Utilisation rate of non-vitamin K antagonist oral anticoagulant, and associated factors of refusal of non-vitamin K antagonist oral anticoagulant usage in atrial fibrillation patients - A study in two Hong Kong general out-patient clinics

Liujing Chen 陳柳靜, Chik-pui Lee 李植沛, Lit-ping Chan 陳列萍, Eric MT Hui 許明通, Maria KW Leung 梁堃華

HK Pract 2023;45:89-96

Summary

Objective: To evaluate the utilisation rate of non-vitamin K antagonist oral anticoagulant (NOAC) and associated factors of NOAC refusal in atrial fibrillation (AF) patients.
Design: A cross-sectional study.
Subjects: All the AF patients, who were regularly followed up in two public general out-patient clinics (GOPC) from November 2019 to March 2020, aged older than 18 years, and eligible for NOACs.
Main outcome Measures: Utilisation rate of NOAC in AF patients. The associated factors of NOAC refusal in AF patients.
Results: A total of 324 patients were included during the study period. Utilisation rate of NOAC in AF patients was 54%. Multivariate analysis revealed that older age, higher financial strain score, lack of sponsor for NOAC and lower CHA2DS2-VASc score were the factors that were significantly associated with NOAC refusal.
Conclusion: In our study, the utilisation rate of NOAC in AF patients in the two local GOPCs studied was 54%, which implied there was still room for improvement. The associated factors of NOAC refusal highlighted the importance of financial support to promote the usage of the NOAC. Further research and strategy to improve guideline attainment should focus on the subgroup patients who were older and who had a lower CHA2DS2-VASc score.
Keywords: Atrial fibrillation, anticoagulation, non-vitamin K antagonist oral anticoagulant, Hong Kong, primary care

摘要

目的 ： 評估非維他命K拮抗劑類口服抗凝血劑(NOAC)在心 房纖顫(AF)患者中的使用率，並且分析患者拒絕此藥的相關 因素 。
設計 ： 橫斷面研究。
對象 ： 2019年11月至2020年3月於兩間普通科門診常規複診 的AF成年患者，並且符合NOAC使用指徵。
主要測量內容 ： NOAC在AF患者中的使用率，以及患者拒 絕此藥的相關因素。
結果 ： 共324位患者加入研究。NOAC在AF患者中的使用 率為54%。多因素分析結果顯示年老、高財務壓力評分、 NOAC缺乏資助、以及CHA2DS2-VASc低分是拒絕NOAC的 相關因素。
結論 ： 本研究顯示，NOAC於兩間本地普通科門診AF患者 中的使用率為54%，仍然有待提高。相關因素分析顯示： 資助NOAC對於提高其使用率相當重要；對於年老以及 CHA2DS2-VASc分數低的患者，仍然需要進一步的分層分 析，尋找提高他們使用NOAC的途徑。
關鍵詞 ： 心房纖顫，抗凝血劑，非維他命K拮抗劑類口服 抗凝血劑，香港，基層醫療

Introduction
Background and objectives

Atrial fibrillation (AF) is the most common cardiac arrhythmia¹, with a lifetime prevalence of one fourth of patients >40 years old.² In a Hong Kong territory-wide community-based AF screening programme, the prevalence of AF detected by smartphone-based wireless single-lead ECG or self-reported by participants was 8.5%.³

Individuals with AF have an increased risk of stroke, and account for up to 1 of 3 stroke cases among the elderly^4,5, potentially leading to permanent disability and death.⁶ Thus, stroke prevention in AF patients is an urgent healthcare and public-health concern.

Anticoagulant is the most important modifiable factor to reduce stroke incidence in AF.⁷ An old drug used to reduce stroke incidence, Warfarin, is a long-established anticoagulant.⁸ However, since the introduction of the non-vitamin K antagonist oral anticoagulant (NOAC) in the 2010s, these new drugs have changed the landscape for stroke prevention in AF patients. Dabigatran was the first NOAC on the market which was approved for stroke prevention in patients with non-valvular AF by the US Food and Drug Administration in 2010. Since then, 3 other NOACs (rivaroxaban, apixaban, and edoxaban) are available in many countries worldwide, including Hong Kong. Of the 3, edoxaban, which was the last of them, became registered in the Hong Kong Drug Office in 2016.

NOACs are not inferior to warfarin when used for stroke prevention^9,10, and some analyses of clinical effectiveness suggests that they are actually preferable.¹¹ NOACs are indicated to prevent stroke in patients with non-valvular AF by both the European Society of Cardiology (ESC) and the American Heart Association (AHA) guidelines for patients with CHA2DS2-VASc Score ≥1 in male, and ≥2 in female.^1,12,13

However, there is still a great gap between guidelines and the clinical utilisation rate of NOACs. A nationwide study in Korea found that NOAC use increased from 0% in 2002 to 14.% in 2016 in AF patients.¹⁴ A Taiwan study of 181214 newly diagnosed AF patients revealed that the NOAC use increased from 0% in 2008 to 26.0% in 2015.¹⁵

In order to promote the usage of the NOACs in AF patients, dabigatran and apixaban were introduced into our general out-patient clinics (GOPCs) of the Hospital Authority (HA) in mid 2019. The prescription of these NOACs was limited to patients with a high risk of stroke whose CHA2DS2-VASc score was 5 or more. In actuality, a number of AF patients still refused NOACs, even through they are eligible for its use. Is this refusal related to the problem of patient affordability? What is the utilisation rate of NOAC in the GOPC setting?

So far, there is no study on the utilisation rate of NOAC in the GOPC setting and no local data on the associated factors for NOAC refusal. Therefore, this study was conducted to evaluate the utilisation rate of NOAC, and the factors independently associated with NOAC refusal in AF patients, aiming at locating the barriers to optimal NOAC use in the GOPC setting.

Method
Design and setting of the study

This was a cross-sectional study. A quantitative method was selected because this approach can objectively reflect the facts.

This study was conducted in two GOPCs in Tai Po, with ethics approval by the Local Ethics Committee (CREC Ref. No. 2019.516). The flowchart in Figure 1 illustrated the patients’ enrolment in the study. A list of patients with International Classification of Primary Care (ICPC) coding of AF(K78) was retrieved from the Clinical Data Analysis and Reporting System (CDARS) of HA. All AF patients who were regularly followed up in participating clinics were invited to attend the Atrial Fibrillation Clinic (AFC). Patients aged older than 18 years were recruited during initial visit in November 2019 to December 2019. A follow-up visit was arranged 3 months after initial visit to review their option of NOAC use and compliance. Study subjects were seen by the principal investigator in the AFC, with active review of the clinical condition and discussion of the NOACs options. The suggestions of NOACs followed stroke prevention guidelines recommended by the European Society of Cardiology (ESC) 2016 and the American Heart Association (AHA) 2014 and 2019.^1,7,13 Patients on warfarin were mainly followed up in Specialist Outpatient Clinics (SOPDs). Hence, we do not initiate warfarin in our GOPCs. Informed consent was also obtained from the study subjects before enrolment into the study.

Figure 1:

Patient selection

Patients having one or more of the following were excluded:
(i) Males with CHA2DS2-VASc Score = 0, and females with CHA2DS2-VASc Score = 1
(ii) Patients with moderate-to-severe mitral stenosis
(iii) Prosthetic valve or valve repair
(iv) Child-Pugh category C hepatic insufficiency.
(v) Severe renal failure with CrCl < 15mL/min or on dialysis
(vi) Clinically significant active bleeding
(vii) HAS-BLED score¹⁶ ≧ 3 or history of non-traumatic intracranial haemorrhage
(viii) Pregnancy or breastfeeding mother
(ix) All Current hospitalisation or hospitalisation within one month prior to inclusion in the study.
(x) Allergy to NOACs
(xi) Those who refuse to join the study, or not competent to consent

Data collection

Baseline demographic and clinical data was collected and recorded via a questionnaire and a review of the medical records in the Hospital Authority’s Clinical Management System (CMS).

The following data were collected (1) sociodemographic information including age, sex, education level, living condition, marital status, financial strain, employment status; (2) drinking and smoking habit; (3) any sponsor for NOACs; (4) duration of atrial fibrillation; (5) CHA2DS2-VASc component details; (6) NOAC use.

Financial strain data was collected instead of income as it was a more powerful control variable than income.¹⁷ The measure of financial strain was based on four items: three items asked respondents whether they had enough money to pay for their needs in food, in medical services, and daily expenses, using a three-point scale ranging from 1 = enough, to 3 = not enough. The fourth question asked respondents to rate how difficult it was for them to pay their monthly bill using a four point scale, ranging from 1 = not difficult at all, to 4 = very difficult. A sum of the scores of these four items was computed, yielding a range from 4-13, with high scores indicating greater strain.¹⁸

Drinker was defined as more than 7 units (for women) or 14 units (for men) of alcohol in a week. Smoker was defined as active smoking or smoking cessation less than 1 year.

Sponsor for NOAC was defined as (a) CHA2DS2- VASc score ≧5 hence no extra charge for NOAC in GOPC, or (b) Civil service eligible persons with full reimbursement of NOAC, or (c) Full reimbursement of NOAC by insurance.

CHA2DS2-VASc score was calculated according to the AHA guideline¹³: congestive heart failure, hypertension, age ≥75 years (doubled), diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism (doubled), vascular disease, age 65 to 74 years, sex category, with theoretical score range 0–9.

Outcome variable

NOAC refusal group was defined as patient refusing to take any NOAC during the consultation in follow up visit. NOAC non-refusal group was defined as those who used NOAC for at least 3 months and was willing to continue NOAC during follow up visit.

NOAC refusal group was coded as “1 = refuse” and “0 = not refuse” for the logistic model.

Statistical analysis

SPSS 24.0 software (SPSS, Chicago, IL, U.S.A.) was used for analysis of data. Continuous data are presented as mean ± standard deviation, and categorical data are shown as number and percentage. Chi-square test was used to compare categorical variables. For continuous variables, independent T-test was used for comparing two groups. Data was compared between the NOAC refusal group and non-refusal group. Binary logistic regression analysis was performed to identify factors significantly associated with the refusal of NOACs. Adequate subject number was based on 10 events per variable (EPV) in logistic regression analysis.¹⁹ A P-value of less than 0.05 was considered to be statistically significant.

Results
(A) Study population

A total of 324 AF patients from two GOPCs were recruited. Average age was 76.6±9.7 years, and 192(59.3%) were male. Baseline demographic data, clinical characteristics, and use of NOACs are shown in Table 1. Among recruited patients, no one was younger than 50 years old. The majority of patients (96.6%) were older than 60 years old and most of the patients had primary school education or below. 215(66.4%) patients had no sponsor for NOACs. 109(33.6%) patients had sponsor for NOACs and 21 patients among them refused NOACs use. 277 patients (85.5%) had a history of hypertension while 99(30.6%) patients had a history of diabetes mellitus. Average CHA2DS2-VASc score was 3.6±1.5 in this population.

(B) Utilisation rate of NOAC

Usage of NOACs were by 175(54.0%) patients. Apixaban was 141 patients, (80.6% of all NOAC user) (Table 1).

Table 1:

(C) Associated factors of NOAC refusal

Comparisons of data between NOAC refusal group and non-refusal group were shown in Table 2.

Multivariate binary logistic regression analysis revealed that older age (Adjusted OR 1.061, CI 1.000- 1.126, P=0.048), higher financial strain score (Adjusted OR 1.592, CI 1.009-2.509, P=0.045), no sponsor for NOAC (Adjusted OR 2.619, CI 1.110-6.177, P=0.028) and lower CHA2DS2-VASc score (Adjusted OR 0.472, CI 0.223-0.997, P=0.049) were independent associated factors for NOAC refusal in AF adults.

Discussion

This study was conducted in AF patients who were regularly followed-up in two GOPCs in Hong Kong. All of the recruited patients were eligible for the use of NOACs. The utilisation rate of NOAC in this population was 54%, with room for improvement.

Overseas studies showed great disparity of NOAC utilisation rate according to study time, region, and type of institute. A Taiwan study with 181,214 newly diagnosed AF patients revealed that the NOAC use increased from 0% in 2008 to 26.0% in 2015.¹⁵ A study of 888,540 AF patients in Korea found that the usage rate of NOAC was significantly different among different medical systems from 37.2% at the tertiary referral hospital and 5.5% at nursing or public health centers.¹⁴ Another nationwide study in Korea showed the proportions of prescribed NOACs to total oral anticoagulants were 5.1%, 36.2%, and 60.8% in 2014, 2015, and 2016, respectively.²⁰ A primary care study in UK between June 2012 and June 2014 revealed that 53% AF patients who were not on anticoagulation agreed to start NOAC.²¹

Table 2:

It was difficult to directly compare our NOAC utilisation rate with previous studies as some of them included warfarinised cases in their study population^14,15,21 and some of them aimed to analyse the NOAC utilisation rate among all the anticoagulant users.²⁰ We could calculate the NOAC utilisation rate in AF patients according to the data of the Taiwan study to be 28.8% in 2015¹⁵, which was lower than our data. However, our data collection started in 2019 and a higher utilisation rate was not surprising. Overall, this study and overseas studies showed that there was still a great gap between guidelines and the clinical utilisation rate of NOAC.

Another aim of this study was intended to identify the associated factors of NOAC refusal in the GOPC setting.

Our result showed that the lack of NOAC sponsor and higher financial strain score were significantly associated with NOAC refusal. NOACs were introduced to GOPC in Hong Kong by the Hospital Authority in mid 2019. However, the prescription was only limited to patients with a high risk of stroke whose CHA2DS2- VASc scores were 5 or above. Eligible patients can be prescribed NOACs in the GOPCs without extra cost after consultation. The non-eligible patients can purchase NOACs in the community pharmacy with a prescription issued by the GOPC doctor (they had to pay HK$600 to HK$1500 per month according to the type of NOAC and dosage). Obviously, for these patients whose CHA2DS2-VAS2 scores were below 5, affordability had a direct impact on their NOAC use, and our findings shared similarities with other studies.

The previous studies demonstrated how health policy and insurance influenced the use of NOACs. In Korea, the policy of health insurance coverage for NOACs was revised in July 2015 to allow a broader coverage, and a study showed a significant growth rate of NOAC prescription after this was introduced.²⁰ Another nationwide study in Korea about newly diagnosed AF patient revealed that partial and full reimbursement of NOAC were independently associated with higher anticoagulant use.¹⁴ There was a study on associated factors for anticoagulants (including NOAC and warfarin) use in 593 non-valvular atrial fibrillation patients in China Jiangsu province, which showed self-paying was negatively associated with anticoagulant therapy in all patients.²²

Based on the result of our study, a broader sponsorship for NOACs in primary care was suggested for the purpose of promoting the usage of NOAC. Nevertheless, we also noticed in the group of 109 patients who were eligible for NOAC sponsorship, 21 patients (19.3%) still refused NOAC use, which meant that patient affordability was not the only factor associated with NOAC refusal. Unfortunately, a sample size of 21 patients would not be sufficient to support further quantitative analysis. Further research involving a bigger sample size in this subgroup of patients would be helpful to identify the barriers of NOAC use apart from the money issue.

A previous study about NOAC adherence showed education level and information about the disease could affect the medication use.²³ In our study, however, patient’s education level was not related to NOAC refusal based on the results of the multivariable analysis. One of the possible explanations was that the patient’s education level might not be equal to their knowledge level which potentially influenced the decision of NOAC use. Furthermore, the education level was difficult to be graded in the elderly (mean age of our study subjects was 76.6 years old) as most of them did not receive formal education. Still, one had to bear in mind that a multitude of factors besides the knowledge level of patients potentially influenced the choice of medication. Patients’ perspectives, perceptions and attitudes cannot be well assessed in a quantitative study. Additional qualitative research is needed to unravel and understand these factors influencing NOAC use in patients.

Older age was an independent predictor of NOAC refusal in our study. The elderly with AF, especially those aged ≥75 years, are considered to have at least a CHA2DS2-VASc score of 2.¹³ This population is the group with the highest risk of stroke and the worst prognosis, thus, oral anticoagulant is certainly recommended. However, the data on the NOACs option in the elderly was not sufficient in previous studies. A small sample study among non-valvular AF patients in China indicated that increasing age was negatively associated with anticoagulant therapy (including NOAC and warfarin).²² Plenty of studies about warfarin use in the elderly showed that overestimation of the bleeding risk and disadvantages associated with advanced age are barriers to the prescription of oral anticoagulants in the elderly.²⁴ Therefore, the underutilisation of NOAC in the elderly group might share similar reasons. Other possible reasons include relatively lower mental capacities to comprehend benefit and risk of NOAC and difficulty to negotiate with family members before decision making.

In general, the decision to prescribe NOAC in the elderly is complicated. It requires not only to balance the stroke risk and bleeding risk, but also the need to consider the patient’s general health, functional and cognitive ability, availability of a caregiver, and patient’s attitude and preference towards anticoagulation. More attention should be paid to this group of elderly patients during consultation and setting up of a special clinic with a multidisciplinary approach to provide patient education, medication monitoring and dosage adjustment might help.

Low CHA2DS2-VASc score was another associated factor for NOAC refusal in our study. A study in Thailand for AF patients aged ≥65 years showed CHA2DS2-VASC score 1, to CHA2DS2-VASC score ≥2, increases the rate of non-prescription of anticoagulant.²⁵ However, the results of this study could not be compared to our study which defined CHA2DS2-VASc score as scale variable and excluded CHA2DS2-VASc Score =0 in male or CHA2DS2-VASc Score =1 in female. A Taiwanese study had similar results as our study. Patients who were not on any antithrombotic therapy tended to have lower CHA2DS2-VASc score (5.1±1.6) than those taking antiplatelet agents (5.6±1.5) or warfarin (5.7±1.5).²⁶ One explanation was that lower CHA2DS2-VASc score implied lower risk of stroke and hence, lower cost effectiveness of NOAC. However, the CHA2DS2-VASc score could not predict the severity of stroke, whether it’s major stroke or TIA. This study demonstrated a barrier in initiating NOAC when patients have lower CHA2DS2-VASc scores. This group of patients should not be neglected, and methods of improving NOAC use including educational programme should be put in place.

Strengths and limitations

There are strengths in this study. Firstly, this is the first paper to quantify the use of NOACs in AF patients in the Hong Kong GOPCs. There are important clinical and policy implications. The use of NOACs in GOPCs will become comparable and traceable. It also facilitates policy holders in resource allocations and planning future medical expenditures. Secondly, patient consultation was conducted by the same principal investigator using the same stroke prevention guideline throughout the study, which could standardise the information that physicians might deliver to patients.

However, there are several limitations in this study. Firstly, study subjects were retrieved according to the ICPC code, and there was a possibility of missing small number of cases if the diagnosis of AF was not coded. Secondly, this study lack generalisability as it was carried out in two public primary care clinics. Warfarin is another anticoagulant eligible for AF patients. However, we do not keep warfarin patients in our GOPC as they were mainly followed up in SOPD. Hence, there was no warfarin user among our subjects. Finally, this study just determined the associated factors but not causative factors of NOAC refusal. Further research is needed to identify major reasons of NOAC refusal, and strategy to improve guideline attainment should be developed and implemented.

Conclusion

With the increasing AF prevalence in our aging population, it is important to identify the barriers of NOAC use in AF patients. By employing a quantitative design to investigate the NOAC utilisation rate, we found that NOAC use in the two studied GOPC groups of AF patients was 54%, which still leaves room for improvement. This study identified four associated factors of NOAC refusal: older age, higher financial strain score, lack of sponsor for NOAC and lower CHA2DS2-VASc score. The result highlighted the importance of financial support to promote the usage of the NOACs. Thus, future resources should be focused on this high-risk group in order to reduce their stroke risk and the subsequent financial burden due to rehabilitation and hospitalisation. Further research should focus on the subgroup of patients associated with NOAC refusal, and strategy to improve guideline attainment should be developed and implemented.

Acknowledgement

The author would like to thank our district coordinator doctor and our clinic in-charge doctor for their advice and supports on this study and the AFC arrangement. Secondly, I would like to express my wholehearted gratitude to the research committee members for their valuable suggestion on my study design. In addition, thanks all senior doctors in my clinic for teaching me the proper writing of a research paper.

Funding/support

This research received no specific grant from any funding agency in the public, commercial, or not-forprofit sectors.

Conflict of interest

All authors have disclosed no conflicts of interest.

References