(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Importance of attributes and willingness to pay for oral anticoagulant therapy in patients with atrial fibrillation in China: A discrete choice experiment

['Jiaxi Zhao', 'Centre For Safe Medication Practice', 'Research', 'Department Of Pharmacology', 'Pharmacy', 'Li Ka Shing Faculty Of Medicine', 'The University Of Hong Kong', 'Hong Kong Sar', 'The University Of Hong Kong Shenzhen Institute Of Research', 'Innovation']

Date: 2021-09

Abstract Background Adherence to oral anticoagulant therapy in patients with atrial fibrillation (AF) in China is low. Patient preference, one of the main reasons for discontinuation of oral anticoagulant therapy, is an unfamiliar concept in China. Methods and findings A discrete choice experiment (DCE) was conducted to quantify patient preference on 7 attributes of oral anticoagulant therapy: antidote (yes/no), food–drug interaction (yes/no), frequency of blood monitoring (no need, every 6/3/1 month[s]), risk of nonfatal major bleeding (0.7/3.1/5.5/7.8[%]), risk of nonfatal stroke (ischemic/hemorrhagic) or systemic embolism (0.6/3.2/5.8/8.4[%]), risk of nonfatal acute myocardial infarction (AMI) (0.2/1.0/1.8/2.5[%]), and monthly out-of-pocket cost (0/120/240/360 RMB) (0 to 56 USD). A total of 16 scenarios were generated by using D-Efficient design and were randomly divided into 2 blocks. Eligible patients were recruited and interviewed from outpatient and inpatient settings of 2 public hospitals in Beijing and Shenzhen, respectively. Patients were presented with 8 scenarios and asked to select 1 of 3 options: 2 unlabeled hypothetical treatments and 1 opt-out option. Mixed logit regression model was used for estimating patients’ preferences of attributes of oral anticoagulants and willingness to pay (WTP) with adjustments for age, sex, education level, income level, city, self-evaluated health score, histories of cardiovascular disease/other vascular disease/any stroke/any bleeding, and use of anticoagulant/antiplatelet therapy. A total of 506 patients were recruited between May 2018 and December 2019 (mean age 70.3 years, 42.1% women). Patients were mainly concerned about the risks of AMI (β: −1.03; 95% CI: −1.31, −0.75; p < 0.001), stroke or systemic embolism (β: −0.81; 95% CI: −0.90, −0.73; p < 0.001), and major bleeding (β: −0.69; 95% CI: −0.78, −0.60; p < 0.001) and were willing to pay more, from up to 798 RMB to 536 RMB (124 to 83 USD) monthly. The least concerning attribute was frequency of blood monitoring (β: −0.31; 95% CI: −0.39, −0.24; p < 0.001). Patients had more concerns about food–drug interactions even exceeding preferences on the 3 risks, if they had a history of stroke or bleeding (β: −2.47; 95% CI: −3.92, −1.02; p < 0.001), recruited from Beijing (β: −1.82; 95% CI: −2.56, −1.07; p < 0.001), or men (β: −0.96; 95% CI: −1.36, −0.56; p < 0.001). Patients with lower educational attainment or lower income weighted all attributes lower, and their WTP for incremental efficacy and safety was minimal. Since the patients were recruited from 2 major hospitals from developed cities in China, further studies with better representative samples would be needed. Conclusions Patients with AF in China were mainly concerned about the safety and effectiveness of oral anticoagulant therapy. The preference weighting on food–drug interaction varied widely. Patients with lower educational attainment or income levels and less experience of bleeding or stroke had more reservations about paying for oral anticoagulant therapies with superior efficacy, safety, and convenience of use.

Author summary Why was this study done? Patient adherence with oral anticoagulant therapy in patients with atrial fibrillation (AF) in China is poor. Patient preference on treatment attributes and their willingness to pay (WTP) might be key factors in patient adherence.

Results from studies using discrete choice experiment (DCE) outside China suggest that preferences and WTP varied in different populations.

Quantitative evaluation of patient preference on attributes of oral anticoagulants and WTP in China remains an area that needs further exploration. What did the researchers do and find? We conducted a survey using a DCE design to investigate preferences on attributes of oral anticoagulants and WTP in patients with AF from Beijing and Shenzhen, 2 large representative cities in China.

Patients were asked to choose their preferred treatment based on their evaluation of the attributes of oral anticoagulant: antidote, food–drug interaction, frequency of blood monitoring, risk of nonfatal major bleeding, risk of nonfatal stroke or systemic embolism, risk of nonfatal acute myocardial infarction (AMI), and monthly out-of-pocket cost.

Patients were mostly concerned about the risks of AMI, stroke or systemic embolism, and major bleeding and were willing to pay 798 to 536 RMB (124 to 83 USD) per month to obtain a better drug with less risks. Food–drug interaction was weighted to be the most important attribute by patients who had history of bleeding or stroke or who were from Beijing. What do these findings mean? The relatively highest preference weights for the risk of both cardiovascular and cerebrovascular events indicate that patients might be more attentive to the safety of vital organs than the main side effects of anticoagulant therapy.

Patients with different demographic or health characteristics have specific preferences. The relatively high preference weight and marginal willingness to pay (mWTP) for food–drug interaction among specific patients suggest that patients might prefer the most direct option offering rapid improvement.

Citation: Zhao J, Wang H, Li X, Hu Y, Yan VKC, Wong CKH, et al. (2021) Importance of attributes and willingness to pay for oral anticoagulant therapy in patients with atrial fibrillation in China: A discrete choice experiment. PLoS Med 18(8): e1003730. https://doi.org/10.1371/journal.pmed.1003730 Academic Editor: Suzanne C. Cannegieter, Leiden University Medical Center, NETHERLANDS Received: February 10, 2021; Accepted: July 12, 2021; Published: August 26, 2021 Copyright: © 2021 Zhao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: As the data was collected via a questionnaire survey, due to the nature of ethical restriction in this study, participants of this study did not agree for their un-anonymized data to be shared publicly. Authors of this article confirm that all anonymized data supporting the findings of this study are available within the article and its supplementary materials. Funding: EWC received funding from National Natural Science Fundation of China for this study (Ref Number: 71704149). (URL is http://www.nsfc.gov.cn/) The funder had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. Competing interests: The authors declare no competing interests. Abbreviations: AF, atrial fibrillation; AMI, acute myocardial infarction; CHA2DS2-VASc, congestive heart failure (C), hypertension (H), age ≥ 75 (A2), diabetes mellitus (D), prior stroke or transient ischemic attack or thromboembolism (S2), vascular disease (V), age between 65 and 74 (A), sex category (Sc); CI, confident interval; DCE, discrete choice experiment; EHRA, European Heart Rhythm Association; INR, international normalized ratio; mWTP, marginal willingness to pay; NOAC, non-vitamin K antagonist oral anticoagulant; RMB, Ren Min Bi (unit of Chinese currency); SD, standard deviation; USD, US dollar (unit of American currency); WTP, willingness to pay

Introduction Oral anticoagulant therapy is a standard management for stroke prevention in patients with atrial fibrillation (AF). In line with the increasing global prevalence of AF over the past decade [1], the age- and sex-standardized prevalence of AF in Chinese over 35 year olds has continued to rise, reaching 0.71% in 2017 [2]. Due to rapid action onset and offset, predictable pharmacodynamics, and fewer drug–food interactions, non-vitamin K antagonist oral anticoagulants (NOACs) are increasingly favored in clinical practice over warfarin worldwide [3] and has been recommended in the latest international clinical practice guidelines [3–7]. However, adherence to NOAC therapy among patients in China is lower than warfarin [8,9], although the use of NOACs was found to be increasing rapidly [10]. One of the main reasons for discontinuation of NOACs is patient preference or affordability [8]. A multicenter cross-sectional study found that out-of-pocket payment was the strongest predictor for declining anticoagulant therapy use in China [11]. In China, although NOACs have 92% of the oral anticoagulant therapy market share in terms of monetary value, only 28% of patients with AF were prescribed a NOAC in 2017 [10]. Currently, for NOACs prescribed in outpatient setting, only 70% of the cost will be subsidized [11]. Achieving a sufficient reduction in the risk of ischemic stroke with an acceptable elevation in bleeding risk is an important clinical trade-off that patients need to consider [12]. In addition, patients’ perspectives concerning difficulties of access to the healthcare service and the complexity of treatment will also influence medication adherence [13,14], a crucial issue for anticoagulant therapy as emphasized in the 2018 European Heart Rhythm Association (EHRA) practical guide [6]. However, current evidence is insufficient to rank the importance of factors affecting patients’ medication adherence in China. Discrete choice experiment (DCE) is a method of eliciting and quantifying preferences and exploring trade-offs between the attributes (characteristics) of alternatives [15]. An Australian DCE study [12] indicated a consensus that patients were likely to choose more effective and safer oral anticoagulant therapies and would prefer to pay more for a reduction in the risk of stroke and bleeding. Although one United States study [16] supported this viewpoint, the lower weight of these 2 attributes compared to blood testing and dietary restrictions heightened the concern for preference heterogeneity between different populations. This heterogeneity existed in the same country. Another US study [17] revealed that patients with cardiovascular disease were not as concerned about the effectiveness and safety, including fatal events, of oral anticoagulants as patients in the previous 2 studies. Inconsistent results, as well as cultural differences between China and high-income countries, which could potentially impact the perspectives of patients receiving treatment, mean that conclusions from previous studies conducted outside China may not be generalizable to Chinese patients. To our knowledge, this is the first study to apply DCE to the investigation of preferences on attributes of oral anticoagulant therapy and willingness to pay (WTP) among patients with AF in China.

Results Characteristics of patients (Table 2) After excluding one patient who did not answer any question, a total of 506 patients (299 from Beijing and 207 from Shenzhen), with mean age 70.3 years, of which 42.1% were women, were successfully recruited and completed consent forms. Regardless of testing scenario, each patient was given 8 scenarios with 2 treatment alternatives and one opt-out option to choose from, and a result of 12,144 (506 × 8 × 3) completed records were expected. After removing 201 empty responses, 11,943 records were kept for further analyses. The first and second blocks were answered by 254 and 252 patients, respectively (S4 File). A total of 36 out of 506 patients (7.1%) did not choose the dominant drug in the testing scenario. Main analysis (Table 3) The positive and negative values of coefficients represent patients’ preference and dislike for the corresponding attribute, respectively. The greater the absolute value of the coefficient, the more weight for changes per unit of that attribute were taken into consideration when patients evaluated the trade-offs. Compared to other attributes, the relatively higher weights for the risk of nonfatal AMI (β: −1.03; 95% CI: −1.31, −0.75; p < 0.001), the risk of nonfatal stroke or systemic embolism (β: −0.81; 95% CI: −0.90, −0.73; p < 0.001), and the risk of nonfatal major bleeding (β: −0.69; 95% CI: −0.78, −0.60; p < 0.001) indicate that the issues patients had the most concern about during anticoagulant therapy were safety and effectiveness. The impact of the frequency of blood monitoring on decision-making was the weakest (β: −0.31; 95% CI: −0.39, −0.24; p < 0.001), and patients disliked frequent hospital visits. Patients preferred treatment with the antidote (β: 0.43; 95% CI: 0.17, 0.69; p = 0.001), while avoiding food–drug interactions (β: −0.63; 95% CI: −0.88, −0.38; p < 0.001). Patients were willing to pay most of the cost for oral anticoagulant therapy: 798 RMB (β: −798; 95% CI: −1,247, −518), 632 RMB (β: −632; 95% CI: −880, −488), and 536 RMB (β: −532; 95% CI: −762, −403), respectively, to reduce 1% of the risk of nonfatal AMI, nonfatal stroke or systemic embolism, and nonfatal major bleeding. They began to hesitate about paying more to be free of food–drug interaction (β: −489; 95% CI: −764, −288) or to obtain the extra antidote for side effects and reactions (β: 338; 95% CI: 125, 639) and were willing to pay only the minimum to reduce the frequency of blood monitoring (β: −244; 95% CI: −347, −180). Subgroup analyses (Table 4) Men were slightly more concerned about almost all of the attributes compared to women and were willing to pay double that of women. The impact of food–drug interaction on men (β: −0.96; 95% CI: −1.36, −0.56; p < 0.001) was much stronger than that on women (β: −0.34; 95% CI: −0.73, 0.05; p = 0.09), and they were willing to pay over 4 times more than women (805 versus 174 RMB [125 versus 27 USD]). As is widely known, education level correlates with income level. The similarity in patterns of preferences on all attributes could also be observed for both subgroups; patients with lower educational attainment (at or under high school) or who had a lower income (≤5,000 RMB/month) were less concerned about all of the attributes and were only willing to pay the minimum to obtain extra benefits or get rid of disadvantages. mWTP varied widely among patients with higher levels of education and income. History of any bleeding or stroke influenced patients’ preference. Anyone with a history valued all attributes at the same high weights and were willing to pay up to 5 times of those without a history. For this group, food–drug interaction was the most concerning attribute (β: −2.47; 95% CI: −3.92, −1.02; p < 0.001). We stratified the analysis by cities to test heterogeneity. Compared to Shenzhen, the effect size of all attributes on patients from Beijing was chiefly doubled. The food–drug interaction had the strongest impact on patients from Beijing (β: −1.82; 95% CI: −2.56, −1.07; p < 0.001). However, they were only willing to pay less than the amount that patients from Shenzhen were prepared to pay for most of the attributes. Sensitivity analyses (S5–S11 Files) First, we removed patients who did not initially select the dominated alternative in the testing scenario to assess data quality. Preferences and mWTP for each attribute were consistent with the primary analysis (S5 File). With consideration that the linear assumption might be violated in the ordinal variable frequency of blood monitoring since the intervals between each level were unequal, we also treated this as a categorical variable (S6 File). Compared to nonnecessity for monitoring, patients preferred less frequent 6 monthly hospital visits (β: 0.43; 95% CI: 0.18, 0.69; p < 0.001). The extent of dislike deteriorated as the frequency increased (every 3 months β: −0.30; 95% CI: −0.54, −0.05; p = 0.02 and every 1 month β: −0.85; 95% CI: −1.09, −0.61; p < 0.001). The consistent results of the sensitivity analyses, which adjusted blocks of scenarios, stratified by outpatient and inpatient settings, and excluded patients who did not have any medical insurance, from main analysis, indicated that there would not be obvious influence from those aforementioned factors on patients’ preferences (S7–S9 Files). By adopting a nested logit model to assess the effects of opt-out on decision-making, the preferences for all attributes were principally similar to that in the main analysis. However, due to the inability to deal with the correlated alternative selection, the results from multinomial logistic regression were not consistent with the main analysis (S10 and S11 Files).

Discussion In view of the underutilization of oral anticoagulant therapy and poor medication adherence among patients with AF in China, this study provides an in-depth exploration of the crucial factors that could affect patients’ preferences, allowing either possible product improvements or therapeutic focused adjustments. In addition, as apixaban and edoxaban have not yet been approved for use in China, the results from mWTP could assist with reasonable pricing for the remaining NOACs in China. The magnitude of concerns about the attributes vary in different populations. In general, the risks of AMI, stroke or systemic embolism, and major bleeding, which could intuitively reflect drug safety and effectiveness, emerged as characteristics for which patients were willing to pay most for improvement. The relatively higher preference weights for the risk of both cardiovascular and cerebrovascular events indicate that patients might be more attentive to the safety of vital organs than the common severe side effects during anticoagulant therapy. Disease conditions of the heart or brain necessitate inevitable long-term medical care, which place a heavy burden on most families in China. This finding is consistent with another US study [17] that was of relative importance for AMI and stroke compared to major bleeding in our study ( = 1.49, = 1.17) and is similar to theirs ( = 2.87, = 2.26). However, the relative importance of blood monitoring compared to major bleeding in our study ( = 0.45) was 9 to 15 times higher than that in 2 studies ( = 0.049) [12] ( = 0.032) [17], which suggested that Chinese patients might have different perspectives about the attributes compared to patients from high-income countries. A recent study found that the general spatial accessibility to hierarchical healthcare facilities in Shenzhen is unevenly distributed and concentrated [33]. Also, the supply of healthcare resources at primary care facilities is far from sufficient. This might reveal the reason why patients in our study were more concerned about blood monitoring. More frequent blood test monitoring is inconvenient for many patients who do not reside near their primary healthcare facility. Another study focused on oral anticoagulation adherence of patients in China suggested that consecutive international normalized ratio (INR) monitoring was a major concern of patients, especially for those who are older, women, and have previous stroke history, as they are at a higher risk of bleeding and need more intensive INR monitoring [34]. Patients with different demographic or health characteristics have their specific preferences. The most surprising finding was that the preference weight of food–drug interaction was the highest among men, patients who had been recruited from Beijing, and who had a history of bleeding or stroke. The dietary habits of Chinese is complex. Green vegetables, rich in vitamin K, which could influence the effectiveness of vitamin K antagonist oral anticoagulants, are usually a main dietary preference. Drinking alcohol is also taboo while under drug treatment. However, one of the main social activities in China is attending banquets, especially common among northern residents and men. This might be the reason that the preference weight among these patients was high. Patients with history of bleeding or stroke who were concerned mainly with food–drug interactions could have experienced severe adverse reactions caused by poor dietary habits. In addition, the relatively high mWTP for this attribute suggest that patients might prefer the most direct option offering rapid improvement. Currently, patients with AF in China are generally required to pay an additional 150 to 300 RMB per month for anticoagulant therapy using NOAC. According to our results, patients were willing to pay a little more than the current monthly expenditure for oral anticoagulants with better safety, effectiveness, and convenience profiles. Nevertheless, if the drug attribute preferences of patients with low income were the same as that of the overall population, their WTP would be understandably low. It is worth noting that the preferences of patients with high income and high educational attainment varied widely. The insignificance of out-of-pocket costs suggests that current incremental cost was within the affordable range for this population.

Limitations There were several limitations in this study. First, as discussed in other stated preference surveys, the true preferences were not revealed because the decisions made were merely hypothetical. Second, all scenarios only considered a limited number of attributes. Other attributes, such as the risk of bleeding with different severity levels and side effect related death, for example, may also reflect other preferences. Nevertheless, those 7 included attributes were the one mostly concerned by patients in other DCE studies. Regardless of sample size increase, too many attributes would also increase patient workload, affecting their decision-making due to information overload and reduce the reliability of the result. Some attributes, such as death, could be a distraction and bias the result. Third, we cannot fully ascertain whether patients prefer NOACs or warfarin due to the unlabeled design. However, patients might be biased by the names of alternatives without an understanding of the products. Moreover, the unlabeled approach may help identify patients’ preferences more accurately. Forth, due to the nature of DCE, there may not be a flawless analytic approach to obtain a comprehensive result. However, we have conducted several sensitivity analyses using different models with consistent results. Fifth, the levels of risks of major bleeding, stroke and systemic embolism, and myocardial infarction were obtained from studies mainly on non-Chinese/Asian populations. The estimated preferences on these attributes might be less generalizable to a Chinese population. However, the findings may not be too biased due to the linear assumption, as long as the differences of those risks between the 2 populations are not large and are still useful especially for the firsthand evidence from an exploratory study. Lastly, the 2 hospitals that were used to recruit patients are the major hospitals in 2 developed and affluent cities in China. Patients from these 2 hospitals may not be representative of patients in China in general.

Conclusions Patients with AF in China were mainly concerned with the safety and effectiveness of oral anticoagulant therapy. The preference weight on food–drug interaction varied widely. Patients with lower educational attainment or low income and less experience of bleeding or stroke were more conservative about paying for oral anticoagulant therapies with superior efficacy, safety, and convenience of use. Ethical approval and informed consent This study was approved by Medical Ethics Committees of the Chinese People’s Liberation Army General Hospital (Reference number: S2018-043-01) and HKU-Shenzhen Hospital (Reference number: [2018]78).

Acknowledgments We thank Dr. Kelvin K.H. Yiu and his colleagues from the Department of Cardiology, The University of Hong Kong Shenzhen Hospital for their assistance in coordinating the patient recruitment process. We thank Ms. Yue Wei, Ms. Le Gao, Ms. Lam Lam, and Ms. Ying Shi, who helped with data entry. We acknowledge that EWC has received honorarium from the Hospital Authority, research grants from Narcotics Division of the Security Bureau of HKSAR, National Health and Medical Research Council (NHMRC, Australia), National Natural Science Foundation of China (NSFC), Research Fund Secretariat of the Food and Health Bureau (HMRF, HKSAR), Research Grants Council (RGC, HKSAR), Wellcome Trust; Amgen, AstraZeneca, Bayer, Bristol-Myers Squibb, Janssen, Pfizer, RGA, and Takeda outside the submitted work. XL has received the research grant from Health and Medical Research Fund, Food and Health Bureau of Hong Kong Government; research and educational grants from Janssen and Pfizer; internal funds from the University of Hong Kong; consultancy fees, unrelated to this work. CKHW has received research funding from the EuroQoL Group Research Foundation, the Hong Kong Research Grants Council, and the Hong Kong Health and Medical Research Fund, Food and Health Bureau. GYHL is a consultant and speaker for BMS/Pfizer, Boehringer Ingelheim, and Daiichi-Sankyo; no fees are received personally. Aforementioned funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

[END]

[1] Url: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1003730

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.