Cost‐utility of once‐weekly insulin icodec versus once‐daily basal insulins for type 2 diabetes in China

Abstract

Objective

To comprehensively assess the cost‐utility of icodec, the first once‐weekly basal insulin, at 700 and 1050 U formulations compared with once‐daily basal insulin analogues in Chinese adults with type 2 diabetes within reimbursement context.

Methods

A long‐term cost‐utility analysis was performed using ONWARDS 5 data from the perspective of the Chinese healthcare system. Icodec (700 and 1050 U) was compared with once‐daily insulin glargine U100, glargine U300, and degludec over a 40‐year time horizon, applying a 5% annual discount rate. Outcomes included diabetes‐related complications, costs, and quality‐adjusted life‐years (QALYs). Sensitivity and scenario analyses were conducted to evaluate robustness and individual comparator results.

Results

Compared with once‐daily basal insulin analogues, icodec was associated with reduced cardiovascular mortality (RR = 0.97) and lower cumulative incidence of microvascular and macrovascular complications. Icodec yielded an additional 0.261 QALYs and increased life expectancy by 0.031 years, while reducing total costs by $2544.93 (700 U) and $2,916.92 (1050 U). Cost savings were primarily driven by fewer hypoglycemia events. Probabilistic sensitivity analysis showed 100% cost‐effectiveness at below 0.4 times China's GDP per capita. Across all scenario analyses, icodec was projected to be cost‐saving or cost‐effective over a lifetime horizon compared with once‐daily basal insulin comparators.

Conclusions

Once‐weekly icodec represents a cost‐effective and clinically beneficial alternative to once‐daily basal insulin analogues for Chinese adults with type 2 diabetes within the current reimbursement framework.

1. INTRODUCTION

The prevalence of diabetes in China has risen significantly over the past 3 decades from less than 1% in 1980 to 12.4% in 2018. ¹ According to the 11th International Diabetes Federation Diabetes Atlas, China has the largest diabetes population in the world, with an estimated 148.0 million people aged 20–79 years having diabetes in 2024. ² Type 2 diabetes (T2D) accounts for approximately 96% of these cases. ³ Among individuals with T2D, 76.4% reported at least one kind of complication, ⁴ which has become the leading cause of both death ⁵ and direct medical expenses in this population. ⁶ Effective glycemic control is crucial to preventing these complications; however, only 16.5% of Chinese T2D patients achieve adequate glycemic control. ¹ This is particularly concerning given the considerable disease burden in China. The nation reports the highest estimated number of diabetes‐related deaths globally at approximately 0.76 million and incurs the world's second‐highest diabetes‐related health expenditure at US$168.9 billion. ²

Insulin represents an essential glucose‐lowering therapy for patients with progressive T2D. Both Chinese and international guidelines recommend initiation of insulin therapy, primarily basal insulins, when non‐insulin glucose‐lowering medications fail to achieve adequate glycemic control.^{7, 8} It is estimated that approximately 20%–30% of patients with T2D will require insulin therapy over the course of their disease. ⁹ Despite their clinical importance, a recent study revealed that 82.1% of T2D patients in China exhibit psychological insulin resistance, and the dosing frequency is a key determinant of medication adherence in chronic disease. ¹⁰ With the research and development of basal insulin over the past century, the injection frequency was finally decreased from at least once daily to once weekly.

Insulin icodec, the world's first once‐weekly insulin analogue, was designed to improve glycemic control compared with once‐daily (OD) basal insulins. Clinical trials conducted across multiple regions, including China and the European Union (EU), demonstrated that icodec provides greater HbA1c reductions than OD basal insulins.^{11, 12, 13} Specifically, in two phase 3a randomized controlled trials (RCT) involving insulin‐naive patients with T2D, icodec showed superior efficacy versus insulin glargine U100 and insulin degludec.^{11, 12} The efficacy and safety of icodec have also been confirmed in post hoc analyses among Asian populations.^{14, 15} Subsequent investigation in the ONWARDS 5 trial, which incorporated both RCT and real‐world study elements, further assessed icodec's effectiveness and safety against OD analogues (degludec, glargine U300, and glargine U100), enhancing the generalizability of findings. ¹³ These three OD basal insulins represent the most commonly used third‐generation analogues in China and were all included in the second round of China's National Volume‐Based Procurement (VBP) specifically for insulin analogues. ¹⁶

Based on this demonstrated clinical efficacy, China approved icodec in June 2024, concurrently with the EU and other countries. Following successful price negotiations later that year, icodec was included in China's National Reimbursement Drug List (NRDL). Its NRDL price took effect on 1 January 2025, positioning China as the first country to cover this therapy under social health insurance based on health technology assessment (HTA). ¹⁷ Two icodec formulations are currently available in China: 1 mL: 700 U and 1.5 mL: 1050 U. Their unit prices differ in accordance with China's differential pricing rules. ¹⁸

Given the growing clinical interest in icodec, several economic evaluations have been conducted to assess its value. Some of them utilized hypothetical pricing to explore the potential cost‐effective thresholds of icodec relative to degludec or glargine U100.^{19, 20, 21} In the Canadian healthcare context, icodec was found to be a dominant or cost‐effective alternative to glargine U100, detemir, and degludec for T2D management. ²² Within China, recent economic evidence has also suggested that icodec offers long‐term advantages over degludec and glargine U100 among insulin‐naïve populations.^{23, 24} However, previous studies have not undertaken a comprehensive comparison of both the 700 and 1050 U formulations of icodec against all relevant OD analogues. In addition, the value of icodec compared with glargine U300 in China remains unknown. To address these gaps, this study evaluated the long‐term cost‐utility of icodec (700 and 1050 U) at its NRDL price in China, compared with OD analogues from a healthcare system perspective. As China has formally integrated HTA into reimbursement decisions for NRDL, ²⁵ this analysis provides empirical validation of icodec's value‐based pricing and offers a transferable framework for informing reimbursement decisions in other settings.

2. METHODS

This cost‐utility analysis was conducted in accordance with the Consolidated Health Economic Evaluation Reporting Standards (CHEERS). ²⁶ As this analysis does not directly involve human participants, it is exempt from institutional review board or ethics review per the Common Rule.

2.1. Decision model

This analysis used the validated Swedish Institute for Health Economics Diabetes Cohort Model (IHE‐DCM), a published and expert‐reviewed Markov cohort model simulating macrovascular and microvascular complications and premature mortality in T2D. ²⁷ The model operates in Microsoft Excel and accommodates both deterministic and probabilistic sensitivity analyses to address parameter uncertainty. It has been increasingly used in recent years for the economic evaluation of anti‐diabetic treatments globally.^{28, 29, 30}

Input parameters included baseline characteristics and risk factors of the cohort, treatment algorithms and treatment effects, unit costs, quality of life (QoL), and risk equations. Primary outcomes encompassed two domains: health outcomes measured in life years and quality‐adjusted life years (QALYs), alongside economic outcomes comprising total direct medical costs. The latter incorporated expenditures for anti‐hyperglycemic treatments as well as microvascular and macrovascular complications. We simulated icodec and OD analogues in parallel model runs. The complete model structure is detailed in Figure S1.

2.2. Model inputs: baseline characteristics, treatment effects, health utilities, and costs

Baseline characteristics were derived from the ONWARDS 5 trial (NCT04760626), a phase 3a randomized clinical trial incorporating real‐world elements. ¹³ This multinational study compared once‐weekly icodec with a dosing guide app versus OD analogues (glargine U100, glargine U300 and degludec) in 1085 insulin‐naive adults with T2D. The mean age of patients in the cohort at baseline was 59.3 years, and 43% were female, with the mean diabetes duration of 11.9 years (Table 1). Other baseline parameters are detailed in Table S1.

TABLE 1.

Key model input parameters.

Parameter	Base‐case point estimate	SE	Source
Clinical parameters
Baseline cohort characteristics
Age, years	59.3	—	Bajaj et al. 13
Female, %	43	—	Bajaj et al. 13
Diabetes duration, years	11.9	—	Bajaj et al. 13
HbA1c, %	8.92	—	Bajaj et al. 13
BMI, kg/m2	32.76	—	Bajaj et al. 13
Treatment effects
HbA1c reduction, %
Icodec	1.68	0.09	Bajaj et al. 13
OD analogues	1.31	0.12	Bajaj et al. 13
Intensification therapy	1.18	0.05	Mathieu et al. 31
SBP reduction, mmHg
Icodec	0.54	0.58	Bajaj et al. 13
OD analogues	1.42	0.57	Bajaj et al. 13
Intensification therapy	0.56	0.77	Mathieu et al. 31
DBP reduction, mmHg
Icodec	1.60	0.37	Bajaj et al. 13
OD analogues	1.76	0.37	Bajaj et al. 13
TC reduction, mmol/L
Icodec	0.18	0.052	Bajaj et al. 13
OD analogues	0.12	0.085	Bajaj et al. 13
Intensification therapy	0.12	0.053	Mathieu et al. 31
LDL‐C reduction, mmol/L
Icodec	0.02	0.040	Bajaj et al. 13
OD analogues	0.04	0.060	Bajaj et al. 13
Intensification therapy	0.07	0.041	Mathieu et al. 31
HDL‐C change, mmol/L
Icodec	0.03	0.010	Bajaj et al. 13
OD analogues	−0.03	0.017	Bajaj et al. 13
Intensification therapy	0.00	0.013	Mathieu et al. 31
TG reduction, mmol/L
Icodec	0.53	0.069	Bajaj et al. 13
OD analogues	0.39	0.107	Bajaj et al. 13
BMI increase, kg/m2
Icodec	0.85	0.13	Bajaj et al. 13
OD analogues	0.56	0.17	Bajaj et al. 13
Intensification therapy	0.75	0.09	Mathieu et al. 31
Clinically significant hypoglycemic event, per patient‐year
Icodec	0.19	0.00	Bajaj et al. 13
OD analogues	0.14	0.00	Bajaj et al. 13
Intensification therapy	5.61	0.00	Mathieu et al. 31
Severe hypoglycemic event, per patient‐year
Icodec	0.00	0.00	Bajaj et al. 13
OD analogues	0.00	0.00	Bajaj et al. 13
Intensification therapy	0.02	0.00	Mathieu et al. 31
Anti‐hyperglycemic treatment cost, $
Annual drug costs
Icodec (700 U)	598.74	59.87	—
Icodec (1050 U)	544.49	54.45	—
OD analogues	352.51	35.25	—
Intensification therapy	712.07	71.21	—
Annual needle costs
Icodec	14.47	1.45	—
OD analogues	101.69	10.17	—
Intensification therapy	406.49	40.65	—
Background cost	477.30	47.73	Xie et al. 45
Adverse event cost
Clinically significant hypoglycemic event	125.34	12.53	Su et al. 46
Severe hypoglycemic event	1964.55	196.46	Su et al. 46
Utility
Baseline	0.936	0.120	Li et al. 47
Microvascular complications
Eye disease
Background retinopathy	−0.023	0.002	Mok et al. 48
Proliferative retinopathy	−0.023	0.002	Mok et al. 48
Macular edema	−0.019	0.002	Li et al. 47
Proliferative retinopathy and macular edema	−0.023	0.002	Mok et al. 48
Severe visual loss	−0.049	0.005	Li et al. 47
Lower extremity disease
Symptomatic neuropathy	−0.026	0.003	Li et al. 47
Peripheral vascular disease	−0.032	0.003	Li et al. 47
Lower extremity amputation event	−0.177	0.018	Mok et al. 48
History of lower extremity amputation	−0.177	0.018	Mok et al. 48
Kidney disease
Microalbuminuria	−0.030	0.003	Li et al. 47
Macroalbuminuria	−0.030	0.003	Li et al. 47
End stage renal disease	−0.092	0.009	Li et al. 47
Macrovascular complications
Ischemic heart disease	−0.017	0.002	Mok et al. 48
Myocardial infarction
First myocardial infarction event	−0.041	0.004	CADTH 49
History of first myocardial infarction	−0.041	0.004	CADTH 49
Subsequent myocardial infarction event	−0.012	0.001	CADTH 49
History of subsequent myocardial infarction	−0.012	0.001	CADTH 49
Stroke			CADTH 49
First stroke event	−0.052	0.005	CADTH 49
History of first stroke	−0.052	0.005	CADTH 49
Subsequent stroke event	−0.04	0.004	CADTH 49
History of subsequent stroke	−0.04	0.004	CADTH 49
Heart failure	−0.05	0.005	Mok et al. 48
Demographic factors
Age, per 10 years	−0.024	0.002	Bagust et al.50, 51
Female	−0.012	0.001	Li et al. 47
Diabetes duration, per 10 years	−0.016	0.002	Bagust et al.50, 51
Overweigh, per BMI over 25	−0.006	0.001	Beaudet et al. 52
Adverse events, per episode
Hypoglycemia
Clinically significant hypoglycemia	−0.014	0.001	CADTH 49
Severe hypoglycemia	−0.047	0.005	CADTH 49
Disutility of injection frequency (vs. once‐weekly injection)
Once daily injection	−0.011	0.001	Liu et al. 33
Three times daily injections a	−0.023	0.002	Liu et al. 33

Note: Intensification therapy was glargine U100 plus aspart. For cost parameters lacking empirical variance data, the standard error (SE) was assumed to be 10% of the mean value in the probabilistic sensitivity analysis.

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HbA_1c, haemoglobin A_1c; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; SBP, systolic blood pressure; SE, standard error; TC, total cholesterol; TG, triglycerides.

^a Assume the disutility of injection for the intensification therapy.

Treatment effects were also sourced from ONWARDS 5. Patients remained on their assigned therapy until HbA1c exceeded 8.31%, a threshold derived from the baseline HbA1c of the insulin glargine U100 plus insulin aspart arm in ONWARDS 4. ³¹ This regimen represents China's most common intensification therapy. Upon exceeding this threshold, all patients were assumed to switch to intensification therapy, with equivalent post‐switch clinical efficacy applied across treatment arms.

Costs were assessed from the Chinese healthcare system perspective, encompassing direct medical costs for glucose‐lowering medications, needles, and diabetes‐related complications. Annual treatment costs were calculated using consumable doses and unit prices. For icodec (700 U), drug costs were based on its 2024 NRDL price. Pricing for the 1050 U formulation followed China's differential pricing policy, yielding lower unit costs. Prices for OD analogues were sourced from China's second insulin VBP program. Consistent with China's innovative drug evaluation framework, the highest VBP price was applied in the base‐case analysis. Dosing regimens and utilization proportions were derived from ONWARDS 5 and ONWARDS 4, respectively. Needle costs reflected national bidding averages. All costs for complication events and hypoglycemic episodes were obtained from published literature (Table S2). Costs were converted to 2024 using the Chinese Consumer Price Index and expressed in US dollars ($), using exchange rate of $1 = 7.1884 CNY (31 December 2024). Health outcomes and costs were discounted annually at 5%, consistent with Chinese Guidelines for Pharmacoeconomic Evaluation. ³²

Health utility and complication‐related disutility used in the model were primarily derived from published studies of Chinese and Asian populations. The disutility associated with injection frequency, based on a China‐specific study, was also integrated. ³³

2.3. Statistical analysis

A base case analysis was conducted to evaluate the long‐term cost‐effectiveness of two icodec formulations versus OD analogues. This study adopted the perspective of the Chinese healthcare system over a 40‐year time horizon, reflecting the chronic nature of T2D and capturing lifetime complications.

Both one‐way and probabilistic sensitivity analyses were performed specifically on two formulations to assess result robustness. The one‐way sensitivity analyses involved varying key parameters, including time horizon, discount rate, alternative HbA1c thresholds for intensification treatment, compared treatment cost, complication cost, icodec effects, and icodec price. Probabilistic sensitivity analysis employed nonparametric bootstrapping with 1000 iterations to quantify parameter uncertainty.

Scenario analyses were informed by post hoc subgroup efficacy data from the ONWARDS 5 trial, which provided direct comparator‐specific treatment effects of icodec versus glargine U100, glargine U300, and degludec. ³⁴ Both icodec formulations (700 and 1050 U) were evaluated against these specific OD analogue subgroups.

3. RESULTS

3.1. Modelled clinical outcomes

Long‐term simulation outcomes demonstrated comparable overall mortality between groups (RR = 1.00; Figure S2). However, icodec was associated with a reduction in cardiovascular mortality (RR = 0.97). The cumulative incidence of both microvascular and macrovascular complications was lower in the icodec group versus comparators, with pronounced reductions observed for eye diseases and kidney diseases, as detailed in Table 2.

TABLE 2.

Mortality and complication cumulative incidence in long‐term simulations.

Variable	Icodec	OD analogues	Relative risk
Mortality
Overall mortality	98.13%	98.26%	1.00
Cardiovascular event mortality	17.68%	18.17%	0.97
Microvascular complications
Eye disease
Background retinopathy	23.87%	27.95%	0.85
Proliferative retinopathy	2.55%	3.40%	0.75
Macular edema	13.47%	15.39%	0.88
Proliferative retinopathy and macular edema	7.56%	8.70%	0.87
Severe visual loss	14.05%	14.54%	0.97
Lower extremity disease
Symptomatic neuropathy	7.10%	8.07%	0.88
Peripheral vascular disease	26.61%	26.35%	1.01
Lower extremity amputation	24.31%	24.80%	0.98
Kidney disease
Microalbuminuria	23.48%	25.08%	0.94
Macroalbuminuria	12.52%	14.87%	0.84
End stage renal disease	9.16%	10.12%	0.90
Macrovascular complications
Ischemic heart disease	13.93%	14.39%	0.97
Myocardial infarction
First myocardial infarction	20.86%	21.58%	0.97
Second or subsequent myocardial infarction	2.87%	2.97%	0.97
Stroke
First stroke	8.16%	8.25%	0.99
Second or subsequent stroke	3.10%	3.10%	1.00
Heart failure	8.35%	8.26%	1.01
Cardiovascular disease (myocardial infarction and stroke)	26.88%	27.60%	0.97

Across both treatment arms, more than 80% of patients ultimately required intensification therapy over the modelled time horizon (Figure S3). Notably, the timing differed between groups. In the icodec arm, the mean time to intensification was approximately 8 years, whereas patients receiving OD basal insulin required intensification earlier, at a mean of approximately 5 years.

3.2. Base case analysis

Treated with icodec could prolong patients' life expectancy by 0.031 life years and improve quality‐adjusted life years by 0.261 QALYs (7.290 vs. 7.029) compared to the OD analogues (Table 3). The QALY improvement was mainly due to icodec's better efficacy and once‐weekly injection frequency, which could prolong patients' life years with a lower incidence of complications and better quality of life (Table S3).

TABLE 3.

Base case outcomes of long‐term cost‐effectiveness analysis for icodec.

Outcomes	Icodec (700 U)	Icodec (1050 U)	OD analogues	Increment (icodec 700 U vs. OD analogues)	Increment (icodec 1050 U vs. OD analogues)
Life years, year	12.296	12.296	12.265	0.031	0.031
QALYs	7.29	7.29	7.029	0.261	0.261
Total cost, $	45 615.58	45 243.45	48 160.37	−2544.93	−2916.92
Anti‐hyperglycemic treatment cost, $	21 313.51	20 941.52	23 259.00	−1945.50	−2317.48
Microvascular cost, $	14 956.74	14 956.74	15 420.12	−463.39	−463.39
Macrovascular cost, $	9345.20	9345.20	9481.25	−136.05	−136.05
ICER, $/QALY	—	—	—	Dominant	Dominant

Note: Anti‐hyperglycemic treatment cost included insulin cost, needle cost, background cost, and adverse events cost.

Abbreviations: ICER, incremental cost‐effectiveness ratio; OD, once‐daily; QALY, quality‐adjusted life year.

Over the model horizon, icodec 700 U demonstrated a $2544.93 cost reduction versus OD analogues, primarily driven by hypoglycemia‐related savings of $1423.68 (Figure S4). Icodec 1050 U showed greater economic advantage, with total costs reduced by $2916.92 compared to OD analogues, attributable to its lower unit pricing. In the base‐case analysis, both icodec formulations (700 and 1050 U) were cost‐effective versus OD analogues, offering greater long‐term clinical benefits at a lower cost over a lifetime horizon.

3.3. Sensitivity analysis

One‐way sensitivity analyses confirmed the robustness of base‐case findings, consistently demonstrating icodec's capacity to improve QALYs while reducing costs across all parameter variations (Table S4). The treatment intensification threshold emerged as the most influential parameter, generating the greatest deviation in the ICER from base‐case estimates. When the price of icodec increased by 50%, it was projected to remain cost‐saving over a 40‐year horizon. Even at a 150% price increase, icodec (1050 U) was no longer cost‐saving but remained cost‐effective, with an ICER of $10 281.10 per QALY, below China's per capita GDP threshold ($13 319.93/QALY). The ICER for icodec (700 U) under the same scenario was $13 850.19 per QALY (1.04 × GDP per capita).

Probabilistic sensitivity analysis further corroborated these findings. The scatterplot confirmed that all 1000 simulated incremental cost‐effectiveness ratio estimates fell below China's 2024 GDP‐based willingness‐to‐pay threshold of $13 319.93 (Figure 1). Concordantly, the cost‐effectiveness acceptability curve demonstrated a 100% probability of icodec being cost‐effective versus once‐daily insulin analogues at this threshold value (Figure S5).

FIGURE 1.

Scatterplot of probabilistic sensitivity analysis. (A) The scatterplot of probabilistic sensitivity analysis for icodec 700 U. (B) The scatterplot of probabilistic sensitivity analysis for icodec 1050 U.

3.4. Scenario analysis

Scenario analyses demonstrated robust cost‐effectiveness advantages of both icodec formulations against specific OD analogue subgroups (Table 4). Compared to glargine U100 and degludec subgroups, icodec 700 U yielded incremental gains of 0.303 and 0.275 QALYs with cost reductions of $4153.64 and $2636.75, respectively. In contrast, when compared with glargine U300, icodec 700 U remained cost‐effective, with an ICER of $4312.64 per QALY gained, well below 0.4 times the GDP per capita in China. Icodec 1050 U was consistently projected to achieve higher QALYs at lower total costs across all comparator subgroups.

TABLE 4.

Scenario analysis results of icodec versus OD analogue subgroups.

Comparison groups	Incremental QALYs	Incremental costs, $	ICER, $/QALY
Icodec 700 U vs. degludec	0.275 (7.298 vs. 7.023)	−2636.75 (45 541.57 vs. 48 178.32)	Dominant
Icodec 700 U vs. glargine U100	0.303 (7.267 vs. 6.964)	−4153.64 (45 834.12 vs. 49 987.76)	Dominant
Icodec 700 U vs. glargine U300	0.053 (7.277 vs. 7.225)	−227.31 (45 563.41 vs. 45 336.10)	4312.64 (<0.4 times GDP per capita)
Icodec 1050 U vs. degludec	0.275 (7.298 vs. 7.023)	−3008.88 (45 169.44 vs. 48 178.32)	Dominant
Icodec 1050 U vs. glargine U100	0.303 (7.267 vs. 6.964)	−4525.76 (45 461.99 vs. 49 987.76)	Dominant
Icodec 1050 U vs. glargine U300	0.053 (7.277 vs. 7.225)	−144.68 (45 191.42 vs. 45 336.10)	Dominant

4. DISCUSSION

This study provides the first comprehensive cost‐utility analysis of both the 700 U and 1050 U formulations of icodec with all relevant once‐daily basal insulins, evaluated at both the class level and the individual comparator level. In addition, this analysis extends the existing evidence by quantifying the economic value of icodec relative to glargine U300 in the Chinese healthcare setting. Both the 700 and 1050 U formulations of icodec were projected to be cost‐saving versus OD analogues among patients with T2D in China, offering superior long‐term clinical outcomes over a lifetime horizon. The reduced unit cost of the 1050 U formulation further contributed to additional savings. Sensitivity and scenario analyses confirmed that these findings remained robust across a wide range of input parameters and assumptions.

In comparison with previous economic evaluations of therapies for T2D, icodec was associated with a clinically meaningful incremental gain of 0.261 QALYs versus OD analogues.^{35, 36, 37, 38} The cost‐utility results of our study are consistent with findings from previous evaluations conducted in Canada and Italy, supporting the generalizability of the economic value of icodec across different healthcare systems.^{21, 22}

Our findings should be interpreted in the context of the ONWARDS 5 trial, which reported a statistically significant HbA1c reduction of 0.38% with icodec relative to OD analogues at 52 weeks. ¹³ This glycemic improvement translates into two key clinical benefits. First, effective glycemic control plays a critical role in preventing complications of T2D. ³⁹ Long‐term simulations in our model indicate that the superior efficacy of icodec in glycemic control significantly reduces both mortality and the incidence of microvascular and macrovascular complications. Second, improved glycemic management delays the need for treatment intensification. In our study, intensification therapy was initiated when HbA1c exceeded 8.31%. Such delay is clinically meaningful, as intensification treatment is associated with a higher risk of hypoglycemia, which in turn negatively impacts health‐related quality of life and increases healthcare costs. ³¹ This also explains the apparent discrepancy between short‐term trial data and our economic projections. While ONWARDS 5 reported comparable or slightly higher hypoglycemia rates for icodec, the lifetime model captures the cumulative benefit of postponing the high‐hypoglycemia risk period inherent to intensification therapy.

Although the projected gain in life expectancy (0.031 years) appears modest, it must be considered alongside the substantially higher gain in QALYs (0.261). This discrepancy suggests that the economic value of icodec is driven predominantly by improvements in quality of life and reductions in morbidity related to chronic complications, rather than by mortality benefits alone. Furthermore, the lifetime simulation ensures that approximately 98% of the cohort reached the mortality endpoint, supporting the long‐term reliability of these estimates. The pattern of incremental QALYs exceeding incremental life‐years is well documented in economic evaluations of interventions for T2D,^{29, 35, 36, 40} and reflects the increasing emphasis on improving the quality of living with T2D rather than extending survival alone.

Non‐adherence to insulin therapy remains a critical challenge in diabetes management, with common patient‐reported barriers including needle fear and social stigma associated with public injections. ⁴¹ As the first once‐weekly basal insulin, icodec may offer a promising approach to improving treatment adherence and potentially transforming insulin treatment paradigms. Although direct evidence comparing adherence to icodec versus OD analogues is limited, insights from studies of other injectable therapies provide indirect support. For instance, once‐weekly glucagon‐like peptide‐1 receptor agonists have demonstrated improved adherence and greater HbA1c reductions compared to daily regimens. ⁴² Furthermore, a multinational utility investigation identified significantly greater disutility associated with higher injection frequencies, with the most pronounced burden observed in China, ⁴³ highlighting the importance of incorporating injection‐related disutility into economic evaluations. While the ONWARDS 5 trial provided clinical inputs for this analysis, its controlled setting does not capture real‐world adherence differences between once‐weekly icodec and OD analogues. In real‐world practice, the reduced injection frequency of icodec is expected to enhance adherence, thereby improving long‐term clinical outcomes. Consequently, our results likely represent a conservative estimate of icodec's long‐term cost‐utility.

Scenario analyses confirmed the base‐case findings, albeit with minor differences. Icodec 700 U was projected to be cost‐saving compared with degludec and glargine U100, while remaining cost‐effective versus glargine U300. This differential outcome is likely attributable to the lower unit cost of glargine U300, which was $0.0098 per unit lower than the highest bid price observed for glargine U100 and degludec. Additionally, due to the limited sample size in the glargine U300 subgroup, ¹³ the post hoc results should be interpreted with caution. Greater confidence in the clinical outcomes for this subgroup would necessitate a larger patient cohort. Conversely, icodec 1050 U was consistently projected to achieve higher QALYs at lower total costs across all OD basal insulin comparators, including degludec, glargine U100, and glargine U300. This robust result was primarily driven by the reduced unit cost of icodec 1050 U, which benefits from China's differential pricing policies. This cost advantage could position icodec 1050 U as a potentially more attractive formulation in China's insulin market.

5. LIMITATIONS

Several potential limitations should be acknowledged. First, long‐term projections are based on 52‐week clinical trial data, which introduces uncertainty regarding complications and mortality. This was addressed through extensive sensitivity and scenario analyses, which demonstrated the robustness of the findings. Second, the clinical efficacy inputs were derived from multinational trials in which a digital dosing guide app was utilized for the icodec arm. As this app is currently unavailable in China, the trial‐based HbA1c reductions might be higher than those achievable in real‐world Chinese practice. To mitigate this uncertainty, we performed sensitivity analyses by reducing the efficacy of icodec by up to 20%, under which icodec remained a cost‐saving option. Third, the risk equations used in the IHE‐DCM were not specifically developed for the Chinese population. Moreover, the direction of potential bias is difficult to determine, as differences in cardiovascular and renal risk profiles between Western and Chinese populations may lead to multidirectional rather than systematic biases across modelled outcomes. ⁴⁴ Forth, this study focused on cost data specific to China. Although one‐way sensitivity analyses tested higher icodec prices, including a 150% increase, the findings should be generalized to other settings with caution. Finally, as icodec is a newly introduced therapy, real‐world evidence on long‐term adherence and clinical outcomes relative to OD analogues remains unavailable.

6. CONCLUSIONS

In conclusion, icodec is projected to be cost‐saving or cost‐effective compared with OD analogues, offering improved health outcomes over a lifetime horizon. This finding supports the alignment of icodec's reimbursement decision in China with the principles of value‐based healthcare. The innovative weekly administration of icodec has the potential to significantly improve patient adherence, further underscoring its capacity to transform diabetes management. These results provide compelling post hoc justification for the inclusion of icodec in China's NRDL and offer a robust, transferable framework for value‐based reimbursement decisions globally.

AUTHOR CONTRIBUTIONS

L.‐Y. Zhang had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Design: Tong, Xing, L.‐Y. Zhang, L. Zhang. Conduct/data collection: Tong, Xing, L.‐Y. Zhang, L. Zhang. Analysis: Tong, Xing. Writing manuscript: Tong, Xing, L. Zhang.

FUNDING INFORMATION

This study was funded by the National Natural Science Foundation of China (grant no. 72374049) and Novo Nordisk (China) Pharmaceutical Co. Ltd.

CONFLICT OF INTEREST STATEMENT

L.‐Y. Zhang, the Principle Investigator of this study is supported by Novo Nordisk (China) Pharmaceutical Co. Ltd., Beijing, China. Xing and L. Zhang are employees of Novo Nordisk (China) Pharmaceutical Co. Ltd. and may hold stock options.

Supporting information

FIGURE S1. Structure of the IHE‐DCM model.

TABLE S1: Baseline demographics and risk factors derived from ONWARDS 5.

TABLE S2: Costs of complication events and hypoglycemia.

TABLE S3: Breakdown of quality‐adjusted life years (QALYs).

FIGURE S2: Overall survival rate in both groups.

FIGURE S3: Proportion of patients by time to initial treatment intensification.

FIGURE S4: Direct costs over patients' lifetimes ($).

TABLE S4: One‐way sensitivity analysis of long‐term cost‐effectiveness analysis.

FIGURE S5: Cost‐effectiveness acceptability curve for icodec 700 and 1050 U.

Tong X, Xing C, Zhang L, Zhang L. Cost‐utility of once‐weekly insulin icodec versus once‐daily basal insulins for type 2 diabetes in China. Diabetes Obes Metab. 2026;28(5):3872‐3882. doi: 10.1111/dom.70567

DATA AVAILABILITY STATEMENT

The data and materials generated or analysed during this study are available from the corresponding author upon reasonable request.