Cost-effectiveness of different strategies with biologics for the treatment of moderate to severe Crohn’s disease in China

Abstract

BACKGROUND

Biologics are the preferred treatment for patients with moderate to severe Crohn’s disease (CD). Four biologics included in China’s National Reimbursement Drug List are available for CD treatment. Due to loss of response, patients need switching to another biologic, making it necessary to evaluate the cost-effectiveness of different biologic treatment sequences.

AIM

To assess the cost-effectiveness of sequential treatment strategies with National Reimbursement Drug List-included biologics for moderate to severe CD in China.

METHODS

From a healthcare system perspective, a Markov model was constructed to evaluate the cost-effectiveness of four biologics [infliximab, adalimumab (ADA), ustekinumab (UST), and vedolizumab] applied in different treatment sequences for moderate to severe CD patients. Using one times the GDP per capita ($13444.68, 2024) in China as the willingness-to-pay threshold, the absolute net monetary benefit (NMB) and incremental cost-effectiveness ratio (ICER) were calculated. Both costs and utilities were discounted at an annual rate of 5%. Sensitivity analyses were conducted on key parameters.

RESULTS

With $13444.68 as willingness-to-pay, among the 17 treatment sequences evaluated for biologic-naïve patients, sequence 3 (ADA-UST) yielded the highest absolute NMB of $35850.93. Compared with sequence 1 (vedolizumab-UST), sequence 3 had the most favorable ICER of $2285.38/quality-adjusted life year. For biologic-exposed patients, sequence 3 still demonstrated the optimal NMB and ICER results.

CONCLUSION

Adding biologic treatment lines provides greater health benefits for patients with moderate to severe CD. Among the various sequential strategies, the treatment sequence combining ADA and UST is more likely to be the optimal cost-effective option in China.

Core Tip: Using a Markov model, this study evaluated the cost-effectiveness of 17 sequential biologic strategies for moderate-to-severe Crohn’s disease in China. Under a willingness-to-pay threshold of one time the 2024 GDP per capita, the findings suggest that while additional lines of biologic therapy improve health outcomes, they concurrently increase costs. The sequence initiating with adalimumab and followed by ustekinumab emerged as the most cost-effective option, yielding the highest net monetary benefit and a favorable incremental cost-effectiveness ratio for both biologic-naïve and biologic-exposed patients.

INTRODUCTION

Crohn’s disease (CD) is a chronic inflammatory bowel disease (IBD) that can affect the entire digestive tract from the mouth to the anus, with predominant involvement of the ileum, cecum, and colon[1]. CD presents with complex clinical manifestations, including abdominal pain, diarrhea, intra-abdominal abscesses, recurrent fistula formation, and perianal lesions, all of which significantly impair patients’ quality of life[2]. Over the past three decades, the global prevalence of IBD has shown a continuous upward trend, potentially imposing substantial social and economic burdens on governments and healthcare systems[3]. The incidence of CD tends to increase with national industrialization[4]. Currently, the incidence of CD in urban China is 0.71 per 100000 person-years, with a higher incidence among males than females and a peak onset age between 30 years and 34 years old[5]. The long-term progression of CD could lead to severe disease-related complications. The cumulative risk of surgery at one year after diagnosis is 16%, with high-risk patients facing an even greater likelihood of adverse outcomes such as surgery, hospitalization, and disability[6,7].

The development of biologics has revolutionized the treatment of moderate to severe CD, shifting the paradigm from the use of corticosteroids, 5-aminosalicylic acid, and non-targeted immunosuppressants aimed at symptom control, to the early application of biologics to achieve clinical and endoscopic remission[8]. As of 2025, four biologics have been included in China’s National Reimbursement Drug List (NRDL) for the treatment of moderate to severe CD: (1) Infliximab (IFX); (2) Adalimumab (ADA); (3) Ustekinumab (UST); and (4) Vedolizumab (VDZ). Although biologics can control clinical symptoms and promote mucosal healing, their high cost may increase the economic burden on patients. Additionally, due to the risk of loss of response, patients may need switching therapies during the course of treatment. While previous studies have evaluated the cost-effectiveness of individual biologics in the Chinese CD population[9-12], there is a lack of cost-effectiveness analyses on sequential treatment strategies involving multiple biologics.

Therefore, from a healthcare system perspective, this study aims to evaluate the cost-effectiveness of different sequential treatment strategies involving the four NRDL-covered biologics for patients with moderate to severe CD in China. The findings are expected to inform clinical decision-making, optimize therapeutic outcomes, and ultimately improve long-term prognosis and quality of life for patients.

MATERIALS AND METHODS

Model overview

The four biologics selected in the study (IFX, ADA, UST, VDZ) have been both included in NRDL and clinical practice guidelines[2]. However, existing guidelines did not explicitly recommend specific treatment sequences or detailed switch strategies. Therefore, in order to systematically and comprehensively evaluate the possible sequencing of different biologics in the real world and their economics, 17 biologic treatment sequences were defined. Crohn’s Disease Activity Index (CDAI) score was used to evaluate the efficacy and disease status, which was a commonly used efficacy endpoint in clinical trials. Specifically, clinical remission was defined as a post-treatment CDAI score < 150, and clinical response was defined as a reduction in CDAI score ≥ 100 points from baseline to the end of treatment. Based on the disease progression and clinical practice of CD, as well as reference to internationally published pharmacoeconomic evaluations[13], a Markov model (Figure 1) was constructed to simulate the disease state transitions in patients with moderate to severe CD. Patients entered the model in the “active disease” state and initiate treatment with a first-line biologic. Treatment typically consisted of two phases: (1) Induction phase; and (2) Maintenance phase. The induction phase employed a higher dose or increased frequency of administration for one cycle, followed by the maintenance phase, during which treatment continued at a reduced dosing regimen until loss of response. During induction, patients who achieved a clinical response or clinical remission entered corresponding states in the maintenance phase, where they continued the same biologic therapy. Patients remained in the “clinical response” or “clinical remission” state until loss of response, at which point they returned to the “active disease” state and switched to another biologic.

Figure 1.

Treatment sequence and model structure for moderate to severe Crohn’s disease. ADA: Adalimumab; IFX: Infliximab; UST: Ustekinumab; VDZ: Vedolizumab.

Patients who did not achieve clinical response or clinical remission during the induction phase remained in the “active disease” state and switched to another biologic, repeating the “induction + maintenance” treatment process. Those who remained in the “active disease” state after two or three treatment switches underwent surgical intervention, which involved resection of the affected bowel segment. Most patients experienced symptom relief after surgery and transitioned to the “post-surgical remission” state, while some experienced ongoing disease activity and entered the “post-surgical active disease” state. Since CD could affect the entire digestive tract, recurrence was possible even after intestinal resection; therefore, patients in “post-surgical remission” could also transition to the “post-surgical active disease” state. “Death” was an absorbing state, and transition to this state was possible from any health state.

In accordance with the dosing patterns of biologics for CD, the model cycle length was set at 8 weeks, and the time horizon was set to 10 years or a lifetime.

Clinical efficacy and transition probability

Patients were categorized into two groups based on prior exposure to biologic therapy: (1) Biological-naïve; and (2) Biological-exposed. The efficacy of each biologic in treating biological-naïve and biological-exposed patients was calculated based on the odds ratios (ORs) compared to placebo (PBO). The ORs for clinical response rates and clinical remission rates were derived from two network meta-analyses[14,15]. For biologics with alternative administration regimens resulting in different OR values, the OR corresponding to the dosage and frequency consistent with the cost calculations in the Markov model was selected for analysis. The probabilities for the biologics were calculated based on the formula: P_bio = (P_pbo × OR_{bio vs pbo})/(1 - P_pbo + P_pbo × OR_{bio vs pbo}), where P_bio and P_pbo were the probability of clinical response or clinical remission for biologics and PBO, separately. The probability for PBO came from a meta-analysis of clinical trials in moderate to severe CD[16]. For missing efficacy data of biologic-exposed populations related to IFX and ADA, to avoid overestimating the efficacy of these two biologics, a conservative assumption was adopted[13], using the lowest reported OR among other biologics as a substitute (Table 1 and Supplementary Table 1).

Table 1.

Summary of efficacy and safety of biologics for treating moderate to severe Crohn’s disease

	Biological-naïve		Biological-exposed		Serious adverse event (%)
	Response rate (%)	Remission rate (%)	Response rate (%)	Remission rate (%)
Induction phase
IFX	86.12	43.27	30.68	12.28	6.61
ADA	41.26	33.63	35.38	28.29	4.46
UST	44.04	30.67	38.18	17.96	5.28
VDZ	32.67	25.09	30.68	12.28	6.87
PBO	22.00	11.00	20.00	10.00	7.00
Maintenance phase
IFX	62.17	47.34	34.06	17.45	16.14
ADA	87.80	84.51	34.06	17.45	11.27
UST	89.84	84.90	69.90	49.76	13.16
VDZ	61.13	46.81	34.06	17.45	16.72
PBO	39.00	32.00	26.00	16.00	17.00

ADA: Adalimumab; IFX: Infliximab; PBO: Placebo; UST: Ustekinumab; VDZ: Vedolizumab.

Safety parameters focused specifically on the incidence of serious adverse events (SAEs). The SAE rates were calculated based on the ORs of SAEs for each biologic compared to PBO[14], and the incidence rates of SAEs for PBO during both the induction and maintenance phases (Table 1 and Supplementary Table 2)[16].

The efficacy and safety parameters for both the induction and maintenance phases were converted into transition probabilities per model cycle and incorporated into the model, following the approach by Lohan et al[17] (Supplementary Table 3). Other transition probability parameters, including transitions probabilities between post-surgical health states and mortality, were presented in Table 2[2,5,11,13,18-20].

Table 2.

Model inputs of cohort characteristics, transition probability, costs, and utility

Parameters	Value	Ref.
Cohort characteristics
Age	30	Xu et al[5], 2023
Female (%)	46.65	Xu et al[5], 2023
Weight (kg)	60
Transition probability
Post-surgical remission	52.68%	Bouhnik et al[13], 2023
Post-surgical active disease	47.32%	Bouhnik et al[13], 2023
Recurrence after surgery	3.50%	Bouhnik et al[13], 2023
Death rate	Age-gender-specific death rate	Cai[20], 2012
Costs
Infliximab	$204.40/100 mg IV	Induction: 5 mg/kg, 0/2/6 weeks; maintenance: 5 mg/kg, every 8 weeks
Adalimumab	$144.71/40 mg SC	Induction: 160 mg 0 week, 80 mg 2 weeks; maintenance: 40 mg, every 2 weeks
Ustekinumab	$302.11/130 mg IV $958.59/90 mL SC	Induction: 390 mg; maintenance: 90 mg, every 12 weeks
Vedolizumab	$699.27/300 mg IV	Induction: 300 mg, 0/2/6 weeks; maintenance: 300 mg, every 8 weeks
SAE	$1507.46/cycle	Zhou et al[11], 2021
Surgery	$9849.27/cycle	Average colectomy cost of China Healthcare Security Diagnosis Related Group 2.0
Conventional therapy (induction)	$125.42/cycle	Inflammatory Bowel Disease Group[2], 2024
Conventional therapy (maintenance)	$103.79/cycle	Inflammatory Bowel Disease Group[2], 2024
Utility
Active disease	0.40	Bouhnik et al[13], 2023; Lindsay et al[18], 2008
Clinical remission	0.83	Bouhnik et al[13], 2023; Lindsay et al[18], 2008
Clinical response	0.55	Bouhnik et al[13], 2023; Lindsay et al[18], 2008
Surgery	0.40	Bouhnik et al[13], 2023
Post-surgical remission	0.67	Bouhnik et al[13], 2023; Lindsay et al[18], 2008
Post-surgical active disease	0.40	Bouhnik et al[13], 2023
Disutility of SAE	0.07	Worbes-Cerezo et al[19], 2019

IV: Intravenous injection; SAE: Serious adverse event; SC: Subcutaneous injection.

Costs

The cost parameters from the healthcare system perspective included expenses for biologics, management of adverse events, surgery, and conventional therapy for recurrence patients after surgery. All biologic costs were calculated using the latest publicly available winning bid prices. For the four biologics included in the NRDL, these bid prices correspond to the official medical insurance payment prices. For IFX and ADA, which have several biosimilars available, the average market price was used to calculate the drug cost. For UST and VDZ, which are original products, the cost was calculated based on brand name drug price. All the biologics costs were calculated based on the recommended adult dosage specified in the prescribing information. The model assumed that patients received the standard recommended maintenance dose and changed the treatment directly after loss of response, without dose adjustment. For weight-based biologics, the dosage and cost were calculated assuming a standard patient body weight of 60 kg.

The cost of intestinal resection was the average cost of China Healthcare Security Diagnosis Related Group 2.0 for this procedure, representing a bundled payment for the entire surgical procedure. Patients who experienced postoperative recurrence received conventional therapy recommended by the Chinese guidelines for the diagnosis and treatment of CD[2], with adverse events monitored each cycle.

All costs were adjusted to 2024 values using the Chinese healthcare price index and converted to United States dollars based on the 2024 average exchange rate ($1 = CNY 7.12) (Table 2)[2,5,11,13,18-20].

Utilities

As health state preferences in the Chinese population have not been reported, the utility values for each state were derived from published literatures[13,18], which estimated utility values for relevant health states in adult CD patients in the United Kingdom. The disutility associated with SAEs was obtained from Worbes-Cerezo et al[19] (Table 2).

Statistical analysis

The base-case analysis was conducted with the initial biologic-naïve patient cohort. Using one times the GDP per capita ($13444.68, 2024) in China as the willingness-to-pay (WTP) threshold, the absolute net monetary benefit (NMB) was calculated for each treatment sequence, with the formula: NMB = cost × WTP - quality-adjusted life year (QALY)[21]. With sequence 1 as the reference, the incremental cost-effectiveness ratio (ICER) of other sequences was calculated, representing the additional cost required to gain one additional QALY. Both costs and utilities in the model were discounted at an annual rate of 5%, and a half-cycle correction was applied.

One-way sensitivity analysis and probabilistic sensitivity analysis were employed to evaluate the impact of variations in single or multiple key parameters on the economic evaluation results. For the one-way sensitivity analysis, the OR parameters were assigned ranges based on the 95%CI of the OR values obtained from the literature, and other parameters were varied by assuming plausible ranges of the base-case values (± 20%). The probabilistic sensitivity analysis was conducted, using Monte Carlo simulations of 5000 iterations by randomly sampling key parameters from the preset probability distributions. Specifically, cost parameters followed a gamma distribution, OR values were assigned a lognormal distribution, transition probabilities were modeled using a normal distribution, and utility values were assumed to follow a beta distribution. A cost-effectiveness acceptability curve (CEAC) was plotted to illustrate the probability of each treatment sequence being cost-effective compared to sequence 1 across a range of WTP thresholds (Supplementary Table 4).

In addition to the biologic-naïve patients included in the base-case analysis, a scenario analysis was further conducted on biologic-exposed patients. Given that prior biologic exposure has a significant impact on treatment efficacy, and considering that approximately 37%[22] of CD patients in China have already received biologic therapy, this patient population still requires switching to alternative treatments after loss of response. Therefore, a scenario analysis was performed with biologic-exposed patients as the initial cohort. Since IFX was the earliest biologic introduced into China, and accounts for over 73.7%[22] of biologic exposures, treatment sequences containing IFX were excluded in this scenario analysis. This approach assumed that the initial patient population consisted entirely of those who had experienced treatment failure with IFX.

RESULTS

Base case analysis

From a healthcare system perspective and over a ten-year time horizon, the base-case analysis results for Chinese patients with moderate to severe CD were presented in Table 3 and Figure 2. The NMB analysis showed that for biologic-naïve patients, sequence 3 (ADA-UST) yielded the highest absolute NMB at $35850.93, followed by sequence 13 (ADA-UST-VDZ, $34207.27) and sequence 14 (ADA-VDZ-UST, $34062.39). In terms of QALYs, compared to sequence 1 (VDZ-UST), the incremental QALYs of other treatment sequences ranged from a minimum of 0.069 (sequence 16: VDZ-ADA-UST) to a maximum of 0.486 (sequence 13: ADA-UST-VDZ).

Table 3.

Summary of net monetary benefit and incremental cost-effectiveness ratio results in base case analysis comparing different biologic sequences for biologic-naïve Crohn’s disease patients

Sequence	Costs ($)	QALY	Net monetary benefit ($)	Incremental costs ($)	Incremental QALYs	Incremental cost-effectiveness ratio ($/QALY)1
Sequence 1: VDZ-UST	22392.37	3.967	30948.45	-	-	-
Sequence 2: IFX-ADA	23555.90	4.117	31790.65	1163.52	0.149	7799.27
Sequence 3: ADA-UST	23396.38	4.407	35850.93	1004.01	0.439	2285.38
Sequence 4: IFX-UST	24697.41	4.272	32741.92	2305.04	0.305	7561.41
Sequence 5: ADA-VDZ	23127.51	4.248	33981.33	735.13	0.280	2623.04
Sequence 6: UST-VDZ	24271.51	4.335	34005.35	1879.14	0.367	5118.36
Sequence 7: IFX-VDZ	24446.82	4.101	30686.86	2054.45	0.133	15406.35
Sequence 8: IFX-UST-VDZ	27191.67	4.322	30920.52	4799.30	0.355	13523.37
Sequence 9: IFX-ADA-VDZ	26236.50	4.171	29844.52	3844.12	0.204	18861.04
Sequence 10: IFX-UST-ADA	26366.71	4.337	31936.78	3974.34	0.369	10767.13
Sequence 11: IFX-ADA-UST	26397.00	4.333	31856.02	4004.63	0.365	10960.66
Sequence 12: IFX-VDZ-UST	27305.37	4.318	30751.11	4913.00	0.351	14007.32
Sequence 13: ADA-UST-VDZ	25665.78	4.453	34207.27	3273.41	0.486	6737.35
Sequence 14: ADA-VDZ-UST	25762.83	4.450	34062.39	3370.46	0.482	6988.27
Sequence 15: UST-VDZ-ADA	25821.32	4.396	33277.94	3428.95	0.428	8005.84
Sequence 16: VDZ-ADA-UST	24281.88	4.037	29989.99	1889.51	0.069	27285.02
Sequence 17: VDZ-UST-ADA	24256.07	4.040	30061.27	1863.70	0.073	25659.35

¹Compared to sequence 1 (vedolizumab-ustekinumab).

ADA: Adalimumab; IFX: Infliximab; QALYs: Quality-adjusted life years; UST: Ustekinumab; VDZ: Vedolizumab.

Figure 2.

Cost-effectiveness scatter plot of different treatment sequences for biologic-naïve Crohn’s disease patients. ADA: Adalimumab; IFX: Infliximab; QALYs: Quality-adjusted life years; UST: Ustekinumab; VDZ: Vedolizumab.

As shown in Table 3, among the ICERs for all treatment sequences compared to sequence 1, sequence 3 (ADA-UST) had the lowest ICER at $2285.38 per QALY, indicating that it required the lowest additional cost to gain one additional QALY. This was followed by sequence 5 (ADA-VDZ) and sequence 6 (UST-VDZ), with ICERs of $2623.04/QALY and $5118.36/QALY, respectively. The ICERs for sequences 2-17 were all below three times the GDP per capita of China.

When the time horizon was extended to a lifetime, sequence 3 remained the strategy with the highest absolute NMB and the most favorable ICER results (Supplementary Table 5).

Sensitivity analyses

Taking the one-way sensitivity analysis comparing sequence 1 (VDZ-UST) and sequence 3 (ADA-UST) as an example (Figure 3), the tornado diagram indicated that the maintenance phase drug costs of biologics, the OR for clinical response rate of biologics vs PBO during the maintenance phase, and the utility value of the clinical remission state were the key factors influencing the model results. From the results of probabilistic sensitivity analysis between sequence 3 (ADA-UST) and sequence 1 (VDZ-UST), the scatter plot results showed that most of the points were in the first quadrant and within one times the GDP per capita (Figure 4).

Figure 3.

Tornado diagram for one-way sensitivity analysis: Sequence 1 (vedolizumab-ustekinumab) vs sequence 3 (adalimumab-ustekinumab) in biologic-naïve Crohn’s disease patients. ADA: Adalimumab; Bio: Biologics; ICER: Incremental cost-effectiveness ratio; IFX: Infliximab; OR: Odds ratio; PBO: Placebo; QALYs: Quality-adjusted life years; SAEs: Serious adverse events; UST: Ustekinumab; VDZ: Vedolizumab.

Figure 4.

Cost-effectiveness scatter plot of sequence 1 (vedolizumab-ustekinumab) vs sequence 3 (adalimumab-ustekinumab) in biologic-naïve Crohn’s disease patients. PSA: Probabilistic sensitivity analysis; QALYs: Quality-adjusted life years.

According to the CEAC derived from the probabilistic sensitivity analysis (Figure 5), for biologic-naïve CD patients, at a WTP threshold equivalent to one time the GDP per capita, the probabilities of sequences 3, 5, and 6 being cost-effective compared to sequence 1 were 98.60%, 91.84%, and 90.04%, respectively. According to the CEAC derived from the probabilistic sensitivity analysis (Figure 5), for biologic-naïve CD patients, at a WTP threshold equivalent to one time the GDP per capita, the probabilities of sequences 3, 5, and 6 being cost-effective compared to sequence 1 were 98.60%, 91.84%, and 90.04%, respectively.

Figure 5.

Cost-effectiveness acceptability curve of treatment sequences vs sequence 1 for biologic-naïve Crohn’s disease patients. QALYs: Quality-adjusted life years.

Scenario analyses

The scenario analysis results (Table 4) showed that after excluding sequences containing IFX, for biologic-exposed patients over a 10-year time horizon, sequence 3 (ADA-UST) yielded the highest absolute NMB at $31150.98, followed by sequence 1 (VDZ-UST, $30363.95) and sequence 6 (UST-VDZ, $29883.92). In terms of QALYs, compared to sequence 1 (VDZ-UST), the incremental QALYs of other treatment sequences ranged from a minimum of -0.184 (sequence 5: ADA-VDZ) to a maximum of 0.079 (sequence 15: UST-VDZ-ADA). Regarding the ICER, sequence 3 (ADA-UST) was the dominant strategy. When the time horizon was extended to a lifetime, sequence 3 (ADA-UST) remained the treatment strategy with the highest absolute NMB and the most favorable ICER results (Supplementary Table 6).

Table 4.

Summary of net monetary benefit and incremental cost-effectiveness ratio results comparing different biologic sequences for biologic-exposed Crohn’s disease patients

Sequence	Costs ($)	QALY	Net monetary benefit ($)	Incremental costs ($)	Incremental QALYs	Incremental cost-effectiveness ratio ($/QALY)1
Sequence 1: VDZ-UST	21375.37	3.848	30363.95	-	-	-
Sequence 3: ADA-UST	20813.64	3.865	31150.98	-561.73	0.017	-33520.57
Sequence 5: ADA-VDZ	20304.88	3.664	28962.46	-1070.49	-0.184	5822.22
Sequence 6: UST-VDZ	21899.71	3.852	29883.92	524.34	0.003	159090.08
Sequence 13: ADA-UST-VDZ	23620.33	3.924	29137.76	2244.97	0.076	29626.51
Sequence 14: ADA-VDZ-UST	23714.52	3.921	28995.66	2339.16	0.072	32392.94
Sequence 15: UST-VDZ-ADA	23814.60	3.927	28983.03	2439.23	0.079	30987.73
Sequence 16: VDZ-ADA-UST	23315.89	3.920	29393.93	1940.52	0.072	26882.61
Sequence 17: VDZ-UST-ADA	23291.21	3.924	29462.02	1915.84	0.075	25404.50

¹Compared to sequence 1 (vedolizumab-ustekinumab).

ADA: Adalimumab; QALYs: Quality-adjusted life years; UST: Ustekinumab; VDZ: Vedolizumab.

DISCUSSION

Patients with moderate to severe CD often experience a variety of symptoms, including diarrhea, abdominal pain, extraintestinal manifestations, and anorectal complications, which significantly impair their quality of life. Therefore, selecting optimal treatment strategies to achieve symptomatic relief and minimize the risks of surgery, hospitalization, and disease-related complications is essential. While previous studies have compared the efficacy of various biologics through indirect methods such as network meta-analysis, there remains a lack of evidence regarding the comparative health outcomes and cost-effectiveness of different sequential biologic therapy regimens.

This study evaluated the cost-effectiveness of 17 treatment sequences involving four biologics included in the NRDL. The results demonstrated that for both biologic-naïve and biologic-exposed patients, sequence 3 (ADA-UST) yielded the highest absolute NMB. According to the ICER results, multiple treatment sequences (sequences 2-6, 10-11, and 13-15 in biologic-naïve patients; sequence 3, 5 in biologic-exposed patients) exhibited ICER values below one times the GDP per capita of China, indicating that these sequences were more cost-effective relative to sequence 1 (VDZ-UST).

Sensitivity analysis identified the cost and efficacy of biologics as key factors influencing the ICER. When comparing the NMB across different treatment sequences, those incorporating both ADA and UST demonstrated higher absolute NMB values, particularly sequences in which ADA was used as first-line therapy and UST was administered as either first-line or second-line treatment. These results highlight the efficacy and cost advantages of ADA and UST. Consistent with previously reported comparative efficacy outcomes in published studies[14,15], ADA showed superior outcomes in biologic-naïve populations, while UST demonstrated favorable results in both biologic-naïve and biologic-exposed patients, particularly during the maintenance phase. In contrast, VDZ exhibited relatively weaker clinical efficacy, leading to poorer cost-effectiveness performance in treatment sequences containing VDZ. In terms of treatment costs, the maintenance phase expenses of all four biologics were comparable, with ADA being the lowest.

Regarding health outcomes across treatment sequences, four of the five regimens with the highest total QALYs involved the sequential use of three biologics (sequence 10, 13-15), suggesting that utilizing a greater number of treatment lines may improve health outcomes. This finding was consistent with other cost-effectiveness analyses in IBD[13,23], which indicated that the superior efficacy of biologics over conventional therapy made additional treatment lines beneficial for patient health outcomes. However, the significantly higher cost of biologics compared to conventional therapy also increased the total costs of therapy. As a result, sequences containing two biologics (sequences 3, 5, and 6) demonstrated more favorable ICER results.

The scenario analysis results indicated that for biologic-exposed patients, increasing the number of biologic treatment lines also contributed to improved health outcomes. However, due to the significantly higher treatment costs, sequences containing two biologics yielded greater NMB and more favorable ICERs, particularly sequence 3 (ADA-UST) and sequence 1 (VDZ-UST). Sequence 3 (ADA-UST) demonstrated the highest NMB in both biologic-naïve and biologic-exposed populations, suggesting that the efficacy advantages of these two biologics in different patient groups, especially the superior clinical response and remission rates of UST during the maintenance phase, positively influenced the economic outcomes.

In recent years, the diagnosis and treatment of CD in China have improved significantly with the introduction of new therapies and updates to treatment paradigms. However, against the growing patient population, the healthcare system still faced multiple challenges: (1) Relatively insufficient resources and training in IBD subspecialties[24]; (2) Weak patient self-management capabilities and low treatment adherence[25]; and (3) Delays in diagnosis and treatment among some patients due to socioeconomic factors[26]. Together, these factors contributed to increased disease relapse rates, hospitalization rates, and surgical risks. Although the NRDL has consistently included several CD biologics since 2020, substantially enhancing drug accessibility and affordability, the overall utilization rate of biologics in China remained relatively low, with notable regional disparities[27]. Low biologics utilization could be influenced by a variety of factors. Compared with small molecule drugs, biologics have higher requirements for storage, transportation, etc., which may increase the costs for hospitals, and reduce patient compliance. In addition, there were several barriers to switching biologics in clinical practice, including clinical inertia or physician preference, the complex dosage adjustment regimens of biologics, administrative barriers to new drug admission, and patient concerns about the unknown side effect spectrum of new drugs, especially biologics with a short time to market, such as VDZ. Therefore, establishing a comprehensive support system, including standardized training for physicians, systematic education for patients and their families, multi-channel social support, is essential[24].

This study has several limitations. Due to the absence of head-to-head trials comparing different treatment regimens, a recently published network meta-analysis of clinical trials in patients with moderate to severe CD was used to estimate efficacy. Partial efficacy data were lacking for IFX and ADA in biologic-exposed populations; therefore, assumptions were made regarding their efficacy as second-line therapies. Given the significant differences inefficacy between biologic-naïve and biologic-exposed patients, the missing subsequent efficacy values were not assumed to be equivalent to those of first-line treatment but were instead set equal to the efficacy of the least effective drug in the corresponding population to avoid overestimating the effectiveness of the treatment. Although the four biologics could be combined into a greater number of treatment sequences, only the most clinically plausible regimens were selected for analysis based on real-world practice. Even though, several sequences still appeared theoretically possible but may not reflect routine clinical decision making. However, considering the complexity and variability of clinical decision-making, providing therapeutic sequences can still help provide are reference for optimizing the clinical pathway from an economic perspective. The study assumed that all treatment-switching behaviors were related to clinical efficacy. However, in the real world, the complexity of patient decision-making – particularly the interplay of social, economic, and personal factors affecting treatment adherence – may introduce additional uncertainty to the outcomes. Future research could further integrate these multidimensional decision-making factors to study the cost-effectiveness of different treatment sequences.

CONCLUSION

Overall, the findings indicate that for Chinese patients with moderate to severe CD, increasing the number of biologic treatment lines could lead to greater health improvements. In the comparison of cost-effectiveness across different treatment sequences, the combination of ADA and UST yielded the highest NMB and the most favorable ICER results, and is likely to be the most cost-effective option in China. However, our findings indicate that further studies are needed to evaluate the long-term efficacy and safety of different biologics in Chinese CD patients, as well as the economic and epidemiological impacts of CD, using more robust data on costs, quality of life, and treatment adherence.

Data sharing statement

All data supporting the findings of this study are included within the article and Supplementary material. No additional data are available.