to minimize this, primarily by using a model of diabetes that has been extensively published and validated against real-life data, both on first publication and recently following a series of model updates^15,16. Projecting outcomes over patient lifetimes is recommended in guidelines for economic evaluation of interventions for patients with diabetes mellitus¹⁷.

Conclusions

The present long-term modeling analysis found that IDegLira was dominant vs liraglutide added to basal insulin for patients with type 2 diabetes failing to achieve glycemic control on basal insulin in the US, improving clinical outcomes and reducing direct costs. The improvements in risk factors associated with IDegLira resulted in a small reduction in the rates of diabetes-related complications over patient lifetimes. Cost savings were identified as a result of avoided diabetes complications and lower costs of treatment. IDegLira is highly likely to be considered cost-effective from a healthcare payer perspective in the US setting.

Transparency

Declaration of funding

The present cost-effectiveness analysis was supported by funding from Novo Nordisk A/S and Novo Nordisk Inc.

Declaration of financial/other relationships

BH and WV are employees of Ossian Health Economics and Communications, which received consulting fees from Novo Nordisk A/S and Novo Nordisk Inc. to support preparation of the analysis. MM is an employee of Novo Nordisk Inc. JL is an employee and shareholder of Novo Nordisk A/S. Peer reviewers on this manuscript have received an honorarium from JME for their review work, but have no other relevant financial relationships to disclose.

Acknowledgments

No assistance in the preparation of this article is to be declared.