The Efficacy and Cost-Effectiveness of Umeclidinium/Vilanterol versus Tiotropium in Symptomatic Patients with Chronic Obstructive Pulmonary Disease

Background Both long-acting muscarinic antagonists (LAMAs) and long-acting β2-agonists (LABAs) are widely used in the treatment of chronic obstructive pulmonary disease (COPD). A novel LAMA/LABA combination of umeclidinium/vilanterol (UMEC/VI; 62.5 μg/25 μg) is approved for chronic obstructive pulmonary disease (COPD) treatment. Objective This study aimed to assess the efficacy and cost-effectiveness of UMEC/VI versus tiotropium (TIO) 18 μg in symptomatic patients with COPD from the perspective of the Chinese National Healthcare System. Methods A simple analysis included three studies in the meta-analysis that compared UMEC/VI with TIO. A Markov model was developed to estimate the cost-effectiveness of UMEC/VI compared with TIO treatment in symptomatic patients with COPD. First, utilities, clinical efficacy, and adverse events obtained from the literature were utilized as model inputs. Costs were from Chinese average data, including local data. Costs were expressed in dollars based on 2020 prices. Then, the model outputs including drug costs, other medical costs, and total costs, and quality-adjusted life years (QALYs) were estimated. Costs and outcomes were discounted at a 5% annual rate. Furthermore, incremental cost-effective ratios (ICERs) were analyzed. Finally, the influences of changing parameters on the uncertainty of the results were assessed by means of one-way and probabilistic sensitivity analyses. Results This study revealed that UMEC/VI treatment had a higher rate of clinical efficacy in comparison with TIO, and the differences in the rate of adverse events between the two treatments were not significant. The results indicated that UMEC/VI was superior to TIO, which provided an increase in QALYs (0.002) and a total cost savings of $765.67 per patient over 3 years. In the base case, the ICER of UMEC/VI is -$397468.04/QALY compared with TIO, suggesting that UMEC/VI may be considered a dominant option over TIO. According to the Chinese medical system, the probability of UMEC/VI being cost-effective was 61.6% at a willingness-to-pay (WTP) of $31554/QALY. Sensitivity analyses confirmed that the results were robust. Conclusion UMEC/VI could be considered a cost-effective treatment compared with TIO in symptomatic COPD patients from the Chinese National Healthcare System perspective. These results may help decision-makers in China when making judgements on which treatments to administer.


Introduction
Chronic obstructive pulmonary disease (COPD) is a disabling respiratory disease characterized by persistent and progressive airflow limitation. It is a commonly preventable and treatable disease associated with an increased chronic inflammatory response in the lungs to noxious stimuli [1]. It ranks as the fourth leading cause of death worldwide by the World Health Organization (WHO). In China, Chen Wang's Pulmonary Health study [2] indicated that 8.6% of the general Chinese adult population (or 99.9 million Chinese adults) aged 20 years or older in 2015 had spirometry-defined COPD, reaching epidemic proportions. COPD has been a major public health problem and will remain a challenge for clinicians in the 21 st century. In Asia, high rates of smoking and air pollution ensure that COPD will continue to pose an ever-increasing public health problem [3] with high healthcare costs. In China, the annual mean health care cost of a COPD patient is approximately $3093.759. Hospitalization and medicine accounted for most of the expenses. COPD is now the third leading cause of death in China, and it is reported that only 2.6% of respondents in China are aware that they have the disease [4].
Inhaled medication is the cornerstone of the pharmacological treatment for patients with asthma and COPD [5]. is drug delivery system has the advantage of delivering the drug directly into the airway and reducing the risk of systemic side effects through high local concentrations [6]. Inhalation of long-acting bronchodilators, which improve lung function and reduce symptoms, is the basis of COPD maintenance treatment. LABA and LAMA can be used alone or in combination. Administration of an LABA + LAMA combination is recommended for patients with severe symptoms or those whose symptoms persist despite treatment. In the USA, the EU, and several other countries, maintenance treatment for COPD consists of LAMA (UMEC) in combination with LABA (VI). Treatment with UMEC/VI, which has a clinically acceptable safety profile, was more effective than treatment with TIO monotherapy or placebo, resulting in improved lung function [7][8][9]. UMEC/ VI inhaled powder aerosol (once daily) has been extensively used in maintenance treatment for COPD since it entered the Chinese market in 2018. A novel fixed-dose combination of LAMA (UMEC) and LABA (VI) (62.5 μg/25 μg) can relieve symptoms in adults with COPD (given once daily) as a maintenance bronchodilator treatment. is study provides a new alternative treatment for symptomatic COPD patients. However, its cost-effectiveness in China remains unknown. Considering the economic burden associated with COPD, drug choice should be based on the expected clinical and economic benefits.
An economic evaluation is considered a valuable tool for drug choice decision-making, especially in a comparative analysis of costs and health outcomes. We should consider the relationship between resource inputs (costs) and intermediate outputs in economic evaluations. Our country has a large population of close to 100 million people with COPD. e national talks are listed in China's medical reform policy. e new variety of UMEC/VI is included in the national talks catalog, which means that more people may use this medicine.
Currently, there is no analysis of the cost-effectiveness of UMEC/VI vs. TIO in China. From the perspective of China's healthcare system, this study aims to evaluate the cost-effectiveness of UMEC/VI vs. TIO for COPD for the first time.

Methods
A meta-analysis was carried out to estimate the clinical efficacy and safety of UMEC/VI compared with TIO in symptomatic patients with COPD. A cost-effectiveness analysis was performed comparing UMEC/VI with TIO in symptomatic patients with COPD from the perspective of China's healthcare system using data from the literature and clinical studies of UMEC/VI vs. TIO.

Search Strategy and Study
Selection. Searches of Medline, Cochrane Library, ClinicalTrails.gov, China National Knowledge Infrastructure (CNKI) were performed for "Umeclidinium" and "Vilanterol" to collect literature on all randomized clinical trials in humans reported up to 31 January 2020. For each paper selected, a systematic manual search of the bibliographies was carried out. e selected language was limited to English or Chinese. e criteria for inclusion in the meta-analysis were as follows: (1) study type: prospective cohort studies, randomized controlled trials, and randomized crossover studies; (2) study subjects: patients who were diagnosed with COPD; and (3) interventions: UMEC/VI (62.5 μg/25 μg) treatment for 24 weeks was compared with TIO (18 μg). e exclusion criteria in the meta-analysis were inconsistent research topics, drug doses, control drugs, experimental time, and research periods. e process is shown in Supplementary Figure 1.

Data Extraction and Quality
Evaluation. Data were independently abstracted from each trial by two researchers; disagreement was resolved by consensus. e following data were extracted: author and year of publication; the number of clinical trials; study duration; study type; total number of patients; number of intervention groups; number of control groups; average age; sex ratio; GOLD stage classification; current smoker at screening (%); smoking pack-years; trough FEV 1 on day 169, L; number of patients with on-treatment exacerbation; any on-treatment AEs, n (%); on-treatment SAEs; AE reported by 3% of patients in any treatment group, n (%). e quality of the research was assessed according to the Cochrane Collaboration's risk of bias assessment tool. is tool includes several sources of bias, for example, random sequence generation, allocation hiding, blinding of subjects and intervention providers, blinding of result evaluators, incomplete results data, selective results reporting, and other sources of bias.

Data Analysis.
e relative risk (RR) and 95% confidence interval (CIs) between interventions were obtained via Review Manager 5.3 software. e random-effects model was used if there was moderate or high heterogeneity between studies (P value < 0.1 or I 2 value > 50%). ere was statistical significance when the P value was less than 0.05. According to the three severity levels of COPD defined in the COPD clinical guidelines in 2013, the severity of COPD is classified according to the patient's forced expiratory volume in one second (FEV 1 ).

Cost-Effectiveness
ere were three disease states ( Figure 1): 50% ≤ predicted FEV1 < 80% (moderate COPD), 30% ≤ predicted FEV1 < 50% (severe COPD), predicted FEV1 < 30% (very severe COPD), and death. e cycle length of the Markov model was set to three months [10], as the findings from the NHANES III follow-up study for COPD recommended. e time horizon of the model is three years. After entering the model, patients need to receive maintenance COPD treatment plus routine care. In the model, the patients maintained their current health state of disease severity or moved to the next more serious health state within three months. According to the natural course of the disease, any health condition can lead to death ( Figure 1). Cost-effectiveness was then explored through the calculation of incremental cost-effectiveness ratios (ICERs), defined as the difference in mean costs divided by the difference in mean effects. In light of the recommendation by the China Guidelines for Pharmacoeconomic Evaluations 2020 [11], the willingness-to-pay (WTP) for a quality-adjusted life year (QALY) in this study was set to be three times the gross domestic product (GDP) per capita. e average exchange rate of Chinese yuan renminbi for 2020 was 6.899 yuan per US dollar, and the Chinese GDP per capita was $10518 in 2020, according to data from the National Bureau of Statistics(https://www.stats.gov.cn/tjsj/sjjd/202101/t20210 119_1812636.html). e WTP threshold was $31554.

Study Perspective.
Cost-effective analyses were conducted from the perspective of China's healthcare system, and only direct medical costs were included.

Model Inputs
(1) Target Population. Patient characteristics for the basecase analysis were based on three RCTs (NCT01777334, NCT01316913, and NCT01316900) were included in the meta-analysis (Table 1). Eligible patients were administered inhalation therapy of UMEC/VI (62.5 μg/25 μg) for 24 weeks compared with TIO (18 μg) once a day. ere was a little difference among the three studies. e method of merging is to merge and calculate according to the proportion of the number of people in each RCT.
(2) Probabilities. Each model state comes from the reference, the first cycle uses the initial probability from base to new, and all subsequent cycles use the new probability. e transitions between the patient's disease severity and health status were based on the method described by Spencer et al., which uses the average patient time in each health status [12].
e transition probabilities of one cycle (3 months) were calculated from the formula: r � − [ln(1 − P1)] /t1, P2 � 1 − exp(− rt2); r represents the transition rate, and P 1 and P 2 represent the transition probabilities for a given cycle length of t 1 and t 2 , respectively.
(3) Efficacy Parameters. (Change in Trough FEV. 1 ) Treatment effects measured by the FEV 1 for UMEC/VI compared with TIO were obtained from three clinical studies (NCT01777334NCT01316913, and NCT01316900). Trough FEV 1 at day 169 was the primary endpoint [9]. Meta-analysis was performed after extracting trough FEV 1 at day 169.
(4) Exacerbations. Exacerbations in this model were the events that caused patients to seek healthcare. Patients could experience different levels of exacerbations or no exacerbations. Patients were considered event free if they experienced no exacerbations. Patients who experienced nonsevere exacerbations sometimes required a change in treatment, such as antibiotics and/or systemic corticosteroids, and/or contact with doctors. Severe exacerbations required hospitalization. Exacerbation risk was based on patient COPD severity and was obtained from the ECLIPSE study [13]. e number of exacerbations every three months is included in Table 2. Different exacerbation costs were considered in the model.

(5) Adverse Events (AEs).
Adverse events were considered in the model. Adverse events occurred in at least 3% of patients and caused significant costs. Adverse events included nasopharyngitis, upper respiratory tract infection, cough, oropharyngeal pain, back pain, and arthralgia, and the corresponding treatment came from expert consultation ( Table 3).  (Table 4). e monthly prescription cost is estimated based on the recommended dose for each treatment. All drug costs are from http://www.yaozh.com, calculated by the median price of the drug in 2020. e corresponding treatment of adverse reactions (usage, dosage, and number of days) is based on expert consultation, and the corresponding treatment cost comes from a tertiary hospital.
(7) Utility Weights. e annual utility weight used in the model comes from the literature [12] described by Spencer et al. Estimates of health status by disease stage were generated from the Health Survey for England. is included an assessment of both lung function and health status, measured using the Euro QOL questionnaire (EQ-5D), in 283 patients with COPD.

Cost-Effectiveness of UMEC/VI (62.5 μg/25 μg) Compared with TIO (18 μg).
e disease progression of COPD patients with different therapies was simulated by relying on the model method.
e model parameters are shown in Table 2. e cost-effectiveness analysis was used to evaluate the influence of different therapies on patients' long-term quality of life and disease burden. An annual discount rate of 5% was adopted for both effects and costs.

Base-Case Analysis.
e cycle length of the Markov model was set to three months [10]. Based on clinical trials and expert opinions, assuming a cycle of 12 weeks (3 months), most patients will change the treatment after three Canadian Respiratory Journal years, so the time horizon of the model is set to three years [15]. e model can also simulate a time span of 1-25 years as needed. Cost and outcomes were discounted at a 5% annual rate in line with Chinese guidelines for economic evaluation [16]. e model estimated costs (total, drug, nondrug, discounted), QALYs gained and incremental costeffectiveness per QALY gained.

One-Way Sensitivity Analysis.
e effect of changing parameters in the model was examined in one-way sensitivity analyses to test the robustness of the model assumptions and specific parameters.
In the one-way sensitivity analysis, the 95% CI of the parameter is preferred as the upper and lower fluctuation range of the parameter [17]. If the 95% CI cannot be obtained, the parameter fluctuation ±20% was used as the fluctuation range.
Data from sources all use 95% CI (such as ADR incidence, utility value), and ±20% (various costs, number of aggravations) are used for sources that cannot be found. e results of the sensitivity analysis for each input were ranked from the most sensitive to the least sensitive and summarized as a tornado diagram, with the 10 most sensitive parameters presented.

Probabilistic Sensitivity Analyses.
In the probabilistic sensitivity analysis, the second-order Monte Carlo simulation was used to iterate the model 1,000 times to examine the changes in the results of the base-case analysis when all parameters change in their respective distributions. Based on the results of the second-order Monte Carlo simulation, we constructed cost-effectiveness acceptability curves (CEACs). ree clinical studies were ultimately included in the meta-analysis after using our search strategy and selection criteria. Trough FEV 1 at day 169 was the primary endpoint. e selection flow is summarized in Supplementary Figure 1. e basic characteristics of the included RCTs are summarized in Supplementary  Table 1. All the included studies were of good methodological quality. e meta-analysis results ( Figure 2) proved that UMEC/ VI had higher FEV 1 results than TIO (RR: 5.96, 95% CI: 2.41-9.51). Based on the results of the meta-analysis, UMEC/ VI can significantly increase the FEV 1 level of COPD patients compared with TIO.

Number of Patients with On-Treatment Exacerbation.
e analysis (Figure 3) also implied that there was no remarkable difference in the rate of exacerbation between UMEC/VI and TIO treatment (RR: 1.58, 95% CI: 0.98-2.55).

Adverse Events and Severe Adverse Event Results
(1) Adverse Event Results. e analysis (Figure 4) also implied that there was no remarkable difference in the rate of adverse events between UMEC/VI and TIO treatment (RR: 1.08, 95% CI: 0.98-1.19).

Moderate COPD 50%≤predicted FEV1< 80%
Severe COPD 30%≤predicted FEV1< 50% Very severe COPD Predicted FEV1< 30% Death Figure 1: Structure of the decision model used.  No significant difference between UMEC/VI and TIO was observed based on the adverse event results of the meta-analysis.
(2) Severe Adverse Event Results. According to the severe adverse event results of the meta-analysis (Figure 5), the difference between UMEC/VI and TIO (RR: 1.08, 95% CI 0.48-2.44) was not statistically significant.
In brief, it could be concluded that UMEC/VI treatment, which did not increase side effects, was more effective than TIO for COPD. e meta-analysis indicated that not only did UMEC/ VI have more significant effects than TIO on improving FEV 1 (RR: 5.96, 95% CI: 2.41-9.51), but it also did not increase any adverse event results (RR: 1.08, 95% CI: 0.98-1.19) or any severe adverse event results (RR: 1.08, 95% CI 0.48-2.44).
e results of the meta-analysis demonstrated that there was no significant difference between UMEC/VI and TIO in exacerbation rate, incidence of any adverse events or serious adverse events.  In the base case, the ICER of UMEC/VI suggested that UMEC/VI may be considered a superior option to TIO.

One-Way Sensitivity Analysis.
e effect of changing parameters was examined in one-way sensitivity analyses to assess the model robustness and the uncertainty of the input parameters. e results of the sensitivity analysis for each input were ranked from the most sensitive to the least sensitive and plotted on a tornado diagram.  e utility of predicted FEV 1 (50%-80%) was identified to have the greatest influence on the results, followed by the utility of predicted FEV 1 (30%-50%) and the cost of TIO from the results of the one-way sensitivity analysis ( Figure 6).

Probabilistic Sensitivity Analysis.
Probabilistic sensitivity analyses (second-order Monte Carlo simulation), in which all parameters in the model varied at the same time, were conducted to address the uncertainty in the model input values.
e ICERs for the 1000 samples in the probabilistic sensitivity analysis (PSA) are shown in the scatter plot (Figure 7). e acceptability curves (Figure 8) further indicated that the cost-effectiveness of UMEC/VI treatment decreased with increasing WTP thresholds, while TIO treatment increased with increasing WTP thresholds.

Discussion
Not only criteria of efficacy and safety but also cost-effectiveness influences decisions in today's health care environment. Cost-effectiveness analyses can provide information about the economic value of health interventions, often compared to the most commonly used intervention. Cost-effectiveness analyses are tools that can help health managers and decision-makers make informed decisions. e aim of this study was to evaluate the cost-effectiveness of UMEC/VI vs. TIO, the most widely used treatment for COPD. Costs included drug and nondrug expenses, costs related to exacerbation events, and adverse event costs in particular. Including quantity and quality of life and measuring the changes in utility during the patient's life, the QALY is the most acceptable health-related utility measure. As a result of different cumulative numbers of exacerbations over the horizon analyzed, quality of life, percentage of patients with symptoms (dyspnea and cough and sputum), FEV 1 results and survival, QALY differed between the compared treatments in the model developed.
e analysis originally studied the cost-effectiveness of the combination of two bronchodilators in comparison with monotherapy in China. A similar analysis has been carried out in the UK and Spain. In a UK study, UMEC/VI was considered to be a cost-effective alternative to TIO at a certain price [18]. In a study in Spain, UMEC/VI produced an additional 0.03 QALY and €590 vs. TIO, causing an ICER of €21,475/QALY. According to PSA, the probability of UMEC/VI being cost-effective was 80.3% at a WTP of €30,000/QALY [19]. A cost-effectiveness analysis of    [20]. ese results are in agreement with previous studies of UMEC/VI and with systematic reviews and meta-analyses of other therapies, which revealed that FEV 1 benefit is related to improvement in quality of life. e combination of inhaled UMEC with VI has been shown to provide significant improvements in lung function compared with UMEC, VI, or placebo in patients with COPD [21]. Model selection is an important factor in cost-effectiveness analyses, since the model must represent the disease studied. is analysis utilized a published Markov model of disease progression [22]. e model used in COPD is based Tornodo Analysis (ICER) $-3,000,000 $-2,500,000 $-2,000,000 $-1,500,000 $-1,000,000 $-500,000 Utility of predicted FEV1 (50%-80%)    According to the Chinese healthcare system, the probability of UMEC/VI being cost-effective was 61.6% at a willingness-to-pay of $31554/ QALY. on pulmonary function for the transition between different health states. Upon entering the model, patients were prescribed maintenance COPD treatment plus usual care. In the model, patients remained in their current disease severity state or moved to the next more severe state each year. Patients could also experience an exacerbation or remain event free. Death could occur from any health state according to the natural progression of the disease. e use of LAMA + LABA combination treatment is recommended in patients with more severe symptoms or those whose symptoms or obvious limitations persist despite receiving monotherapy [23]. e results obtained here support this recommendation, with UMEC/VI being an efficient treatment option compared with TIO monotherapy. In the basecase results, the patients gained more QALYs from UMEC/ VI than from TIO. e cost-effectiveness analysis presented here could be a conservative estimate since some medical costs, such as costs concerning inhaler misuse, were not factored into the model. Poor inhalation techniques can lead to poor disease control, and the costs related to critical errors are considerable [24,25]. In a recent study, the Ellipta device was compared with other commonly used inhaler devices. Patients who used the Ellipta device had fewer critical errors than those who used the five alternative inhalers [26].
Confirming the results observed within this indirect treatment comparison and comparing the cost-effectiveness of LAMA/LABA vs. LAMA therapies over a longer duration are planned for future studies. Hence, the informed judgements of payers on cost-effectiveness will be needed. Certain assumptions were made in this model about the use of LAMA/LABA combination therapy for a longer duration because data about the timeframes considered in this analysis (3 years, 5 years, 10 years, and lifetime) are not available from current clinical studies. e assumption that there are changes in FEV 1 once a patient escalates to subsequent triple therapy (LAMA/LABA + ICS) needs to be made due to the lack of data on escalation to triple therapy from LAMA/LABA therapy. It could be a conservative assumption that patients treated with UMEC/VI did not experience a change in FEV 1 after initiating ICS therapy. e strength of this study is that it originally studied the cost-effectiveness of the combination of two bronchodilators in comparison with monotherapy in China. In this study, we included the following data: current smoker at screening (%); smoking pack-years; trough FEV 1 on day 169, L; number of patients with on-treatment exacerbation; any on-treatment AEs, n (%); and ontreatment SAEs. Finally, data on baseline SGRQ and 6MWT distances were not available from the sources used for this analysis and had to be estimated within the model. To further facilitate cost-effectiveness evaluations, collecting and including such data in future studies are necessary.
COPD is a chronic progressive condition, and patients with COPD frequently undergo inhalation treatment switches. In the current analysis, we did not consider any inhalation treatment changes. e limitation of these economic evaluations is the quality of the data. In this case, local Chinese data sources were considered. e cost-effectiveness analysis presented here could be a conservative estimate since some medical costs, such as costs concerning inhaler misuse, were not factored in the model. All assumptions were validated by the researchers, and a sensitivity analysis was performed to evaluate uncertainty. Although Chinese costs were used in these sensitivity analyses, we nevertheless provide the most applicable platform by which to assess the cost-effectiveness of UMEC/VI vs. TIO from the perspective of other nations using nationally derived cost inputs.

Conclusion
e results of this cost-effectiveness analysis show that treatment with UMEC/VI in symptomatic COPD patients is a cost-effective option compared with TIO from the perspective of the Chinese NHS, as the ICER was below the threshold commonly accepted in China to consider interventions as efficient.