Article Text
PDF
Abstract
Objectives This analysis aims to better reflect the value of new antibiotic treatment strategies, thereby informing clinical antibiotic use, antimicrobial reimbursement and/or hospital formulary decision-making in China.
Design We adapted a published and validated dynamic disease transmission and cost-effectiveness model to evaluate the clinical and economic outcomes of introducing a new antibiotic, ceftazidime/avibactam (CAZ-AVI) for treating resistant infections in Zhejiang province, China. Outcomes were assessed over a 10-year infectious period and an annual discount rate of 5%. Costs were extracted from the hospital’s Health Information System (HIS) and obtained after data cleaning, aggregation and discounting.
Setting The Chinese healthcare system perspective.
Participants 10 905 patients in a Chinese tier-3 hospital from 2018 to 2021 with any of the three common infections (complicated intra-abdominal infection (cIAI), hospital-acquired/ventilator-associated pneumonia (HAP/VAP) and infections with limited treatment options (LTO)) caused by three common resistant pathogens (Escherichia coli, Klebsiella spp. and Pseudomonas aeruginosa).
Interventions (1) Current treatment strategy (piperacillin-tazobactam (pip/taz) and meropenem); (2) CAZ-AVI at the third line; (3) CAZ-AVI at the second line; (4) CAZ-AVI at the first line; (5) CAZ/AVI first line, two lines diversified (i.e., equal pip/taz and CAZ-AVI at the first line; meropenem at the last line) and (6) CAZ/AVI first line, all-lines diversified.
Primary outcome measures Quality-adjusted life years (QALYs) lost, hospitalisation costs and incremental net monetary benefit (INMB) were used to assess cost-effectiveness.
Results Over 10 years, the introduction of CAZ-AVI to the current treatment strategy led to lower hospitalisation costs and more QALYs across all five treatment strategies, with between 68 284 and 78 571 QALYs gained whilst saving up to US$236.37 for each additional QALY gained. The INMB of introducing CAZ-AVI is estimated up to US$3 550 811 878.
Conclusions Introducing CAZ-AVI had a positive impact on clinical and economic outcomes for treating antimicrobial resistance, and diversifying the antibiotics use early in the treatment might yield the best benefits.
- anti-bacterial agents
- health economics
- public health
Data availability statement
Data are available upon reasonable request. The data sets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
In this study, the dynamic disease transmission and cost-effectiveness model we adapted has been validated and is well-suited for economic evaluations of infectious diseases.
This model takes into account the transmission and diversity value attributes from the broader value framework for antimicrobials, STEDI (spectrum, transmission, enablement, diversity and insurance).
Most of the parameters in the model are derived from real-world data collected from the hospital in Zhejiang province, China, making it highly applicable to the hospital settings in Zhejiang province, China.
Due to the absence of the China-specific utility and treatment efficacy data for piperacillin-tazobactam and meropenem, data from the literature was used which may not be most reflective of local situations.
Length of stay and daily hospitalisation costs data were collected from a single hospital with limited generalisability.
Introduction
The increase in antimicrobial resistance (AMR), accelerated by inappropriate use of antimicrobials, is an important global public health concern, and has emerged as a severe public health problem in China, with recent reports indicating alarming levels of resistance among common pathogens.1 2 For example, the resistance rate of clinical Escherichia coli (E. coli) isolates to ciprofloxacin and levofloxacin in 2022 has been reported as 61.4% and 53.2%, respectively.3 Gram-negative bacteria are particularly troublesome because some of them have become resistant to nearly all the available antibiotics. According to a report by the China Antimicrobial Surveillance Network in 2022, 65.7%–73.0% of the isolates were gram-negative bacteria; the most frequently isolated gram-negative pathogens were E. coli, Klebsiella spp., Pseudomonas aeruginosa (P. aeruginosa) and Acinetobacter baumannii (A. baumannii).3 The overuse of antimicrobials in clinical practice, animal feed and agricultural settings further exacerbates the situation.4 During the Coronavirus Disease 2019 (COVID-19) pandemic, the World Health Organization (WHO) warned about the potential for increased misuse of antibiotics and relaxed antimicrobial stewardship practices, which could exacerbate AMR development.5
The Chinese government and health authorities have taken measures to combat AMR. National surveillance networks for bacterial resistance and the clinical use of antimicrobials were established in 2005, with annual reports being published.6 Since 2011, the National Health Commission has issued several announcements focused on the appropriate use of antimicrobials and antimicrobial stewardship (AMS).7 Additionally, in 2016, 14 ministries, led by the National Health and Family Planning Commission (now the National Health Commission), jointly released the National Action Plan to Contain Antibacterial Resistance (2016–2020).8 Following the launch of these policies studies indicate a general decrease in resistance rates among most tested species from 2018 to 2022 suggesting these policies have been effective in tackling AMR in China.9 Despite the implementation of policies aimed at addressing antibiotic misuse, challenges such as inadequate surveillance systems, insufficient healthcare infrastructure and lack of standardised reporting methods persist. Moreover, public awareness about AMR remains low, necessitating comprehensive educational campaigns and improved training for healthcare professionals.
The development of AMR has been compounded by slowing research and development (R&D) of new antimicrobials, with many of the largest global pharmaceutical companies withdrawing from the market, leading to a clinical pipeline that is inadequate for the current AMR threat.10 Antimicrobials are associated with a low return on investment, compared with other therapeutic areas such as oncology. The investment and time required to develop a new antimicrobial are as high as other drugs; however, the ability to generate revenues is limited by the availability of cheap competitor treatments, AMS restricting the use of antimicrobials with new products often held in reserve and a loss of efficacy due to rapid resistance development.11 This is evident in new drug approvals in China in the recent decade: from 2011 to 2021 in China, most approved were antineoplastic agents12; data from the National Medical Products Administration in 2022 revealed that nearly half of all newly approved drugs were antineoplastic agents but no antibiotics.13 Access to new antibiotics is often restricted by the strict management of antimicrobial entities in hospital formularies. In 2012, the Ministry of Health of the People’s Republic of China (now the National Health Commission of the People’s Republic of China) issued ‘Measures for the Management of Clinical Application of Antibacterial Drugs’, which posed a limit of 50 and 35 antimicrobial entities in tier-3 and tier-2 hospitals, respectively.7 This measure was proposed to prevent inappropriate use of antibiotics; however, it may serve as a barrier to more clinically appropriate recently approved antibiotics. Additionally, access to the national drug reimbursement list has been increasingly difficult for new antibiotics mainly due to high expectations of clinical superiority and minimal impact on the healthcare budget.14 However, for many life-saving antibiotics, non-inferiority trial designs are used due to ethical considerations.
Antibiotics are subject to ‘externalities’, for example, the transmission of infection and the emergence of resistance, that are not considered in conventional economic modelling and health technology assessment (HTA).15 Antibiotics have a broader influence beyond the individual patient who receives them, quantifying these externalities is essential to better demonstrate their full value to the population and healthcare system. The value of antibiotics to the population has been represented in the literature with the acronym STEDI (spectrum, transmission, enablement, diversity and insurance).16
Introducing a new antibiotic and increasing the available treatment options can slow the overall increase of AMR due to reduced selection pressure.17 This practice aligns with AMS guidance and is encouraged in clinical practice. This value attribute is referred to as ‘diversity value’ in the STEDI framework. In addition, new effective antibiotics can help reduce the spread of infections and resistance, referred to as ‘transmission value’. Economic analyses that appropriately quantify the broader value of new antibiotics to populations, and their use in HTA and reimbursement decision-making, are critical for stimulating R&D investment and promoting access to new antibiotics.15 18
A dynamic disease transmission and cost-effectiveness model of antibiotic resistance has been developed to estimate the value of new antibiotics considering transmission and diversity value.19 20 However, in China, to the best of our knowledge, there have been no economic evaluations that quantify the value of new antibiotics using these approaches.
We aim to better understand the clinical and economic value of different treatment strategies before and after introducing a new antibiotic from the Chinese healthcare system perspective. Our findings may contribute to bringing a new aspect of antibiotic value proposition in China and provide valuable insights for policymakers informing decisions to combat AMR and ensure appropriate access to new antibiotics.
Methods
Structure of the model
A previously published and validated dynamic disease transmission and cost-effectiveness model of AMR20 was adapted to estimate the impact of introducing a new antibiotic to the healthcare system in Zhejiang province, China. The model considered outcomes over a 10-year transmission period, where quality-adjusted life years (QALYs) and life years (LYs) were assessed over the patient’s lifetime.
A population-based mathematical model was developed using linear regression equations to capture the transmission dynamics of resistance from a previously developed and validated dynamic disease transmission and cost-effectiveness model of AMR (figure 1).20 Over 1 million permutations of model inputs (population, baseline resistance, treatment strategy, treatment duration and treatment efficacy) were run to produce a data set of inputs and outputs. Linear regression equations were derived from models fitted to the outcome data for time-on-treatment, mortality and resistance, to create a parsimonious and tractable framework to estimate the value of a new antibiotic. Inputs sourced from the literature reflective of the Chinese setting were applied to the regression equations to estimate outcomes. LYs and QALYs were estimated using the mortality equation, applying inputs for utility (infected and non-infected patients) and life expectancy. Cost outcomes were linked to the mortality and time on treatment equations where daily hospitalisation costs were applied. Resistance was captured in the regression equations as a factor of treatment exposure, which is influenced by model settings and inputs relating to treatment sequence, antimicrobial stewardship practices (diversification) and treatment duration, which is also influenced by baseline resistance and treatment efficacy.
Model schematic of the published dynamic disease transmission model, characterised in the model regression equations. (A) Flow diagram of disease transmission and (B) treatment pathway.
Infected patients (resistant or sensitive) were treated according to a defined treatment pathway allowing for health economic analyses of different treatment strategies. Patients were able to receive up to three lines of antibiotic treatment. Patients could be cured by successful treatment or natural resolution of infection or were unsuccessfully treated. Patients unsuccessfully treated were able to start a subsequent treatment line. Patients had a daily risk of mortality associated with the infection; patients unsuccessfully treated after exhausting all treatment options were assumed to die.
The model considered treating infected patients (resistant or sensitive) in a decision-tree treatment pathway allowing for health economic analyses of different treatment strategies. Patients were able to receive up to three lines of antibiotic treatment. Patients could be cured by successful treatment or natural resolution of infection or were unsuccessfully treated. Patients unsuccessfully treated were able to start a subsequent treatment line. Patients had a daily risk of mortality associated with the infection; patients unsuccessfully treated after exhausting all treatment options were assumed to die.
Patient population and setting
The patient population selection was based on the following inclusion criteria: (1) inpatients, (2) male or female, aged 18 years old or above, (3) infected with complicated intra-abdominal infections (cIAI) or hospital-acquired pneumonia and ventilator-associated pneumonia (HAP/VAP) or infections with limited treatment options (LTO), caused by any of the three pathogens (E. coli, Klebsiella spp. and P. aeruginosa) and (4) used antibiotics during hospitalisation. Patients who met these inclusion criteria but with missing data in drug sensitivity were excluded.
Based on these inclusion criteria, 10 905 patients were identified from a tier-3 hospital in Hangzhou city, Zhejiang province, China, from 2018 to 2021 (online supplemental table A.1), data from these patients was used to support parameterising the model.
Supplemental material
Patient and public involvement
No patient involved.
Interventions and comparators
Ceftazidime-avibactam (CAZ-AVI) was assessed as the new antibiotic and was evaluated in the context of being added to the current treatment strategy. CAZ-AVI is an intravenously administered fixed-dose combination of the third-generation cephalosporin ceftazidime and the novel, non-β-lactam β-lactamase inhibitor avibactam. In China, CAZ-AVI was approved in 2019 for treating adults with cIAI, HAP/VAP and other infections caused by aerobic gram-negative organisms in patients with LTO.21
Piperacillin-tazobactam (pip/taz) and meropenem were selected to best represent the current treatment strategy for the infections of interest, in China. Other types of existing antibiotics, such as colistin, were not included in analyses due to insufficient case data.
Six treatment strategies were defined to reflect the clinical practice without CAZ-AVI compared with CAZ-AVI used in various treatment scenarios. These six treatment strategies were: (1) current treatment strategy (i.e., without CAZ-AVI), (2) CAZ-AVI at the third line, (3) CAZ-AVI at the second line, (4) CAZ-AVI at the first line, (5) CAZ-AVI first line, two lines diversified and (6) CAZ-AVI first line, all-lines diversified (figure 2).
Treatment pathways of six modelled treatment scenarios (1—current treatment strategy; 2—CAZ-AVI at the third line; 3—CAZ-AVI at the second line; 4—CAZ-AVI at the first line; 5—CAZ/AVI first line, two lines diversified; 6—CAZ/AVI first line, all-lines diversified). 1st, first; 2nd, second; 3rd, third; Pip/Taz, piperacillin/tazobactam; CAZ-AVI, ceftazidime/avibactam.
The current empirical treatment strategy included pip/taz as the first-line treatment with meropenem as the second-line treatment (figure 2(1)). This strategy was considered as the control for other treatment strategies (2–6) with CAZ-AVI to be compared against.
The first-line diversification and all-line diversification strategies were both designed to explore the impact of AMS strategies that aim to reduce the exposure to first-line treatments. In first-line diversification, patients were split equally (50%) to receive either CAZ-AVI or pip/taz at the first line; patients received the other antibiotic (i.e., pip/taz or CAZ-AVI) at the second line, with meropenem at the last line (figure 2(5)). In comparison, in all-line diversification, patients were split equally (33%) to receive one of the three antibiotics for the first-line treatment, following unsuccessful treatment, patients were treated with the next available treatment in the empirical treatment sequence, regardless of which treatment they started with, until all available treatments were exhausted (figure 2(6)).
Input parameters for deterministic analysis
In line with the China Guidelines for Pharmacoeconomic Evaluations,22 an annual discount rate of 5% was applied to costs and outcomes and a willingness-to-pay (WTP) threshold equal to three times the Gross Domestic Product (GDP) per capita in Zhejiang province, China. To align with the model setting and time horizon, the GDP per capita of Zhejiang province was calculated as a 4-year mean from 2018 to 2021,23 resulting in ¥313 087.99 (US$44 959.44, (¥1=US$0.1436 (9 December 2022)). WTP threshold was converted to baseline value using the Consumer Price Index (CPI) with 2021 as the base year.24
Other key inputs populating the model came from hospital data or literature (table 1, online supplemental table B.1). Where possible regional data was used to reflect the conditions in China most accurately. Life expectancy post treatment success, the hospital length of stay (LOS), additional LOS for mortality and daily hospitalisation costs were calculated using data from the 10 905 patients (2018–2021) treated at the tier-3 hospital in Hangzhou city, Zhejiang province, China. The average life expectancy of the population of Zhejiang province (77.73 years)25 minus the average age of successfully treated inpatients (n=10 905; 68.13 years) yielded the life expectancy post treatment success of 9.60 years. LOS was calculated as a weighted average from inpatient prognosis data, for successfully treated (patients who were cured or improved) and unsuccessfully treated (patients who remained infected or died) patients at 10.91 days and 10.40 days, respectively. Additional LOS for mortality of 4.37 days was derived based on the difference in LOS for the discharge outcome of death (27.54 days) and non-death (23.17 days). Daily hospitalisation costs of a patient, including laboratory costs, hospital costs and medication costs, were converted to the baseline value using the CPI and calculated as ¥3062.73 (US$439.81).24
Key inputs used to populate the model
Due to the lack of local data on utility, we referred to the ‘utility-not infected’ of 0.78 from the University of York Centre for Health Economics26 and the ‘utility-infected’ of 0.62 weighted by literature data.27–29 Similarly, treatment efficacy data were calculated as a weighted average based on literature, as 0.83 and 0.87 for pip/taz and meropenem, respectively.30–34 The treatment efficacy of CAZ-AVI was assumed to be 0.90; there is limited clinical trial evidence of CAZ/AVI in Chinese populations, this assumption is based on existing evidence showing CAZ/AVI is more effective than pip/taz or meropenem.
The distribution of E. coli, Klebsiella spp. and P. aeruginosa in 10 905 infected patients was 33.94%, 33.54% and 32.53%, respectively. Pathogen-specific baseline resistance levels for pip/taz and meropenem were also calculated from the hospital data. The resistance level of CAZ-AVI was assumed to be 0.00%, globally resistance to CAZ/AVI is low and the analysis assumes that before its launch in China there was no natural resistance (table 2, online supplemental table B.2).
Baseline resistance specific inputs
Cost-effectiveness analysis
The cost-effectiveness analysis assessed both clinical and economic outcomes of six treatment strategies in terms of absolute and incremental values. The absolute outcomes based on current and alternative treatment strategies included costs: hospital LOS, hospitalisation costs; and health outcomes: LYs lost, QALYs lost due to infection. Incremental outcomes included incremental net monetary benefit (INMB). The resistance levels for the three antibiotics in each of the six different treatment strategies were projected up to 10 years.
One-way sensitivity analysis
One-way sensitivity analysis was conducted on the discount rate, life expectancy post treatment success, utility (for not infected and infected), LOS (for successful treatment and unsuccessful treatment), additional LOS for mortality, daily hospitalisation cost and treatment efficacy (for pip/taz, meropenem and CAZ-AVI). The selection of these variables was mainly based on literature review and expert opinion, which demonstrated a significant impact on the results in previous studies. All these inputs were adjusted by ±25%, except for the discount rate, which ranged from 0% to 8%. The impact on INMB was assessed between CAZ-AVI first line, all-lines diversified and current treatment strategy.
Results
Resistance level
The projected resistance level of the two antibiotics in the current treatment strategy, pip/taz and meropenem, increased by 41.6% and 35.4% (from 24.89% to 35.26% and 28.65% to 38.80%) respectively, over 10 years (table 3). Among the five treatment strategies, modelling the introduction of CAZ-AVI, the projected resistance to pip/taz and meropenem after 10 years was reduced by 1.30%–4.40% (an absolute reduction of 0.46%–1.55%) and 0.77%–3.51% (an absolute reduction of 0.30%–1.36%). The lowest gain in resistance to pip/taz and meropenem were projected when CAZ-AVI was introduced at the first line. Conversely, resistance was highest for pip/taz and meropenem when CAZ-AVI was used as the third-line treatment, as the exposure to pip/taz and meropenem would have been highest (table 3; figure 3A,B). Development of resistance in the new antibiotic (i.e., CAZ-AVI), after 10 years, was highest when used as first-line treatment and lowest when reserved as last line; projected resistance was not impacted by the diversification strategies with all strategies at 2.91%. Overall resistance rates (i.e., the sum of the changes in the resistance rates of the three antibiotics) were lowest after 10 years when CAZ-AVI was used at the second line and the first line with all-lines diversified; in the other strategies, there was no difference compared to the current treatment strategy (table 3; figure 3C).
Resistance level over 10 years
Projected resistance to (A) pip/taz, (B) meropenem and (C) CAZ-AVI over 10 years for diversified treatment strategies. Baseline resistance values of the two existing antibiotics are calculated from the data collected from the tier-3 hospital, which could be representative of the current population resistance level in China to some extent. The baseline resistance value of CAZ-AVI is assumed to be 0.00%. The year 1 is the first year in the dynamic health economic model. 1—current treatment strategy; 2—CAZ-AVI at the third line; 3—CAZ-AVI at the second line; 4—CAZ-AVI at the first line; 5—CAZ-AVI first line, two lines diversified; 6—CAZ-AVI first line, all-lines diversified. CAZ/AVI, ceftazidime/avibactam; Pip/Taz: piperacillin-tazobactam.
Costs
Under the current treatment strategy, the hospital LOS and hospitalisation costs over 10 years were estimated to be 1 588 763 days and ¥3 898 198 802 (US$559 781 348). The introduction of CAZ-AVI at first or second line resulted in a reduction in hospital LOS (1 539 156–1 578 887) and hospitalisation costs (¥3 768 864 867–3 865 333 768 (US$541 208 969–555 061 929)). However, the use of CAZ-AVI at the third line had the longest hospital LOS of 1 606 829 days and the highest hospitalisation costs of ¥3 934 038 268 (US$564 927 895) (table 4).
Absolute and incremental lifetime outcomes over a 10-year transmission time horizon
Health outcomes
Under the current treatment strategy, it was estimated that, over 10 years, 183 175 LYs and 142 999 QALYs would be lost. Compared with the current treatment strategy, the introduction of CAZ-AVI in the other five treatment strategies all reduced the LYs and QALYs lost; the biggest reduction was demonstrated when CAZ-AVI was used at the first line with all-lines diversified, saving 100 727 LYs and 78 571 QALYs (table 4).
Incremental outcomes of cost-effectiveness analysis
In comparison to the current treatment strategy, the incremental outcomes of the other five treatment strategies were superior, with CAZ/AVI first line, all-lines diversified strategy showing the largest benefits. The estimated INMB for this treatment strategy amounted to ¥24 727 102 215 (US$3 550 811 878) (table 4).
Parameters influencing the incremental outcomes
The impact of adjusting inputs by ±25% (discount rate: 0%–8%) was displayed in a tornado plot (figure 4). The one-way sensitivity analysis demonstrated that the model was most sensitive to discount rate for the estimates of INMB. Furthermore, life expectancy post treatment success and utility of patients not infected were key drivers of INMB. Factors such as additional LOS for mortality, the treatment efficacy of pip/taz and the utility of patients infected had minimal impact on INMB (figure 4).
One-way sensitivity analysis varying key inputs by ±25% (discount rate: 0%–8%) was conducted on INMB, comparing CAV/AVI first line, all-lines diversified with the current treatment strategy over 10 years. CAZ-AVI, ceftazidime/avibactam; INMB, incremental net monetary benefit; LOS, length of stay; Pip/Taz, piperacillin-tazobactam.
Discussion
This study assessed the outcomes of introducing CAZ-AVI as a new antibiotic to the current treatment strategy under five specified treatment strategies for the treatment of the most significant gram-negative hospital-acquired infections (HAIs) in Zhejiang province, China, using a previously published and validated dynamic disease transmission and cost-effectiveness model.20 To our knowledge, this is the first study to estimate the clinical and economic value of a new antibiotic considering the transmission and diversity value attributes from the STEDI framework, in the Chinese setting. Therefore, this analysis, advances the approach to evidence-based economic evaluation of antibiotics in China and can support health policy decision-making to combat AMR.
Introducing CAZ-AVI to the current treatment strategy of the modelled gram-negative HAIs in Zhejiang province, China, was associated with substantial health and economic gains, and a reduction in the rate of resistance gain in pip/taz and meropenem, over 10 years. If these outcomes were translated across the whole of China, the value of CAZ-AVI to the Chinese healthcare system, at a national level, would be significantly greater. The health economic improvements were primarily driven by the availability of an additional effective treatment. When CAZ-AVI was used in the third line, while it increased LYs and QALYs, it led to increased hospital LOS and higher hospitalisation costs, due to treating patients with CAZ-AVI who were unsuccessfully treated with pip/taz and meropenem. All treatment strategies with CAZ-AVI were dominant over the current treatment strategy, offering improved outcomes (LYs and QALYs) and lower costs. Considering INMB, the best outcomes were demonstrated when CAZ-AVI was used at the first line with all-lines diversified (¥24 billion (US$3.55 billion)); in the first-line position the benefits of low resistance in CAZ-AVI were most effectively realised within the treatment strategy. Conversely, when used in the third-line position INMB was ¥3.4 billion (US$486 million) lower (¥21.3 billion (US$3.1 billion)), with the second-line position in between these two strategies (¥22.9 billion (US$3.3 billion)). However, these health economic benefits need to be considered within the context of resistance gain; in the first-line position, resistance to CAZ-AVI developed at the quickest rate; however, the impact on resistance across all treatments can be managed with AMS practices such as diversification. This provides China-specific evidence that by adding a new antibiotic into current clinical practice clinical and economic outcomes can be improved, and equally importantly, a more diversified antibiotic treatment can contribute to slowing overall AMR gain.
In the current analysis the model was parameterised to reflect the setting of Zhejiang province; however, the GDP per capita of Zhejiang province, used to estimate the WTP threshold, is higher than the national average for China. If the WTP threshold was aligned to the Chinese national GDP per capita the overall NMB for introducing CAZ-AVI would be lower per QALY gained. It is important to note in an analysis considering the whole of the Chinese healthcare system although the NMB per QALY would be lower, the overall NMB would be considerably higher due to the number of patients who would be treated if the model was to consider the whole of.
AMR, much like climate change, is a global issue and requires coordinated global action. China has one of the largest global economies; therefore, the introduction of a pull incentive as novel reimbursement models in China, alongside other nations including the USA and the European Union, would provide a significant boost to global antimicrobial incentive required to encourage the sustained development of new antimicrobials. The impact of AMR disproportionately affects low- and middle-income countries, it is important that global policies to incentivise the development of antimicrobials consider methods to also improve access to new treatments in these countries, alongside public health and education policies to address antimicrobial resistance. This study shows that by assessing the value of antimicrobials considering their broader value, which is overlooked with conventional cost-effectiveness methodologies, healthcare providers can better recognise the true value of antimicrobials and therefore help to inform decision and policy aiming to tackle the issues of antimicrobial resistance and development of new effective treatments.
Analysis conducted by Matsumoto et al. established two scenarios and explored the value of three treatment strategies including CAZ-AVI, through pairwise comparisons within each scenario under the Japanese setting.35 Their results showed that introducing CAZ-AVI into clinical practice was associated with considerable clinical and economic benefits, which was also confirmed in our study. However, our study differed by, first, comparing outcomes to the current treatment strategy without the use of the new antibiotic (i.e., CAZ-AVI) as a control, allowing for comparisons of the value added with introducing CAZ-AVI. Second, our analysis explored the use of CAZ-AVI in all treatment lines and by implementing more antibiotic diversification strategies, we investigated the optimal treatment strategy for reducing the development of AMR and achieving improved clinical and economic outcomes.
AMR is associated with increased medical costs, hospitalisation rates, treatment failure rates, more critical illness and increased mortality.36 37 Without urgent action, the world may head towards a ‘post-antibiotic era’, where even minor bacterial infections may cause fatal consequences.38 Jim O’Neill predicted that AMR would cause 10 million deaths and a US$100 trillion economic loss worldwide by 2050; in China, AMR will be responsible for 1 million deaths and US$20 trillion economic losses.39 40 In 2017, China suffered ¥265 billion of direct and indirect economic loss due to AMR, which accounts for 0.37% of the annual GDP.41 Our study strongly suggests that adding CAZ-AVI to clinical practice would produce positive health and economic impact, that is, higher LYs and QALYs, lower hospitalisation costs and shorter hospital LOS. Furthermore, these benefits may be optimised when used at the first line as part of an AMS strategy.
Clinical measures such as optimising the duration of empirical therapy, implementing a restricted antibiotic prescribing system and antibiotic cycling have been proposed internationally to deal with AMR.42 In recent years, China has also taken various measures to address AMR.43–46 Since 2009, the Chinese government has implemented strict policies to curb the overuse and misuse of antibiotics, including restrictive and persuasive interventions targeting the primary healthcare and hospital sectors.43 In 2012, the Ministry of Health of the People’s Republic of China (now the National Health Commission of the People’s Republic of China) issued ‘Measures for the Management of Clinical Application of Antibacterial Drugs’. It was the first comprehensive regulation formally introduced to control AMR, covering various aspects, and required strict compliance at the hospital level. Regulations include the clear organisation and responsibilities for the clinical application management of antimicrobial drugs, specific measures for the clinical application management of antimicrobial drugs, supervision and management by administrative departments, and corresponding legal responsibilities of medical institutions. These requirements reflect the awareness and commitment of China to control AMR. After a decade of verification, these measures have indeed proven to be somewhat effective in combating AMR9; however, they have also brought forth some challenges.
The requirement for tier-3 hospitals to limit their antibiotic formulary to 50 entities significantly restricted the availability of antibiotics in clinical settings, especially limiting access to recently launched antibiotics. Antibiotics were categorised into three levels according to factors such as safety, efficacy, bacterial resistance and price: non-restricted use, restricted use and special use. Many new antibiotics were categorised as special use after entering the hospital’s formulary because their initial resistance rates were low, to prevent a rapid increase in resistance. Additionally, healthcare institutions ranked clinical departments and medical personnel based on their antibiotic usage, usage rates and intensity.7 Consequently, healthcare professionals exercised extreme caution when prescribing antibiotics. Therefore, whilst controlling misuse of antibiotics is critical, it is essential to also consider the accessibility of new antibiotics to ensure the most appropriate treatments are available optimising health outcomes.
Applying research findings to guide the value-based access and/or formulary decision-making of new antibiotics is crucial. International Federation of Pharmaceutical Manufacturers and Associations has proposed economic incentives, reimbursement reform and HTA reform to promote antibiotic R&D, aiming to achieve the accessibility of rational antibiotic use and reflect the overall value of new antibiotics.47 The UK has launched a subscription reimbursement model that was decoupled from sales volume, aiming to appropriately reflect population-level value and incentivise R&D.48 The USA proposed to separate high-value antibiotics from disease-related groups (DRGs) to improve the payment proportion of antibiotics. This strategy was approved by the Food and Drug Administration to treat serious or life-threatening infections. The International Classification of Diseases defined AMR patients as complication/comorbidity, thus obtaining higher DRGs payment standards.49 These crucial steps reflected recognition of the significance of new antibiotics and may ensure the controllability of health insurance funds and clinical outcomes. In China, various pieces of evidence are considered in health insurance decision-making, with HTA playing a significant role.14 However, HTA only accounts for direct benefits to the patient, underestimating the overall value of new antibiotics. We believe that considering the transmission and diversity value of new antibiotics in the process of national health insurance drug selection is in line with the global trend and will provide a better reflection of the true value of antibiotics.
In order to guide the R&D of new antibiotics, the WHO has developed a priority list of antibiotic-resistant bacteria. Among 50 antibiotics currently in clinical trials 32 are targeted at the priority pathogens listed by the WHO. However, compared with the existing antibiotics, most of these 32 new antibiotics in R&D have limited effectiveness, and only 2 are effective against multiresistance gram-negative bacteria.50 Clearly, there is a scarcity of and urgent need for antibiotics that can effectively treat multidrug-resistant bacterial infections. The results of our study have demonstrated the value of new antibiotics, if medical insurance decisions and hospital antibiotic usage policies could reflect this population-level value of new antibiotics, it may instil confidence in pharmaceutical companies to invest in the R&D of new antibiotics considering China’s vast market for antibiotic consumption.
There were several limitations to this study. First, this analysis focused on cIAI, HAP/VAP and infections with LTO caused by three pathogens, but there were far more multidrug-resistant infectious diseases pathogens in clinical practice, this highlighted the considerable value unaccounted for in this analysis; a restricted number of clinical indications and pathogens likely underestimated the total diversity value that could be realised. It is important to consider that the current analysis may not be generalisable to other indications and pathogens. In the future, we will include more indications and pathogens in the study; however, this is dependent on the scope of surveillance efforts. Second, LOS and daily hospitalisation costs data were collected from a single hospital with limited generalisability. Therefore, further work is necessary to improve generalisability by calibrating models in more locations, healthcare facilities and patient populations. Third, the regression equations captured the transmission dynamics from the previously published dynamic transition model, the transition parameters were derived via empirical calibration using historic resistance data from the UK. Therefore, the transmission dynamics may not reflect the infections and resistance gain in China as accurately. And due to the absence of the China-specific utility and treatment efficacy data for pip/taz and meropenem, data from the literature was used which may not be most reflective of local situations, although these factors were shown to not be key drivers of INMB. Furthermore, the model is developed from a disease transmission model where a number of inputs were calibrated empirically from observed data, therefore there are no estimates of uncertainty, when parameters are unbounded the model may result in uninformative of misleading estimates. And due to the number of parameters and the interdependencies a comprehensive and well-supported approach was not possible. However extensive deterministic sensitivity analyses and scenario analyses have been conducted to characterise the uncertainty and sensitivity of the model. Finally, this study only explored the transmission and diversity aspects of the STEDI framework of a new antibiotic. Future research should aim to develop a framework to quantify the additional STEDI value elements; efforts should also include generating data to support calibrating and parameterising such an economic evaluation.
Conclusions
The addition of CAZ-AVI as a new antibiotic in clinical practices in China has a significant impact on clinical and economic outcomes, emphasising the value of including this new antibiotic in clinical practice to enrich the antibiotic arsenal. The health economic benefits realised are influenced by how the new antibiotic is used; however, these benefits should be balanced carefully against the resistance gain in existing and the new antibiotics. This study provides quantitative evidence to inform reimbursement and/or hospital formulary decision-making in China.
Data availability statement
Data are available upon reasonable request. The data sets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by Zhejiang University School of Public Health Medical Ethics Committee with the reference number being ZGL202306-1. Consent for participation in the study was not obtained because we did not have direct contact with patients during the data collection process and we have applied for a waiver of informed consent. This waiver was included as one of the materials for our ethics application.
Acknowledgments
The authors thank James Dennis for providing medical writing support and editorial support, which was funded by Pfizer in accordance with Good Publication Practice (GPP3) guidelines (http://www.ismpp.org/gpp3).
References
Footnotes
WY and XZ contributed equally.
Contributors HD, WY, XZ and YC: Conceptualisation, Methodology. WY: Writing—Original Draft, Data Curation, Investigation, Formal Analysis, Interpretation and Visualisation. XZ, XS, SUK, YC, AA-T, JG and HD: Writing—Review and Editing. DY and PD: Resources, Supervision. HD and XZ: Project Administration, Funding Acquisition. HD is the guarantor.
Funding This work was supported by Pfizer Investment Co. Ltd., who provided support for the model development/analysis and medical writing of this study (the grant number: 2021-SKY-A07054-0010), and by Zhejiang Pharmaceutical Association Economics and Health Technology Evaluation Special Scientific Research Funding Project (the grant number: 2022-SKY-A07054-0009)
Competing interests YC, PD and AA-T are employees of Pfizer. JG is an employee of Health Economics and Outcomes Research Ltd.; Health Economics and Outcomes Research Ltd. received fees from Pfizer in relation to this study. HD and XZ are paid consultants to Pfizer from Zhejiang University and Shandong University, respectively.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.