The treatment pattern and adherence to direct oral anticoagulants in patients with atrial fibrillation aged over 65

The treatment pattern and adherence to direct oral anticoagulants in patients with atrial... a1111111111 a1111111111 Objective In this study, we aimed to assess the utilization pattern (potentially inappropriate dosing and concomitant use of contraindicated drugs) and adherence to direct oral anticoagulants OPENACCESS (DOACs), including apixaban, dabigatran, and rivaroxaban, in patients with atrial fibrillation Citation: Han S, Jeong HS, Kim H, Suh HS (2019) (AF) unsuitable for warfarin. The treatment pattern and adherence to direct oral anticoagulants in patients with atrial fibrillation Methods aged over 65. PLoS ONE 14(4): e0214666. https:// doi.org/10.1371/journal.pone.0214666 We used nationally representative data, namely Health Insurance Review and Assessment Editor: Hugo ten Cate, Maastricht University Service-Aged Patient Sample 2014, that included medical and pharmacy claims of approxi- Medical Center, NETHERLANDS mately 1 million patients aged 65 or older. We included patients who had at least one diag- Received: October 6, 2018 nosis of AF and at least one prescription of DOAC between January 1 and December 31, 2014. In 2014, DOACs were reimbursed only to patients with AF unsuitable for warfarin. Accepted: March 18, 2019 Appropriate dosing and contraindicated drugs were determined according to the Summary Published: April 1, 2019 of Product Characteristics for each DOAC. Multivariate logistic regression was performed to Copyright:© 2019 Han et al. This is an open examine the factors contributing to the concomitant use of contraindicated drugs. To assess access article distributed under the terms of the adherence, we calculated the medication possession ratio (MPR). Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original Results author and source are credited. The percentage of inappropriate dosing was 11.8% among 1,234 patients with AF; it was Data Availability Statement: Data for these the highest in rivaroxaban users (16.8%). Contraindicated drugs were prescribed to 236 analyses were made available to the authors patients (19.1%). Clinics, smaller healthcare institutions, and outpatient visits were signifi- through the Health Insurance Review and cantly related to contraindicated drug use. The mean MPRs were 0.95, 0.93, and 0.91 for Assessment Service (HIRA), and are available upon request to researchers for the purpose of research. apixaban, dabigatran, and rivaroxaban, respectively (P = 0.075). The authors are not allowed to provide the raw data to unauthorized persons who are not involved in Conclusions this study. Other researchers who are interested in using HIRA-Aged Patient Sample (HIRA-APS) data Careful monitoring is warranted in patients with AF aged over 65 who were unsuitable for are able to obtain the these data through a formal warfarin to reduce the incidence of inappropriate dosing and concomitant use of contraindi- application for HIRA-APS 2014 to the HIRA (http:// cated drugs. opendata.hira.or.kr/op/opc/ selectPatDataAplInfoView.do). HIRA evaluates data PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 1 / 13 Drug utilization in older patients with AF access queries according to whether they are Introduction suitable queries for research purposes. The Atrial fibrillation (AF) is one of the risk factors of stroke, where 15–20% of stroke events are authors confirm they did not have any special related to AF [1, 2]. Especially, older patients with AF constitute almost half of all patients with access privileges to the data that other researchers would not have. AF and are known to have increased risk of stroke and bleeding [3]. As a previous study reported, doses of oral anticoagulants (OACs), concomitant drugs, comorbidities, and adher- Funding: This research was supported by the Basic ence largely affect the anticoagulant effect of OACs in older patients with AF [3]. Science Research Program through the National Research Foundation of Korea (NRF) funded by the Warfarin has been widely used for prevention of stroke in patients with AF. Based on the Ministry of Education, Science and Technology results of several clinical trials, direct oral anticoagulants (DOACs) were also approved for the (NRF-2016R1C1B1009198). same purpose [4–7]. Compared to warfarin, DOACs have fewer drug interactions and lower Competing interests: The authors have declared risk of intracranial hemorrhage. In addition, it does not require regular blood tests [8]. that no competing interests exist. Until July 2015, DOACs were reimbursed by the National Health Insurance of Korea only to patients with AF who were at high risk of development of stroke (i.e., CHA DS -VASc score 2 2 of two or more) and unsuitable for warfarin. Although DOACs have been used in Korea for almost 6 years, little is known about the real-world use, particularly the drug utilization pat- terns and medication adherence, in Korea, which is important to develop strategies for optimal stroke prevention. Therefore, in this study, we aimed to assess the drug utilization patterns and medication adherence to DOACs among patients with AF aged over 65 who were unsuitable for warfarin in a real-world setting. Drug utilization patterns included potentially inappropriate dosing, switching, and concomitant use of contraindicated drugs. Methods Data source We used nationally representative data, namely Health Insurance Review and Assessment Ser- vice-Aged Patient Sample (HIRA-APS) 2014 in this study. This data included medical and pharmacy claims of approximately 1 million patients aged 65 or older, who represented 20% of all Korean patients aged over 65 [9]. All patient data in HIRA-APS were fully anonymized to ensure privacy. Data included the following variables: age, sex, insurance information, diagno- sis codes, prescriptions and medical institution information. Diagnoses were coded according to the International Classification of Diseases, Tenth Revision (ICD-10) [9]. The Pusan National University Institutional Review Board found that this study was exempt from ethical review (PNU IRB/2016_103_HR). Study subjects We included patients who had at least one diagnosis of AF (ICD-10 code, I48.0) and at least one prescription of DOACs (apixaban, dabigatran, or rivaroxaban) during the study period between January 1, 2014 and December 31, 2014. For reference, in 2014 in Korea, 2006–2008 version of ICD-10 code was used. In this version of ICD-10, I48 included only I48.0 (atrial fibrillation) and I48.1 (atrial flutter), but not as many subcodes as recent version of ICD-10 does. The type of the first prescribed DOAC was defined as the index DOAC. The index date was defined as the date of the first DOAC prescription with AF diagnosis code. Medication codes used in this study are listed in S1 Table. Included patients were patients with AF unsuitable for warfarin due to poor international normalized ratio (INR) control, hypersensitivity, or contra- indication to warfarin, according to the Korean reimbursement criteria for DOACs in 2014. We excluded patients who had two or more DOACs at the index date. We followed up the patients from the index date until one of the following three occurrences, whichever comes PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 2 / 13 Drug utilization in older patients with AF first: (1) the first occurrence of switching, (2) the last date of DOAC prescription plus days of supply for the last prescription, or (3) the study end date (December 31, 2014). We identified the age, sex, insurance type, level of institution, bed size and region of institution, and type of visit (inpatient or outpatient) at the index date as the baseline characteristics of the study sub- jects. We also calculated the CHA DS -VASc score, AnTicoagulation and Risk Factors in 2 2 Atrial Fibrillation score (ATRIA score), and Charlson Comorbidity Index (CCI) by using ICD-10 codes at the index date (S2, S3 and S4 Tables). Drug utilization patterns We assessed potentially inappropriate dosing, switching, and concomitant use of contraindi- cated drugs. To assess potentially inappropriate dosing, we used the daily doses regardless of the frequency of dosing per day. Then, we divided the patients into two groups to define them as having potentially inappropriate dosing, (1) patients with DOAC dosage higher than stan- dard dose; and (2) patients with DOAC dosage lower than recommended low dose. The stan- dard dose was defined as the maintenance dose recommended for patients with AF and the recommended low dose was defined as the reduced dose recommended for patients with AF whose health status (i.e., weight, age, and kidney function) was not appropriate for receiving standard dose (S5 Table) [10–13]. We also identified patients with total hip or knee replace- ment, deep venous thrombosis or pulmonary embolism by using Korean procedure codes for hip and knee replacement surgery and ICD-10 codes of deep venous thrombosis and pulmo- nary embolism (S6 Table), because their anticoagulation strategies were different from those in patients with AF. Thus, we ignored the following doses to make them not to be considered as any potentially inappropriate dosing: (1) doses for patients with hip and knee replacement surgery (until 5 and 2 weeks after hip and knee replacement surgery, respectively); (2) doses for patients with acute deep venous thrombosis or pulmonary embolism (until 3 weeks after its occurrence). In addition, if patients received doses other than the recommended doses for� 7 days, these doses were not considered as a potentially inappropriate dosing, because we assumed that this was a dose adjustment period. Switching was defined as initiation of another DOAC and its use for� 8 days to minimize the possibility that a short-term overlapping prescription period for switching would be con- sidered as a concomitant use. Contraindicated drugs were defined according to the Summary of Product Characteristics (SmPC) for each DOAC. The medication codes of contraindicated drugs used in this analysis are found in S7 Table. We excluded unfractionated heparin (UFH) from the analysis because it could be used in combination with DOACs when patients should maintain a patent central venous or arterial catheter. We also excluded systemic ketoconazole (P-glycoprotein [P-gp] inhibitor) from the analysis because it was withdrawn from the Korean market in 2013. For P-gp inducers, they have a substantial impact on the plasma concentra- tions of all DOACs according to the SmPC, they were considered as a contraindication only to dabigatran according to the package insert in Korea. The concomitant use of contraindicated drugs was defined as the filling of a prescription of contraindicated drug during the period in which DOAC was also prescribed. However, the concomitant use of other OACs prescribed for� 7 days was not considered as a concomitant use of contraindicated drugs. Medication adherence Medication adherence to DOACs was assessed by calculating the medication possession ratio (MPR). The MPR was calculated by dividing the number of days of medication supplied within the refill interval by the number of days in the refill interval [14]. Based on the taxon- omy of adherence to medication (i.e., initiation, implementation, and discontinuation) [15], PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 3 / 13 Drug utilization in older patients with AF we evaluated adherence of patients with at least two DOAC prescriptions (i.e., the implementa- tion phase of adherence). If the calculated MPR was� 0.8, it was considered an adherent case. If the calculated MPR was > 1, it was considered 1. If switching occurred, we only included the prescriptions of index DOACs. Statistical analysis The baseline characteristics of the study subjects were expressed as the means ± standard devi- ations for continuous variables and frequency with percentage for categorical variables. The differences in the baseline characteristics and results (drug utilization patterns and adherence) were analyzed using analysis of variance (ANOVA) followed by Bonferroni post-hoc test for continuous variables andχ or Fisher’s exact test for categorical variables. Logistic regression analysis was used to assess the factors associated with the concomitant use of contraindicated drugs. To check the performance of the logistic model, we examined the c-statistic. In all statis- tical analyses, a P-value of < 0.05 was considered statistically significant. Results Study population We identified 40,473 patients with AF, of whom 1,234 patients fulfilled our inclusion criteria for the drug utilization pattern analysis. Fig 1 shows the process of patient selection. The base- line characteristics of the patients stratified according to index DOACs are shown in Table 1. The majority of patients were treated with dabigatran (48.1%), followed by rivaroxaban (43.4%) and apixaban (8.6%). The overall mean age was 76.2 years. Dabigatran users (75.7 years) were significantly younger, compared to rivaroxaban users (76.9 years, P = 0.003). Approximately half of the patients had cerebrovascular disease in the three groups receiving different DOACs. Approximately 28.3–34.0% of the patients had congestive heart failure, and 20.9–30.2% were diagnosed with diabetes without complications. S1 Fig (CHA DS -VASc) 2 2 and S2 Fig (ATRIA) show the distribution of patients receiving each DOAC. There was no sig- nificant difference in predicted stroke risk associated with different DOACs, whereas apixaban and rivaroxaban users had a significantly higher predicted bleeding risk, compared to dabiga- tran users (P = 0.005). Drug utilization patterns S3 Fig shows the number of AF patients who were prescribed DOACs monthly in 2014. Dabi- gatran was the most prescribed DOAC in Korea. There was a slight increase in the number of patients using apixaban and rivaroxaban over 2014. The percentage of potentially inappropriate dosing and switching are shown in Table 2. The percentage of potentially inappropriate dosing was the highest in the rivaroxaban group (16.8%), followed by the dabigatran and apixaban groups (9.3 and 0.9%, respectively; P < 0.001). Among the two types of inappropriate dosing, the use of a dose lower than the rec- ommended low dose was more frequent than the use of a dose higher than standard dose. The percentages of switching for dabigatran, rivaroxaban, and apixaban were 5.4, 2.6, and 0.9%, respectively (P = 0.014). Table 2 shows the percentage of the concomitant use of contraindicated drugs in patients with AF aged over 65 who received DOACs. From a total of 1,234 patients, 236 patients (19.1%) were prescribed contraindicated drugs with an average of 1.89 prescriptions per patient. Among the patients treated with apixaban, 12 patients received concomitant warfarin therapy (11.3%). Among the patients treated with dabigatran, 38 patients (6.4%) were PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 4 / 13 Drug utilization in older patients with AF Fig 1. Flow chart of patient selection. https://doi.org/10.1371/journal.pone.0214666.g001 prescribed concomitant warfarin, and 25 patients (4.2%) were prescribed drugs related to P- glycoprotein. Among the rivaroxaban-treated patients, concomitant use of enoxaparin was the most common (10.7%), followed by warfarin (6.1%). In general, concomitant use of other oral anticoagulants was relatively frequent. The frequency of concomitant use of warfarin with DOACs was not significantly different among different DOACs. In the multivariate logistic model, lower level of medical institution and Chungcheong region were significantly associated with higher odds of concomitant use of contraindicated drugs in patients with AF aged over 65 (Table 3). In addition, outpatient visits were 4.7 times more likely to result in concomitant use of contraindicated drug than inpatient settings. Medication adherence The mean MPR of patients who had at least two DOAC prescriptions was 0.92 (Table 4). The mean MPRs for apixaban, dabigatran, and rivaroxaban were 0.95, 0.93, and 0.91, respectively, which were not significantly different from each other (P = 0.075). Non-adherence (MPR < 0.8) was observed in 16.2%, 11.6%, and 8.3% of rivaroxaban, dabigatran, and apixa- ban users, respectively (P = 0.086; Table 4 and S4 Fig). When the mean MPR values were strati- fied by CHA DS -VASc scores, the mean MPR was the highest in patients with a CHA DS - 2 2 2 2 VASc score of 5. PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 5 / 13 Drug utilization in older patients with AF Table 1. Demographic characteristics of the study population. Total, Apixaban, Dabigatran, Rivaroxaban, P value n (%) n (%) n (%) n (%) Overall, n 1234 106 593 535 Sex < 0.001 Female 615 (49.8) 42 (39.6) 273 (46.0) 300 (56.1) Male 619 (50.2) 64 (60.4) 320 (54.0) 235 (43.9) Age, mean ± SD 76.2 ± 6.09 75.8 ± 5.98 75.7 ± 5.93 76.9 ± 6.23 0.003 Insurance type 0.107 NHI 1108 (89.8) 95 (89.6) 533 (89.9) 480 (89.7) Medical Aid 104 (8.4) 11 (10.4) 44 (7.4) 49 (9.2) PVI 22 (1.8) 0 (0) 16 (2.7) 6 (1.1) Level of institution < 0.001 Clinic 49 (4.0) 4 (3.8) 28 (4.7) 17 (3.2) Long-term care hospital 10 (0.8) 0 (0) 8 (1.3) 2 (0.4) Hospital 105 (8.5) 0 (0) 30 (5.1) 75 (14.0) General hospital 629 (41.9) 33 (31.1) 278 (46.9) 205 (38.3) Specialized general hospital 556 (44.9) 69 (65.1) 249 (42.0) 236 (44.1) Bed size < 0.001 � 100 66 (5.3) 4 (3.8) 33 (5.6) 29 (5.4) 101–300 168 (13.6) 2 (1.9) 56 (9.4) 110 (20.6) 301–800 507 (41.1) 40 (37.7) 282 (47.6) 185 (34.6) 801–1000 235 (19.0) 29 (27.4) 118 (19.9) 88 (16.4) >1,000 258 (20.9) 31 (29.2) 104 (17.5) 123 (23.0) Region of institution 0.001 Seoul� Gyeonggi 616 (49.9) 68 (64.2) 290 (48.9) 258 (48.2) Gangwon 45 (3.6) 4 (3.8) 26 (4.4) 15 (2.8) Chungcheong 118 (9.6) 8 (7.5) 63 (10.6) 47 (8.8) Gyeongsang 237 (19.2) 18 (17.0) 96 (16.2) 123 (23.0) Jeolla 205 (16.6) 8 (7.5) 115 (19.4) 82 (15.3) Jeju 13 (1.1) 0 (0) 3 (0.5) 10 (1.9) Type of hospital visit < 0.001 Inpatient 257 (20.8) 14 (13.2) 80 (13.5) 163 (30.5) Outpatient 977 (79.2) 92 (86.8) 513 (86.5) 372 (69.5) Comorbidities Cerebrovascular disease 679 (55.0) 58 (54.7) 382 (64.4) 239 (44.7) < 0.001 Congestive heart failure 356 (28.8) 36 (34.0) 168 (28.3) 152 (28.4) 0.478 Chronic pulmonary disease 141 (11.4) 6 (5.7) 55 (9.3) 80 (15.0) 0.02 Dementia 174 (14.1) 11 (10.4) 86 (14.5) 77 (14.4) 0.514 Diabetes without chronic complication 275 (22.3) 32 (30.2) 131 (22.1) 112 (20.9) 0.111 Mild liver disease 89 (7.2) 2 (1.9) 40 (6.7) 47 (8.8) 0.036 Myocardial infection 30 (2.4) 1 (0.9) 11 (1.9) 18 (3.4) 0.151 Peripheral vascular disease 61 (4.9) 7 (6.6) 21 (3.5) 33 (6.2) 0.090 Peptic ulcer disease 93 (7.5) 7 (6.6) 43 (7.3) 43 (8.0) 0.821 Rheumatologic disease 11 (0.9) 2 (1.9) 3 (0.5) 6 (1.1) 0.211 Diabetes with chronic complication 66 (5.3) 9 (8.5) 31 (5.2) 26 (4.9) 0.311 Hemiplegia or paraplegia 56 (4.5) 3 (2.8) 36 (6.1) 17 (3.2) 0.045 Any malignancy, including leukemia and lymphoma 55 (4.5) 4 (3.8) 17 (2.9) 34 (6.4) 0.017 Renal disease 26 (2.1) 8 (7.5) 8 (1.3) 10 (1.9) < 0.001 (Continued ) PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 6 / 13 Drug utilization in older patients with AF Table 1. (Continued ) Total, Apixaban, Dabigatran, Rivaroxaban, P value n (%) n (%) n (%) n (%) Moderate or severe liver disease 2 (0.2) 0 (0) 1 (0.2) 1 (0.2) 1.000 AIDS/HIV 2 (0.2) 0 (0) 0 (0) 2 (0.4) 0.352 Metastatic solid tumor 6 (0.5) 2 (1.9) 0 (0) 4 (0.7) 0.014 Risk score, mean ± SD CHA DS -VASc 4.51 ± 1.33 4.55 ± 1.35 4.51 ± 1.25 4.50 ± 1.42 0.956 2 2 � c ATRIA 2.21 ± 1.53 2.53 ± 1.69 2.08 ± 1.42 2.30 ± 1.59 0.005 CCI 1.91 ± 1.56 2.09 ± 2.15 1.90 ± 1.38 1.91 ± 1.61 0.483 SD, standard deviation; PVI, patriots & veterans insurance; CHA DS -VASc, score based on congestive heart failure, hypertension, age� 75 years, diabetes mellitus, 2 2 stroke, vascular disease, age 65–74 years, and sex category; ATRIA, AnTicoagulation and Risk Factors In Atrial Fibrillation; CCI, Charlson Comorbidity Index. Statistically significant, P < 0.05 P = 0.002 for dabigatran versus rivaroxaban, P > 0.05 for apixaban versus dabigatran, and P > 0.05 for apixaban versus rivaroxaban based on Bonferroni post-hoc test Fisher’s exact test was used P = 0.017 for apixaban versus dabigatran, P = 0.049 for dabigatran versus rivaroxaban, and P > 0.05 for apixaban versus rivaroxaban based on Bonferroni post-hoc test. https://doi.org/10.1371/journal.pone.0214666.t001 Discussion This study examined the drug utilization patterns (potentially inappropriate dosing, switching, and concomitant use of contraindicated drugs) and medication adherence to DOACs in patients with AF aged over 65 in Korea. To the best of our knowledge, this is the first study to evaluate the real-world use of apixaban, dabigatran, and rivaroxaban in Korea. First, we found that in Korea, the most commonly prescribed DOAC among patients with AF aged over 65 who were unsuitable for warfarin during the whole study period was dabigatran. Second, 16.8% of the patients treated with rivaroxaban received potentially inappropriate dose (higher than the standard doses or lower than the recommended low doses), followed by dabigatran and apixaban (9.3 and 0.9%, respectively). Third, switching was the most frequent in the dabi- gatran group, whereas it occurred in 2.6 and 0.9% of the patients treated with rivaroxaban and apixaban, respectively. These differences in switching frequency might be attributed to the dif- ferent launch time of each DOAC or other issues, such as tolerability, which needs further investigation. Fourth, concomitant use of other anticoagulants and enoxaparin was relatively frequent. Lower level of medical institution, Chungcheong region, and outpatient visits were significantly associated with concomitant use of contraindicated drugs. Fifth, medication adherence measured as the mean MPR was relatively high in all DOAC groups, which were not significantly different from each other (0.95, 0.93, and 0.91 for apixaban, dabigatran, and rivaroxaban, respectively). Dabigatran was the most commonly prescribed DOAC among Korean patients with AF aged over 65 during the whole study period, followed by rivaroxaban and apixaban. This trend was also observed by Olesen et al., the Danish nationwide administrative registries study, where dabigatran was the first launched DOAC in Denmark [16]. In Korea, rivaroxaban was launched in July 2009, followed by dabigatran in June 2011, apixaban in April 2013, and edox- aban in February 2016. The market share of DOACs might be owing to the launch time, pre- scription preference, or characteristics of treatments and included patients. Because this study only included patients with AF unsuitable for warfarin, these inclusion criteria might have affected the prescription trend. In addition, at the time of this study, edoxaban was not avail- able in the dataset that was used in this study. Further investigations with longer follow-up periods are needed to understand the prescription trend and impact of the market entry of edoxaban in 2015. PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 7 / 13 Drug utilization in older patients with AF Table 2. Percentage of potentially inappropriate dosing, switching, and concomitant use of contraindicated drugs of direct oral anticoagulants in patients with atrial fibrillation aged over 65. Apixaban, Dabigatran, Rivaroxaban, P value n (%) n (%) n (%) Overall, n 106 593 535 Potentially inappropriate dosing and switching Dose > Standard dose 0 (0) 11 (1.9) 18 (3.4) 0.061 < Recommended low-dose 1 (0.9) 44 (7.4) 72 (13.5) < 0.001 Total 1 (0.9) 55 (9.3) 90 (16.8) < 0.001 Switch 1 (0.9) 32 (5.4) 14 (2.6) 0.014 Concomitant use of contraindicated drugs LMWH Enoxaparin 1 (0.9) 21 (3.5) 57 (10.7) < 0.001 Dalteparin 0 (0) 1 (0.2) 0 (0) 0.582 Heparin derivative Fondaparinux 0 (0) 0 (0) 0 (0) - OAC Warfarin 12 (11.3) 38 (6.4) 33 (6.1) 0.140 Apixaban - 9 (1.5) 11 (2.1) 0.298 Dabigatran 0 (0) - 2 (0.4) 0.270 Rivaroxaban 1 (0.9) 25 (4.2) - < 0.001 Potent inhibitor of P-gp Cyclosporine - 1 (0.2) - - Itraconazole - 5 (0.8) - - Dronedarone - 5 (0.8) - - Inducer of P-gp Rifampicin - 5 (0.8) - - St. John’s wort - 0 (0) - - Carmazepine - 6 (1.0) - - Phenytoin - 3 (0.5) - - LMWH, Low molecular weight heparin; OAC, Oral anticoagulant; P-gp, P-glycoprotein. Using DOACs at least 8 days at doses which were not recommended in the Summary of Product Characteristics Using another DOACs rather than the DOAC prescribed at the index date for at least 8 days Concomitant use of contraindicated drugs for least 1 day were included in this analysis, except for the OAC category (concomitant use of OACs for at least 8 days was included) Contraindicated drugs with all DOACs Contraindicated drugs with dabigatran only. https://doi.org/10.1371/journal.pone.0214666.t002 It should be noted that understanding of the reimbursement criteria is important to inter- pret the study findings. DOACs were not covered for all AF patients until June 30, 2015 in Korea. They were only reimbursed for patients with AF unsuitable for warfarin due to poor international normalized ratio (INR) control, hypersensitivity or contraindications to warfa- rin. DOACs started to be reimbursed as the first-line therapy for AF patients from July 1, 2015. Since the available dataset only included claims in 2014, the study findings explain the use of DOACs in patients with AF aged over 65 who were unsuitable for warfarin. Thus, caution is needed when interpreting the study results; in addition, it is not appropriate to extend the study findings to all DOACs users. PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 8 / 13 Drug utilization in older patients with AF Table 3. Factors associated with concomitant use of contraindicated drugs in patients with atrial fibrillation aged over 65 who received direct oral anticoagulants. Factors Odds Ratio (95% CI) Level of institution Tertiary hospital 1 General hospital 1.018 (0.691–1.215) Clinic� Long-term care hospital� Hospital 2.434 (1.215–4.876) Bed size � 100 1 101–300 0.382 (0.207–0.704) 301–800 0.361 (0.178–0.732) 801–1000 0.232 (0.105–0.516) > 1,000 0.298 (0.13–0.684) Region of institution Seoul� Gyeonggi 1 Gangwon 0.723 (0.424–1.232) Chungcheong 1.563 (1.008–2.423) Gyeongsang 1.262 (0.923–1.724) Jeolla 1.142 (0.832–1.568) Jeju 0.776 (0.333–1.811) Type of hospital visit Inpatient 1 Outpatient 4.718 (3.692–6.028) CI, confidence interval. Statistically significant, P < 0.05 c-statistic = 0.798. https://doi.org/10.1371/journal.pone.0214666.t003 Nevertheless, understanding of this special population which is patients with AF unsuitable for warfarin is still important. Because a large proportion of patients with atrial fibrillation (AF) had still received warfarin (44.2% in 2016 in Korea) in the DOAC era, significant number Table 4. Medication adherence to direct oral anticoagulants in patients with atrial fibrillation aged over 65. Total Apixaban Dabigatran Rivaroxaban P value Overall, n 921 72 492 357 MPR for adherence, mean ± SD All 0.92 ± 0.16 0.95 ± 0.12 0.93 ± 0.15 0.91 ± 0.18 0.075 Female 0.92 ± 0.17 0.95 ± 0.14 0.92 ± 0.16 0.91 ± 0.18 0.637 Male 0.92 ± 0.16 0.95 ± 0.11 0.93 ± 0.15 0.90 ± 0.18 0.057 MPR stratified by CHA DS -VASc score, mean ± SD 2 2 < 4 0.90 ± 0.18 0.97 ± 0.11 0.92 ± 0.15 0.86 ± 0.22 0.059 4 0.93 ± 0.15 0.97 ± 0.46 0.93 ± 0.15 0.92 ± 0.16 0.304 5 0.94 ± 0.14 0.91 ± 0.18 0.94 ± 0.12 0.93 ± 0.16 0.526 6 0.91 ± 0.19 0.98 ± 0.05 0.92 ± 0.19 0.89 ± 0.21 0.299 > 6 0.91 ± 0.19 0.91 ± 0.18 0.89 ± 0.21 0.94 ± 0.14 0.695 Non-adherent patients, n (%) 121 (13.1) 6 (8.3) 57 (11.6) 58 (16.2) 0.086 MPR, medication possession ratio; SD, standard deviation. Statistically significant, P < 0.05. https://doi.org/10.1371/journal.pone.0214666.t004 PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 9 / 13 Drug utilization in older patients with AF of patients are expected to get warfarin [17]. As time goes by, these warfarin users could become unsuitable for warfarin (e.g., poor INR control, severe kidney/liver disease, severe hypertension). Thus, understanding issues in this population is important and this study pro- vides an insight into the potentially inappropriate anticoagulation therapy in patients with AF unsuitable for warfarin. For the majority of AF patients included in this study, apixaban, dabigatran, and rivaroxa- ban were appropriately prescribed (99.1%, 90.7%, and 83.1%, respectively). We also found that using lower dose (< the recommended low dose) of DOACs was more frequent than using higher dose (> the standard dose) of DOACs. Using inappropriate low dose of DOACs is related to an increase in stroke risk [18]. According to the SmPC for apixaban, dabigatran, and rivaroxaban, older age, renal insufficiency, lower body weight, and concomitant use of inter- acting medications can increase the blood concentration of DOACs, which tends to increase the bleeding risk. Two previous studies reported that these factors might contribute to the use of inappropriate low doses of DOACs [19, 20]. The risks of stroke and bleeding should be bal- anced out to get the best treatment effects of DOACs. Every drug has both benefits and risks, thus a careful approach is needed to manage each patient. The use of a dose lower than the rec- ommended low dose might increase the risk of stroke, which could endanger the patient. Thus, further research examining the outcomes of using doses lower than the recommended low dose of DOACs is warranted. Switching to a different DOAC was not prevalent as inappropriate dosing of DOACs. This might be due to the length of the follow-up period, which was less than one year. Zhou et al. reported differences in adherence among switchers and non-switchers. Interestingly, they found that the mean MPR values were relatively higher among switchers than among non- switchers [21]. It seemed that switching might be one of the factors associated with adherence to DOACs. However, in this study, we did not compare the adherence between switchers and non-switchers, which might be further investigated. As mentioned above, the concomitant use of contraindicated drugs can affect the bleeding risk or treatment effectiveness by modifying the blood concentration of DOACs [18]. In our analysis, the concomitant use of enoxaparin and warfarin was relatively frequent. Both drugs have a potential to increase the bleeding risk. A previous study found that approximately 7% of dabigatran users were concomitantly receiving low molecular weight heparin [22]. Thus, a medication monitoring system to detect such contraindicated combinations is warranted to reduce the risk of bleeding. Our data indicated that the mean MPR values were similar and high for the three DOACs. The differences in the mean MPR among the three DOACs might be attributed to the differ- ences in the dosing interval and adverse reaction profile [23]. However, since we did not find significant differences in the MPR values among the three DOACs, we concluded that the compliance to the three DOACs was high. This might be attributable to their easy administra- tion (oral formulations) and absence of need for careful monitoring in hospital, unlike warfa- rin. Patients included in this study have already used warfarin, thus they might be aware of the convenience of DOAC administration. Moreover, they might be at high risk since they were unsuitable for warfarin; therefore, they pay more attention to taking their medications. Another study showed that adherent patients had older age and higher CHADS scores [21]. These results are consistent with our findings. This study has several limitations. First, the data used in our study, HIRA-APS, did not include laboratory data (e.g., renal function) or body weight data, which did not allow us to evaluate whether the DOAC dose was appropriately reduced or not. Thus, we used the concept of “lower than the recommended low dose”. For example, <15 mg QD rivaroxaban in patients with AF was considered as “lower than the recommended low dose”. Second, we could not PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 10 / 13 Drug utilization in older patients with AF determine whether the patients received all medications prescribed to them, which is one of the inherent limitations of studies using claims data. Third, the first DOAC prescription date of the study population could be before January 1, 2014. However, we did not identify the first DOAC prescription date to define OAC-naïve users. Because we thought that if we included only OAC-naïve users, it would be difficult to reflect the results of patients unsuitable for war- farin due to poor INR, who are the major components of the study population. In addition, the follow-up period of this study (the mean follow-up period was 194.44 days) may not long enough to evaluate adherence to DOACs. However, the impact of length of follow-up on MPR may not be significant in our study population aged over 65, because it has been reported that adherence tend to remain consistent over time in older patients with AF (especially patients aged over 70) [24]. Conclusions In conclusion, potentially inappropriate dosing of DOACs in terms of dose reduction (lower doses than the recommended low dose) was observed in DOACs users, which cannot be neglected. Medication adherence was generally high among DOACs users. Further research is needed to determine whether this dose reduction led to higher risk of stroke; in addition, stud- ies are required to test whether the high medication adherence lasted for longer follow-up peri- ods and to determine how this might impact the outcomes. Further evaluation of these phenomena in real-world settings will assist clinicians to find the best treatment options for patients who are different from the strictly refined patients included in the clinical trials. Con- comitant use of contraindicated drugs was mainly related to the use of other anticoagulants and enoxaparin. This was significantly associated with smaller healthcare institutions and out- patient visits. Therefore, an alert system of concomitant use of contraindicated drugs needs to be implemented. Supporting information S1 Fig. Distribution of the number of patients with direct oral anticoagulants stratified by CHA DS -VASc scores. 2 2 (DOCX) S2 Fig. Distribution of the number of patients with direct oral anticoagulants stratified by The AnTicoagulation and Risk Factors In Atrial Fibrillation (ATRIA) scores. (DOCX) S3 Fig. Monthly number of patients who were prescribed direct oral anticoagulants. (DOCX) S4 Fig. Proportion of adherent patients (medication possession ratio� 0.80). (DOCX) S1 Table. Medication code of direct oral anticoagulants in Korea. (DOCX) S2 Table. ICD-10 codes for CHA DS -VASc score. 2 2 (DOCX) S3 Table. ICD-10 codes for ATRIA score. (DOCX) S4 Table. ICD-10 codes for charlson comorbidities index. (DOCX) PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 11 / 13 Drug utilization in older patients with AF S5 Table. Indications and recommended doses of direct oral anticoagulants in 2014. (DOCX) S6 Table. Diagnosis and procedure codes used to find patients with deep vein thrombosis, pulmonary embolism, and hip or knee replacement. (DOCX) S7 Table. Medication codes of contraindicated drugs of direct oral anticoagulants. (DOCX) Acknowledgments We used the Health Insurance Review and Assessment Service-Aged Patient Sample 2014 (HIRA-APS-2014-0069); however, we declare that the results do not reflect the positions of either the Health Insurance Review and Assessment Service or the Ministry of Health and Welfare in Korea. Author Contributions Conceptualization: Hwa Seop Jeong, Hae Sun Suh. Data curation: Sola Han, Hwa Seop Jeong, Hyungtae Kim. Formal analysis: Sola Han, Hwa Seop Jeong. Investigation: Sola Han, Hwa Seop Jeong, Hyungtae Kim, Hae Sun Suh. Writing – original draft: Sola Han. Writing – review & editing: Hae Sun Suh. References 1. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framing- ham Study. Stroke. 1991; 22(8):983–8. PMID: 1866765. 2. Ko JS, Kim NH. Anticoagulation in Atrial Fibrillation. Korea J Med. 2011; 81(4):434–43. 3. Sinnaeve P, Brueckmann M, Clemens A, Oldgren J, Eikelboom J, Healey J. Stroke prevention in elderly patients with atrial fibrillation: challenges for anticoagulation. J Intern Med. 2012; 271(1):15–24. https:// doi.org/10.1111/j.1365-2796.2011.02464.x PMID: 21995885 4. Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, et al. Dabigatran versus warfa- rin in patients with atrial fibrillation. N Engl J Med. 2009; 361(12):1139–51. https://doi.org/10.1056/ NEJMoa0905561 PMID: 19717844. 5. Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011; 365(10):883–91. https://doi.org/10.1056/ NEJMoa1009638 PMID: 21830957. 6. Granger CB, Alexander JH, McMurray JJ, Lopes RD, Hylek EM, Hanna M, et al. Apixaban versus war- farin in patients with atrial fibrillation. N Engl J Med. 2011; 365(11):981–92. https://doi.org/10.1056/ NEJMoa1107039 PMID: 21870978. 7. Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, et al. Edoxaban versus war- farin in patients with atrial fibrillation. N Engl J Med. 2013; 369(22):2093–104. https://doi.org/10.1056/ NEJMoa1310907 PMID: 24251359. 8. January CT, Wann LS, Alpert JS, Calkins H, Cigarroa JE, Cleveland JC, Jr., et al. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Car- diology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014; 130(23):e199–267. https://doi.org/10.1161/CIR.0000000000000041 PMID: 24682347; PubMed Central PMCID: PMCPMC4676081. PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 12 / 13 Drug utilization in older patients with AF 9. Kim L, Kim J, Kim S. A guide for the utilization of Health Insurance Review and Assessment Service national patient samples. Epidemiol Health. 2014; 36:e2014008. https://doi.org/10.4178/epih/e2014008 PMID: 25078381 10. Summary of Product Characteristics (INN–rivaroxaban). Available at: https://www.ema.europa.eu/ documents/product-information/xarelto-epar-product-information_en.pdf. Accessed 2 Jan 2019. 11. Summary of Product Characteristics (INN–dabigatran etexilate). Available at: https://www.ema.europa. eu/documents/product-information/pradaxa-epar-product-information_en.pdf. Accessed 2 Jan 2019. 12. Summary of Product Characteristics (INN–apixaban). Available at: https://www.ema.europa.eu/ documents/product-information/eliquis-epar-product-information_en.pdf. Accessed 2 Jan 2019. 13. Jung BC, Kim NH, Nam GB, Park HW, On YK, Lee YS, et al. The Korean Heart Rhythm Society’s 2014 statement on antithrombotic therapy for patients with nonvalvular atrial fibrillation: Korean Heart Rhythm Society. Korean Circ J. 2015; 45(1):9–19. https://doi.org/10.4070/kcj.2015.45.1.9 PMID: 25653698 14. Peterson AM, Nau DP, Cramer JA, Benner J, Gwadry-Sridhar F, Nichol M. A checklist for medication compliance and persistence studies using retrospective databases. Value Health. 2007; 10(1):3–12. https://doi.org/10.1111/j.1524-4733.2006.00139.x PMID: 17261111 15. Vrijens B, De Geest S, Hughes DA, Przemyslaw K, Demonceau J, Ruppar T, et al. A new taxonomy for describing and defining adherence to medications. Br J Clin Pharmacol. 2012; 73(5):691–705. https:// doi.org/10.1111/j.1365-2125.2012.04167.x PMID: 22486599 16. Olesen JB, Sorensen R, Hansen ML, Lamberts M, Weeke P, Mikkelsen AP, et al. Non-vitamin K antag- onist oral anticoagulation agents in anticoagulant naive atrial fibrillation patients: Danish nationwide descriptive data 2011–2013. Europace. 2015; 17(2):187–93. https://doi.org/10.1093/europace/euu225 PMID: 25236181. 17. Ko Y-J, Kim S, Park K, Kim M, Yang BR, Kim M-S, et al. Impact of the Health Insurance Coverage Policy on Oral Anticoagulant Prescription among Patients with Atrial Fibrillation in Korea from 2014 to 2016. J Korean Med Sci. 2018; 33(23):e163. https://doi.org/10.3346/jkms.2018.33.e163 PMID: 29853819 18. Reilly PA, Lehr T, Haertter S, Connolly SJ, Yusuf S, Eikelboom JW, et al. The effect of dabigatran plasma concentrations and patient characteristics on the frequency of ischemic stroke and major bleed- ing in atrial fibrillation patients: the RE-LY Trial (Randomized Evaluation of Long-Term Anticoagulation Therapy). J Am Coll Cardiol. 2014; 63(4):321–8. https://doi.org/10.1016/j.jacc.2013.07.104 PMID: 19. Barra ME, Fanikos J, Connors JM, Sylvester KW, Piazza G, Goldhaber SZ. Evaluation of Dose- Reduced Direct Oral Anticoagulant Therapy. Am J Med. 2016; 129(11):1198–204. https://doi.org/10. 1016/j.amjmed.2016.05.041 PMID: 27341955. 20. Howard M, Lipshutz A, Roess B, Hawes E, Deyo Z, Burkhart JI, et al. Identification of risk factors for inappropriate and suboptimal initiation of direct oral anticoagulants. J Thromb Thrombolysis. 2017; 43:149–56. https://doi.org/10.1007/s11239-016-1435-3 PMID: 27757787. 21. Zhou M, Chang HY, Segal JB, Alexander GC, Singh S. Adherence to a Novel Oral Anticoagulant Among Patients with Atrial Fibrillation. J Manag Care Spec Pharm. 2015; 21(11):1054–62. https://doi. org/10.18553/jmcp.2015.21.11.1054 PMID: 26521117. 22. Carter AA, Leblanc K, Woods A, Lowe D. Utilization of Dabigatran for Atrial Fibrillation at 3 Tertiary Care Centres. Can J Hosp Pharm. 2015; 68(5):369–77. PMID: 26478582; PubMed Central PMCID: PMCPMC4605460. 23. Beyer-Westendorf J, Ehlken B, Evers T. Real-world persistence and adherence to oral anticoagulation for stroke risk reduction in patients with atrial fibrillation. Europace. 2016; 18(8):1150–7. https://doi.org/ 10.1093/europace/euv421 PMID: 26830891. 24. Harper P, Pollock D, Stephens M. Dabigatran persistence and adherence in New Zealand: a nationwide retrospective observational study. BMJ Open. 2018; 8(4):e020212. https://doi.org/10.1136/bmjopen- 2017-020212 PMID: 29626048 PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 13 / 13 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLoS ONE Public Library of Science (PLoS) Journal

The treatment pattern and adherence to direct oral anticoagulants in patients with atrial fibrillation aged over 65

PLoS ONE, Volume 14 (4) – Apr 1, 2019

Loading next page...
 
/lp/public-library-of-science-plos-journal/the-treatment-pattern-and-adherence-to-direct-oral-anticoagulants-in-IVXd0GuQTK
Publisher
Public Library of Science (PLoS) Journal
Copyright
Copyright: © 2019 Han 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: Data for these analyses were made available to the authors through the Health Insurance Review and Assessment Service (HIRA), and are available upon request to researchers for the purpose of research. The authors are not allowed to provide the raw data to unauthorized persons who are not involved in this study. Other researchers who are interested in using HIRA-Aged Patient Sample (HIRA-APS) data are able to obtain the these data through a formal application for HIRA-APS 2014 to the HIRA (http://opendata.hira.or.kr/op/opc/selectPatDataAplInfoView.do). HIRA evaluates data access queries according to whether they are suitable queries for research purposes. The authors confirm they did not have any special access privileges to the data that other researchers would not have. Funding: This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (NRF-2016R1C1B1009198). Competing interests: The authors have declared that no competing interests exist.
eISSN
1932-6203
D.O.I.
10.1371/journal.pone.0214666
Publisher site
See Article on Publisher Site

Abstract

a1111111111 a1111111111 Objective In this study, we aimed to assess the utilization pattern (potentially inappropriate dosing and concomitant use of contraindicated drugs) and adherence to direct oral anticoagulants OPENACCESS (DOACs), including apixaban, dabigatran, and rivaroxaban, in patients with atrial fibrillation Citation: Han S, Jeong HS, Kim H, Suh HS (2019) (AF) unsuitable for warfarin. The treatment pattern and adherence to direct oral anticoagulants in patients with atrial fibrillation Methods aged over 65. PLoS ONE 14(4): e0214666. https:// doi.org/10.1371/journal.pone.0214666 We used nationally representative data, namely Health Insurance Review and Assessment Editor: Hugo ten Cate, Maastricht University Service-Aged Patient Sample 2014, that included medical and pharmacy claims of approxi- Medical Center, NETHERLANDS mately 1 million patients aged 65 or older. We included patients who had at least one diag- Received: October 6, 2018 nosis of AF and at least one prescription of DOAC between January 1 and December 31, 2014. In 2014, DOACs were reimbursed only to patients with AF unsuitable for warfarin. Accepted: March 18, 2019 Appropriate dosing and contraindicated drugs were determined according to the Summary Published: April 1, 2019 of Product Characteristics for each DOAC. Multivariate logistic regression was performed to Copyright:© 2019 Han et al. This is an open examine the factors contributing to the concomitant use of contraindicated drugs. To assess access article distributed under the terms of the adherence, we calculated the medication possession ratio (MPR). Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original Results author and source are credited. The percentage of inappropriate dosing was 11.8% among 1,234 patients with AF; it was Data Availability Statement: Data for these the highest in rivaroxaban users (16.8%). Contraindicated drugs were prescribed to 236 analyses were made available to the authors patients (19.1%). Clinics, smaller healthcare institutions, and outpatient visits were signifi- through the Health Insurance Review and cantly related to contraindicated drug use. The mean MPRs were 0.95, 0.93, and 0.91 for Assessment Service (HIRA), and are available upon request to researchers for the purpose of research. apixaban, dabigatran, and rivaroxaban, respectively (P = 0.075). The authors are not allowed to provide the raw data to unauthorized persons who are not involved in Conclusions this study. Other researchers who are interested in using HIRA-Aged Patient Sample (HIRA-APS) data Careful monitoring is warranted in patients with AF aged over 65 who were unsuitable for are able to obtain the these data through a formal warfarin to reduce the incidence of inappropriate dosing and concomitant use of contraindi- application for HIRA-APS 2014 to the HIRA (http:// cated drugs. opendata.hira.or.kr/op/opc/ selectPatDataAplInfoView.do). HIRA evaluates data PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 1 / 13 Drug utilization in older patients with AF access queries according to whether they are Introduction suitable queries for research purposes. The Atrial fibrillation (AF) is one of the risk factors of stroke, where 15–20% of stroke events are authors confirm they did not have any special related to AF [1, 2]. Especially, older patients with AF constitute almost half of all patients with access privileges to the data that other researchers would not have. AF and are known to have increased risk of stroke and bleeding [3]. As a previous study reported, doses of oral anticoagulants (OACs), concomitant drugs, comorbidities, and adher- Funding: This research was supported by the Basic ence largely affect the anticoagulant effect of OACs in older patients with AF [3]. Science Research Program through the National Research Foundation of Korea (NRF) funded by the Warfarin has been widely used for prevention of stroke in patients with AF. Based on the Ministry of Education, Science and Technology results of several clinical trials, direct oral anticoagulants (DOACs) were also approved for the (NRF-2016R1C1B1009198). same purpose [4–7]. Compared to warfarin, DOACs have fewer drug interactions and lower Competing interests: The authors have declared risk of intracranial hemorrhage. In addition, it does not require regular blood tests [8]. that no competing interests exist. Until July 2015, DOACs were reimbursed by the National Health Insurance of Korea only to patients with AF who were at high risk of development of stroke (i.e., CHA DS -VASc score 2 2 of two or more) and unsuitable for warfarin. Although DOACs have been used in Korea for almost 6 years, little is known about the real-world use, particularly the drug utilization pat- terns and medication adherence, in Korea, which is important to develop strategies for optimal stroke prevention. Therefore, in this study, we aimed to assess the drug utilization patterns and medication adherence to DOACs among patients with AF aged over 65 who were unsuitable for warfarin in a real-world setting. Drug utilization patterns included potentially inappropriate dosing, switching, and concomitant use of contraindicated drugs. Methods Data source We used nationally representative data, namely Health Insurance Review and Assessment Ser- vice-Aged Patient Sample (HIRA-APS) 2014 in this study. This data included medical and pharmacy claims of approximately 1 million patients aged 65 or older, who represented 20% of all Korean patients aged over 65 [9]. All patient data in HIRA-APS were fully anonymized to ensure privacy. Data included the following variables: age, sex, insurance information, diagno- sis codes, prescriptions and medical institution information. Diagnoses were coded according to the International Classification of Diseases, Tenth Revision (ICD-10) [9]. The Pusan National University Institutional Review Board found that this study was exempt from ethical review (PNU IRB/2016_103_HR). Study subjects We included patients who had at least one diagnosis of AF (ICD-10 code, I48.0) and at least one prescription of DOACs (apixaban, dabigatran, or rivaroxaban) during the study period between January 1, 2014 and December 31, 2014. For reference, in 2014 in Korea, 2006–2008 version of ICD-10 code was used. In this version of ICD-10, I48 included only I48.0 (atrial fibrillation) and I48.1 (atrial flutter), but not as many subcodes as recent version of ICD-10 does. The type of the first prescribed DOAC was defined as the index DOAC. The index date was defined as the date of the first DOAC prescription with AF diagnosis code. Medication codes used in this study are listed in S1 Table. Included patients were patients with AF unsuitable for warfarin due to poor international normalized ratio (INR) control, hypersensitivity, or contra- indication to warfarin, according to the Korean reimbursement criteria for DOACs in 2014. We excluded patients who had two or more DOACs at the index date. We followed up the patients from the index date until one of the following three occurrences, whichever comes PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 2 / 13 Drug utilization in older patients with AF first: (1) the first occurrence of switching, (2) the last date of DOAC prescription plus days of supply for the last prescription, or (3) the study end date (December 31, 2014). We identified the age, sex, insurance type, level of institution, bed size and region of institution, and type of visit (inpatient or outpatient) at the index date as the baseline characteristics of the study sub- jects. We also calculated the CHA DS -VASc score, AnTicoagulation and Risk Factors in 2 2 Atrial Fibrillation score (ATRIA score), and Charlson Comorbidity Index (CCI) by using ICD-10 codes at the index date (S2, S3 and S4 Tables). Drug utilization patterns We assessed potentially inappropriate dosing, switching, and concomitant use of contraindi- cated drugs. To assess potentially inappropriate dosing, we used the daily doses regardless of the frequency of dosing per day. Then, we divided the patients into two groups to define them as having potentially inappropriate dosing, (1) patients with DOAC dosage higher than stan- dard dose; and (2) patients with DOAC dosage lower than recommended low dose. The stan- dard dose was defined as the maintenance dose recommended for patients with AF and the recommended low dose was defined as the reduced dose recommended for patients with AF whose health status (i.e., weight, age, and kidney function) was not appropriate for receiving standard dose (S5 Table) [10–13]. We also identified patients with total hip or knee replace- ment, deep venous thrombosis or pulmonary embolism by using Korean procedure codes for hip and knee replacement surgery and ICD-10 codes of deep venous thrombosis and pulmo- nary embolism (S6 Table), because their anticoagulation strategies were different from those in patients with AF. Thus, we ignored the following doses to make them not to be considered as any potentially inappropriate dosing: (1) doses for patients with hip and knee replacement surgery (until 5 and 2 weeks after hip and knee replacement surgery, respectively); (2) doses for patients with acute deep venous thrombosis or pulmonary embolism (until 3 weeks after its occurrence). In addition, if patients received doses other than the recommended doses for� 7 days, these doses were not considered as a potentially inappropriate dosing, because we assumed that this was a dose adjustment period. Switching was defined as initiation of another DOAC and its use for� 8 days to minimize the possibility that a short-term overlapping prescription period for switching would be con- sidered as a concomitant use. Contraindicated drugs were defined according to the Summary of Product Characteristics (SmPC) for each DOAC. The medication codes of contraindicated drugs used in this analysis are found in S7 Table. We excluded unfractionated heparin (UFH) from the analysis because it could be used in combination with DOACs when patients should maintain a patent central venous or arterial catheter. We also excluded systemic ketoconazole (P-glycoprotein [P-gp] inhibitor) from the analysis because it was withdrawn from the Korean market in 2013. For P-gp inducers, they have a substantial impact on the plasma concentra- tions of all DOACs according to the SmPC, they were considered as a contraindication only to dabigatran according to the package insert in Korea. The concomitant use of contraindicated drugs was defined as the filling of a prescription of contraindicated drug during the period in which DOAC was also prescribed. However, the concomitant use of other OACs prescribed for� 7 days was not considered as a concomitant use of contraindicated drugs. Medication adherence Medication adherence to DOACs was assessed by calculating the medication possession ratio (MPR). The MPR was calculated by dividing the number of days of medication supplied within the refill interval by the number of days in the refill interval [14]. Based on the taxon- omy of adherence to medication (i.e., initiation, implementation, and discontinuation) [15], PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 3 / 13 Drug utilization in older patients with AF we evaluated adherence of patients with at least two DOAC prescriptions (i.e., the implementa- tion phase of adherence). If the calculated MPR was� 0.8, it was considered an adherent case. If the calculated MPR was > 1, it was considered 1. If switching occurred, we only included the prescriptions of index DOACs. Statistical analysis The baseline characteristics of the study subjects were expressed as the means ± standard devi- ations for continuous variables and frequency with percentage for categorical variables. The differences in the baseline characteristics and results (drug utilization patterns and adherence) were analyzed using analysis of variance (ANOVA) followed by Bonferroni post-hoc test for continuous variables andχ or Fisher’s exact test for categorical variables. Logistic regression analysis was used to assess the factors associated with the concomitant use of contraindicated drugs. To check the performance of the logistic model, we examined the c-statistic. In all statis- tical analyses, a P-value of < 0.05 was considered statistically significant. Results Study population We identified 40,473 patients with AF, of whom 1,234 patients fulfilled our inclusion criteria for the drug utilization pattern analysis. Fig 1 shows the process of patient selection. The base- line characteristics of the patients stratified according to index DOACs are shown in Table 1. The majority of patients were treated with dabigatran (48.1%), followed by rivaroxaban (43.4%) and apixaban (8.6%). The overall mean age was 76.2 years. Dabigatran users (75.7 years) were significantly younger, compared to rivaroxaban users (76.9 years, P = 0.003). Approximately half of the patients had cerebrovascular disease in the three groups receiving different DOACs. Approximately 28.3–34.0% of the patients had congestive heart failure, and 20.9–30.2% were diagnosed with diabetes without complications. S1 Fig (CHA DS -VASc) 2 2 and S2 Fig (ATRIA) show the distribution of patients receiving each DOAC. There was no sig- nificant difference in predicted stroke risk associated with different DOACs, whereas apixaban and rivaroxaban users had a significantly higher predicted bleeding risk, compared to dabiga- tran users (P = 0.005). Drug utilization patterns S3 Fig shows the number of AF patients who were prescribed DOACs monthly in 2014. Dabi- gatran was the most prescribed DOAC in Korea. There was a slight increase in the number of patients using apixaban and rivaroxaban over 2014. The percentage of potentially inappropriate dosing and switching are shown in Table 2. The percentage of potentially inappropriate dosing was the highest in the rivaroxaban group (16.8%), followed by the dabigatran and apixaban groups (9.3 and 0.9%, respectively; P < 0.001). Among the two types of inappropriate dosing, the use of a dose lower than the rec- ommended low dose was more frequent than the use of a dose higher than standard dose. The percentages of switching for dabigatran, rivaroxaban, and apixaban were 5.4, 2.6, and 0.9%, respectively (P = 0.014). Table 2 shows the percentage of the concomitant use of contraindicated drugs in patients with AF aged over 65 who received DOACs. From a total of 1,234 patients, 236 patients (19.1%) were prescribed contraindicated drugs with an average of 1.89 prescriptions per patient. Among the patients treated with apixaban, 12 patients received concomitant warfarin therapy (11.3%). Among the patients treated with dabigatran, 38 patients (6.4%) were PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 4 / 13 Drug utilization in older patients with AF Fig 1. Flow chart of patient selection. https://doi.org/10.1371/journal.pone.0214666.g001 prescribed concomitant warfarin, and 25 patients (4.2%) were prescribed drugs related to P- glycoprotein. Among the rivaroxaban-treated patients, concomitant use of enoxaparin was the most common (10.7%), followed by warfarin (6.1%). In general, concomitant use of other oral anticoagulants was relatively frequent. The frequency of concomitant use of warfarin with DOACs was not significantly different among different DOACs. In the multivariate logistic model, lower level of medical institution and Chungcheong region were significantly associated with higher odds of concomitant use of contraindicated drugs in patients with AF aged over 65 (Table 3). In addition, outpatient visits were 4.7 times more likely to result in concomitant use of contraindicated drug than inpatient settings. Medication adherence The mean MPR of patients who had at least two DOAC prescriptions was 0.92 (Table 4). The mean MPRs for apixaban, dabigatran, and rivaroxaban were 0.95, 0.93, and 0.91, respectively, which were not significantly different from each other (P = 0.075). Non-adherence (MPR < 0.8) was observed in 16.2%, 11.6%, and 8.3% of rivaroxaban, dabigatran, and apixa- ban users, respectively (P = 0.086; Table 4 and S4 Fig). When the mean MPR values were strati- fied by CHA DS -VASc scores, the mean MPR was the highest in patients with a CHA DS - 2 2 2 2 VASc score of 5. PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 5 / 13 Drug utilization in older patients with AF Table 1. Demographic characteristics of the study population. Total, Apixaban, Dabigatran, Rivaroxaban, P value n (%) n (%) n (%) n (%) Overall, n 1234 106 593 535 Sex < 0.001 Female 615 (49.8) 42 (39.6) 273 (46.0) 300 (56.1) Male 619 (50.2) 64 (60.4) 320 (54.0) 235 (43.9) Age, mean ± SD 76.2 ± 6.09 75.8 ± 5.98 75.7 ± 5.93 76.9 ± 6.23 0.003 Insurance type 0.107 NHI 1108 (89.8) 95 (89.6) 533 (89.9) 480 (89.7) Medical Aid 104 (8.4) 11 (10.4) 44 (7.4) 49 (9.2) PVI 22 (1.8) 0 (0) 16 (2.7) 6 (1.1) Level of institution < 0.001 Clinic 49 (4.0) 4 (3.8) 28 (4.7) 17 (3.2) Long-term care hospital 10 (0.8) 0 (0) 8 (1.3) 2 (0.4) Hospital 105 (8.5) 0 (0) 30 (5.1) 75 (14.0) General hospital 629 (41.9) 33 (31.1) 278 (46.9) 205 (38.3) Specialized general hospital 556 (44.9) 69 (65.1) 249 (42.0) 236 (44.1) Bed size < 0.001 � 100 66 (5.3) 4 (3.8) 33 (5.6) 29 (5.4) 101–300 168 (13.6) 2 (1.9) 56 (9.4) 110 (20.6) 301–800 507 (41.1) 40 (37.7) 282 (47.6) 185 (34.6) 801–1000 235 (19.0) 29 (27.4) 118 (19.9) 88 (16.4) >1,000 258 (20.9) 31 (29.2) 104 (17.5) 123 (23.0) Region of institution 0.001 Seoul� Gyeonggi 616 (49.9) 68 (64.2) 290 (48.9) 258 (48.2) Gangwon 45 (3.6) 4 (3.8) 26 (4.4) 15 (2.8) Chungcheong 118 (9.6) 8 (7.5) 63 (10.6) 47 (8.8) Gyeongsang 237 (19.2) 18 (17.0) 96 (16.2) 123 (23.0) Jeolla 205 (16.6) 8 (7.5) 115 (19.4) 82 (15.3) Jeju 13 (1.1) 0 (0) 3 (0.5) 10 (1.9) Type of hospital visit < 0.001 Inpatient 257 (20.8) 14 (13.2) 80 (13.5) 163 (30.5) Outpatient 977 (79.2) 92 (86.8) 513 (86.5) 372 (69.5) Comorbidities Cerebrovascular disease 679 (55.0) 58 (54.7) 382 (64.4) 239 (44.7) < 0.001 Congestive heart failure 356 (28.8) 36 (34.0) 168 (28.3) 152 (28.4) 0.478 Chronic pulmonary disease 141 (11.4) 6 (5.7) 55 (9.3) 80 (15.0) 0.02 Dementia 174 (14.1) 11 (10.4) 86 (14.5) 77 (14.4) 0.514 Diabetes without chronic complication 275 (22.3) 32 (30.2) 131 (22.1) 112 (20.9) 0.111 Mild liver disease 89 (7.2) 2 (1.9) 40 (6.7) 47 (8.8) 0.036 Myocardial infection 30 (2.4) 1 (0.9) 11 (1.9) 18 (3.4) 0.151 Peripheral vascular disease 61 (4.9) 7 (6.6) 21 (3.5) 33 (6.2) 0.090 Peptic ulcer disease 93 (7.5) 7 (6.6) 43 (7.3) 43 (8.0) 0.821 Rheumatologic disease 11 (0.9) 2 (1.9) 3 (0.5) 6 (1.1) 0.211 Diabetes with chronic complication 66 (5.3) 9 (8.5) 31 (5.2) 26 (4.9) 0.311 Hemiplegia or paraplegia 56 (4.5) 3 (2.8) 36 (6.1) 17 (3.2) 0.045 Any malignancy, including leukemia and lymphoma 55 (4.5) 4 (3.8) 17 (2.9) 34 (6.4) 0.017 Renal disease 26 (2.1) 8 (7.5) 8 (1.3) 10 (1.9) < 0.001 (Continued ) PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 6 / 13 Drug utilization in older patients with AF Table 1. (Continued ) Total, Apixaban, Dabigatran, Rivaroxaban, P value n (%) n (%) n (%) n (%) Moderate or severe liver disease 2 (0.2) 0 (0) 1 (0.2) 1 (0.2) 1.000 AIDS/HIV 2 (0.2) 0 (0) 0 (0) 2 (0.4) 0.352 Metastatic solid tumor 6 (0.5) 2 (1.9) 0 (0) 4 (0.7) 0.014 Risk score, mean ± SD CHA DS -VASc 4.51 ± 1.33 4.55 ± 1.35 4.51 ± 1.25 4.50 ± 1.42 0.956 2 2 � c ATRIA 2.21 ± 1.53 2.53 ± 1.69 2.08 ± 1.42 2.30 ± 1.59 0.005 CCI 1.91 ± 1.56 2.09 ± 2.15 1.90 ± 1.38 1.91 ± 1.61 0.483 SD, standard deviation; PVI, patriots & veterans insurance; CHA DS -VASc, score based on congestive heart failure, hypertension, age� 75 years, diabetes mellitus, 2 2 stroke, vascular disease, age 65–74 years, and sex category; ATRIA, AnTicoagulation and Risk Factors In Atrial Fibrillation; CCI, Charlson Comorbidity Index. Statistically significant, P < 0.05 P = 0.002 for dabigatran versus rivaroxaban, P > 0.05 for apixaban versus dabigatran, and P > 0.05 for apixaban versus rivaroxaban based on Bonferroni post-hoc test Fisher’s exact test was used P = 0.017 for apixaban versus dabigatran, P = 0.049 for dabigatran versus rivaroxaban, and P > 0.05 for apixaban versus rivaroxaban based on Bonferroni post-hoc test. https://doi.org/10.1371/journal.pone.0214666.t001 Discussion This study examined the drug utilization patterns (potentially inappropriate dosing, switching, and concomitant use of contraindicated drugs) and medication adherence to DOACs in patients with AF aged over 65 in Korea. To the best of our knowledge, this is the first study to evaluate the real-world use of apixaban, dabigatran, and rivaroxaban in Korea. First, we found that in Korea, the most commonly prescribed DOAC among patients with AF aged over 65 who were unsuitable for warfarin during the whole study period was dabigatran. Second, 16.8% of the patients treated with rivaroxaban received potentially inappropriate dose (higher than the standard doses or lower than the recommended low doses), followed by dabigatran and apixaban (9.3 and 0.9%, respectively). Third, switching was the most frequent in the dabi- gatran group, whereas it occurred in 2.6 and 0.9% of the patients treated with rivaroxaban and apixaban, respectively. These differences in switching frequency might be attributed to the dif- ferent launch time of each DOAC or other issues, such as tolerability, which needs further investigation. Fourth, concomitant use of other anticoagulants and enoxaparin was relatively frequent. Lower level of medical institution, Chungcheong region, and outpatient visits were significantly associated with concomitant use of contraindicated drugs. Fifth, medication adherence measured as the mean MPR was relatively high in all DOAC groups, which were not significantly different from each other (0.95, 0.93, and 0.91 for apixaban, dabigatran, and rivaroxaban, respectively). Dabigatran was the most commonly prescribed DOAC among Korean patients with AF aged over 65 during the whole study period, followed by rivaroxaban and apixaban. This trend was also observed by Olesen et al., the Danish nationwide administrative registries study, where dabigatran was the first launched DOAC in Denmark [16]. In Korea, rivaroxaban was launched in July 2009, followed by dabigatran in June 2011, apixaban in April 2013, and edox- aban in February 2016. The market share of DOACs might be owing to the launch time, pre- scription preference, or characteristics of treatments and included patients. Because this study only included patients with AF unsuitable for warfarin, these inclusion criteria might have affected the prescription trend. In addition, at the time of this study, edoxaban was not avail- able in the dataset that was used in this study. Further investigations with longer follow-up periods are needed to understand the prescription trend and impact of the market entry of edoxaban in 2015. PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 7 / 13 Drug utilization in older patients with AF Table 2. Percentage of potentially inappropriate dosing, switching, and concomitant use of contraindicated drugs of direct oral anticoagulants in patients with atrial fibrillation aged over 65. Apixaban, Dabigatran, Rivaroxaban, P value n (%) n (%) n (%) Overall, n 106 593 535 Potentially inappropriate dosing and switching Dose > Standard dose 0 (0) 11 (1.9) 18 (3.4) 0.061 < Recommended low-dose 1 (0.9) 44 (7.4) 72 (13.5) < 0.001 Total 1 (0.9) 55 (9.3) 90 (16.8) < 0.001 Switch 1 (0.9) 32 (5.4) 14 (2.6) 0.014 Concomitant use of contraindicated drugs LMWH Enoxaparin 1 (0.9) 21 (3.5) 57 (10.7) < 0.001 Dalteparin 0 (0) 1 (0.2) 0 (0) 0.582 Heparin derivative Fondaparinux 0 (0) 0 (0) 0 (0) - OAC Warfarin 12 (11.3) 38 (6.4) 33 (6.1) 0.140 Apixaban - 9 (1.5) 11 (2.1) 0.298 Dabigatran 0 (0) - 2 (0.4) 0.270 Rivaroxaban 1 (0.9) 25 (4.2) - < 0.001 Potent inhibitor of P-gp Cyclosporine - 1 (0.2) - - Itraconazole - 5 (0.8) - - Dronedarone - 5 (0.8) - - Inducer of P-gp Rifampicin - 5 (0.8) - - St. John’s wort - 0 (0) - - Carmazepine - 6 (1.0) - - Phenytoin - 3 (0.5) - - LMWH, Low molecular weight heparin; OAC, Oral anticoagulant; P-gp, P-glycoprotein. Using DOACs at least 8 days at doses which were not recommended in the Summary of Product Characteristics Using another DOACs rather than the DOAC prescribed at the index date for at least 8 days Concomitant use of contraindicated drugs for least 1 day were included in this analysis, except for the OAC category (concomitant use of OACs for at least 8 days was included) Contraindicated drugs with all DOACs Contraindicated drugs with dabigatran only. https://doi.org/10.1371/journal.pone.0214666.t002 It should be noted that understanding of the reimbursement criteria is important to inter- pret the study findings. DOACs were not covered for all AF patients until June 30, 2015 in Korea. They were only reimbursed for patients with AF unsuitable for warfarin due to poor international normalized ratio (INR) control, hypersensitivity or contraindications to warfa- rin. DOACs started to be reimbursed as the first-line therapy for AF patients from July 1, 2015. Since the available dataset only included claims in 2014, the study findings explain the use of DOACs in patients with AF aged over 65 who were unsuitable for warfarin. Thus, caution is needed when interpreting the study results; in addition, it is not appropriate to extend the study findings to all DOACs users. PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 8 / 13 Drug utilization in older patients with AF Table 3. Factors associated with concomitant use of contraindicated drugs in patients with atrial fibrillation aged over 65 who received direct oral anticoagulants. Factors Odds Ratio (95% CI) Level of institution Tertiary hospital 1 General hospital 1.018 (0.691–1.215) Clinic� Long-term care hospital� Hospital 2.434 (1.215–4.876) Bed size � 100 1 101–300 0.382 (0.207–0.704) 301–800 0.361 (0.178–0.732) 801–1000 0.232 (0.105–0.516) > 1,000 0.298 (0.13–0.684) Region of institution Seoul� Gyeonggi 1 Gangwon 0.723 (0.424–1.232) Chungcheong 1.563 (1.008–2.423) Gyeongsang 1.262 (0.923–1.724) Jeolla 1.142 (0.832–1.568) Jeju 0.776 (0.333–1.811) Type of hospital visit Inpatient 1 Outpatient 4.718 (3.692–6.028) CI, confidence interval. Statistically significant, P < 0.05 c-statistic = 0.798. https://doi.org/10.1371/journal.pone.0214666.t003 Nevertheless, understanding of this special population which is patients with AF unsuitable for warfarin is still important. Because a large proportion of patients with atrial fibrillation (AF) had still received warfarin (44.2% in 2016 in Korea) in the DOAC era, significant number Table 4. Medication adherence to direct oral anticoagulants in patients with atrial fibrillation aged over 65. Total Apixaban Dabigatran Rivaroxaban P value Overall, n 921 72 492 357 MPR for adherence, mean ± SD All 0.92 ± 0.16 0.95 ± 0.12 0.93 ± 0.15 0.91 ± 0.18 0.075 Female 0.92 ± 0.17 0.95 ± 0.14 0.92 ± 0.16 0.91 ± 0.18 0.637 Male 0.92 ± 0.16 0.95 ± 0.11 0.93 ± 0.15 0.90 ± 0.18 0.057 MPR stratified by CHA DS -VASc score, mean ± SD 2 2 < 4 0.90 ± 0.18 0.97 ± 0.11 0.92 ± 0.15 0.86 ± 0.22 0.059 4 0.93 ± 0.15 0.97 ± 0.46 0.93 ± 0.15 0.92 ± 0.16 0.304 5 0.94 ± 0.14 0.91 ± 0.18 0.94 ± 0.12 0.93 ± 0.16 0.526 6 0.91 ± 0.19 0.98 ± 0.05 0.92 ± 0.19 0.89 ± 0.21 0.299 > 6 0.91 ± 0.19 0.91 ± 0.18 0.89 ± 0.21 0.94 ± 0.14 0.695 Non-adherent patients, n (%) 121 (13.1) 6 (8.3) 57 (11.6) 58 (16.2) 0.086 MPR, medication possession ratio; SD, standard deviation. Statistically significant, P < 0.05. https://doi.org/10.1371/journal.pone.0214666.t004 PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 9 / 13 Drug utilization in older patients with AF of patients are expected to get warfarin [17]. As time goes by, these warfarin users could become unsuitable for warfarin (e.g., poor INR control, severe kidney/liver disease, severe hypertension). Thus, understanding issues in this population is important and this study pro- vides an insight into the potentially inappropriate anticoagulation therapy in patients with AF unsuitable for warfarin. For the majority of AF patients included in this study, apixaban, dabigatran, and rivaroxa- ban were appropriately prescribed (99.1%, 90.7%, and 83.1%, respectively). We also found that using lower dose (< the recommended low dose) of DOACs was more frequent than using higher dose (> the standard dose) of DOACs. Using inappropriate low dose of DOACs is related to an increase in stroke risk [18]. According to the SmPC for apixaban, dabigatran, and rivaroxaban, older age, renal insufficiency, lower body weight, and concomitant use of inter- acting medications can increase the blood concentration of DOACs, which tends to increase the bleeding risk. Two previous studies reported that these factors might contribute to the use of inappropriate low doses of DOACs [19, 20]. The risks of stroke and bleeding should be bal- anced out to get the best treatment effects of DOACs. Every drug has both benefits and risks, thus a careful approach is needed to manage each patient. The use of a dose lower than the rec- ommended low dose might increase the risk of stroke, which could endanger the patient. Thus, further research examining the outcomes of using doses lower than the recommended low dose of DOACs is warranted. Switching to a different DOAC was not prevalent as inappropriate dosing of DOACs. This might be due to the length of the follow-up period, which was less than one year. Zhou et al. reported differences in adherence among switchers and non-switchers. Interestingly, they found that the mean MPR values were relatively higher among switchers than among non- switchers [21]. It seemed that switching might be one of the factors associated with adherence to DOACs. However, in this study, we did not compare the adherence between switchers and non-switchers, which might be further investigated. As mentioned above, the concomitant use of contraindicated drugs can affect the bleeding risk or treatment effectiveness by modifying the blood concentration of DOACs [18]. In our analysis, the concomitant use of enoxaparin and warfarin was relatively frequent. Both drugs have a potential to increase the bleeding risk. A previous study found that approximately 7% of dabigatran users were concomitantly receiving low molecular weight heparin [22]. Thus, a medication monitoring system to detect such contraindicated combinations is warranted to reduce the risk of bleeding. Our data indicated that the mean MPR values were similar and high for the three DOACs. The differences in the mean MPR among the three DOACs might be attributed to the differ- ences in the dosing interval and adverse reaction profile [23]. However, since we did not find significant differences in the MPR values among the three DOACs, we concluded that the compliance to the three DOACs was high. This might be attributable to their easy administra- tion (oral formulations) and absence of need for careful monitoring in hospital, unlike warfa- rin. Patients included in this study have already used warfarin, thus they might be aware of the convenience of DOAC administration. Moreover, they might be at high risk since they were unsuitable for warfarin; therefore, they pay more attention to taking their medications. Another study showed that adherent patients had older age and higher CHADS scores [21]. These results are consistent with our findings. This study has several limitations. First, the data used in our study, HIRA-APS, did not include laboratory data (e.g., renal function) or body weight data, which did not allow us to evaluate whether the DOAC dose was appropriately reduced or not. Thus, we used the concept of “lower than the recommended low dose”. For example, <15 mg QD rivaroxaban in patients with AF was considered as “lower than the recommended low dose”. Second, we could not PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 10 / 13 Drug utilization in older patients with AF determine whether the patients received all medications prescribed to them, which is one of the inherent limitations of studies using claims data. Third, the first DOAC prescription date of the study population could be before January 1, 2014. However, we did not identify the first DOAC prescription date to define OAC-naïve users. Because we thought that if we included only OAC-naïve users, it would be difficult to reflect the results of patients unsuitable for war- farin due to poor INR, who are the major components of the study population. In addition, the follow-up period of this study (the mean follow-up period was 194.44 days) may not long enough to evaluate adherence to DOACs. However, the impact of length of follow-up on MPR may not be significant in our study population aged over 65, because it has been reported that adherence tend to remain consistent over time in older patients with AF (especially patients aged over 70) [24]. Conclusions In conclusion, potentially inappropriate dosing of DOACs in terms of dose reduction (lower doses than the recommended low dose) was observed in DOACs users, which cannot be neglected. Medication adherence was generally high among DOACs users. Further research is needed to determine whether this dose reduction led to higher risk of stroke; in addition, stud- ies are required to test whether the high medication adherence lasted for longer follow-up peri- ods and to determine how this might impact the outcomes. Further evaluation of these phenomena in real-world settings will assist clinicians to find the best treatment options for patients who are different from the strictly refined patients included in the clinical trials. Con- comitant use of contraindicated drugs was mainly related to the use of other anticoagulants and enoxaparin. This was significantly associated with smaller healthcare institutions and out- patient visits. Therefore, an alert system of concomitant use of contraindicated drugs needs to be implemented. Supporting information S1 Fig. Distribution of the number of patients with direct oral anticoagulants stratified by CHA DS -VASc scores. 2 2 (DOCX) S2 Fig. Distribution of the number of patients with direct oral anticoagulants stratified by The AnTicoagulation and Risk Factors In Atrial Fibrillation (ATRIA) scores. (DOCX) S3 Fig. Monthly number of patients who were prescribed direct oral anticoagulants. (DOCX) S4 Fig. Proportion of adherent patients (medication possession ratio� 0.80). (DOCX) S1 Table. Medication code of direct oral anticoagulants in Korea. (DOCX) S2 Table. ICD-10 codes for CHA DS -VASc score. 2 2 (DOCX) S3 Table. ICD-10 codes for ATRIA score. (DOCX) S4 Table. ICD-10 codes for charlson comorbidities index. (DOCX) PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 11 / 13 Drug utilization in older patients with AF S5 Table. Indications and recommended doses of direct oral anticoagulants in 2014. (DOCX) S6 Table. Diagnosis and procedure codes used to find patients with deep vein thrombosis, pulmonary embolism, and hip or knee replacement. (DOCX) S7 Table. Medication codes of contraindicated drugs of direct oral anticoagulants. (DOCX) Acknowledgments We used the Health Insurance Review and Assessment Service-Aged Patient Sample 2014 (HIRA-APS-2014-0069); however, we declare that the results do not reflect the positions of either the Health Insurance Review and Assessment Service or the Ministry of Health and Welfare in Korea. Author Contributions Conceptualization: Hwa Seop Jeong, Hae Sun Suh. Data curation: Sola Han, Hwa Seop Jeong, Hyungtae Kim. Formal analysis: Sola Han, Hwa Seop Jeong. Investigation: Sola Han, Hwa Seop Jeong, Hyungtae Kim, Hae Sun Suh. Writing – original draft: Sola Han. Writing – review & editing: Hae Sun Suh. References 1. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framing- ham Study. Stroke. 1991; 22(8):983–8. PMID: 1866765. 2. Ko JS, Kim NH. Anticoagulation in Atrial Fibrillation. Korea J Med. 2011; 81(4):434–43. 3. Sinnaeve P, Brueckmann M, Clemens A, Oldgren J, Eikelboom J, Healey J. Stroke prevention in elderly patients with atrial fibrillation: challenges for anticoagulation. J Intern Med. 2012; 271(1):15–24. https:// doi.org/10.1111/j.1365-2796.2011.02464.x PMID: 21995885 4. Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, et al. Dabigatran versus warfa- rin in patients with atrial fibrillation. N Engl J Med. 2009; 361(12):1139–51. https://doi.org/10.1056/ NEJMoa0905561 PMID: 19717844. 5. Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011; 365(10):883–91. https://doi.org/10.1056/ NEJMoa1009638 PMID: 21830957. 6. Granger CB, Alexander JH, McMurray JJ, Lopes RD, Hylek EM, Hanna M, et al. Apixaban versus war- farin in patients with atrial fibrillation. N Engl J Med. 2011; 365(11):981–92. https://doi.org/10.1056/ NEJMoa1107039 PMID: 21870978. 7. Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, et al. Edoxaban versus war- farin in patients with atrial fibrillation. N Engl J Med. 2013; 369(22):2093–104. https://doi.org/10.1056/ NEJMoa1310907 PMID: 24251359. 8. January CT, Wann LS, Alpert JS, Calkins H, Cigarroa JE, Cleveland JC, Jr., et al. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Car- diology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014; 130(23):e199–267. https://doi.org/10.1161/CIR.0000000000000041 PMID: 24682347; PubMed Central PMCID: PMCPMC4676081. PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 12 / 13 Drug utilization in older patients with AF 9. Kim L, Kim J, Kim S. A guide for the utilization of Health Insurance Review and Assessment Service national patient samples. Epidemiol Health. 2014; 36:e2014008. https://doi.org/10.4178/epih/e2014008 PMID: 25078381 10. Summary of Product Characteristics (INN–rivaroxaban). Available at: https://www.ema.europa.eu/ documents/product-information/xarelto-epar-product-information_en.pdf. Accessed 2 Jan 2019. 11. Summary of Product Characteristics (INN–dabigatran etexilate). Available at: https://www.ema.europa. eu/documents/product-information/pradaxa-epar-product-information_en.pdf. Accessed 2 Jan 2019. 12. Summary of Product Characteristics (INN–apixaban). Available at: https://www.ema.europa.eu/ documents/product-information/eliquis-epar-product-information_en.pdf. Accessed 2 Jan 2019. 13. Jung BC, Kim NH, Nam GB, Park HW, On YK, Lee YS, et al. The Korean Heart Rhythm Society’s 2014 statement on antithrombotic therapy for patients with nonvalvular atrial fibrillation: Korean Heart Rhythm Society. Korean Circ J. 2015; 45(1):9–19. https://doi.org/10.4070/kcj.2015.45.1.9 PMID: 25653698 14. Peterson AM, Nau DP, Cramer JA, Benner J, Gwadry-Sridhar F, Nichol M. A checklist for medication compliance and persistence studies using retrospective databases. Value Health. 2007; 10(1):3–12. https://doi.org/10.1111/j.1524-4733.2006.00139.x PMID: 17261111 15. Vrijens B, De Geest S, Hughes DA, Przemyslaw K, Demonceau J, Ruppar T, et al. A new taxonomy for describing and defining adherence to medications. Br J Clin Pharmacol. 2012; 73(5):691–705. https:// doi.org/10.1111/j.1365-2125.2012.04167.x PMID: 22486599 16. Olesen JB, Sorensen R, Hansen ML, Lamberts M, Weeke P, Mikkelsen AP, et al. Non-vitamin K antag- onist oral anticoagulation agents in anticoagulant naive atrial fibrillation patients: Danish nationwide descriptive data 2011–2013. Europace. 2015; 17(2):187–93. https://doi.org/10.1093/europace/euu225 PMID: 25236181. 17. Ko Y-J, Kim S, Park K, Kim M, Yang BR, Kim M-S, et al. Impact of the Health Insurance Coverage Policy on Oral Anticoagulant Prescription among Patients with Atrial Fibrillation in Korea from 2014 to 2016. J Korean Med Sci. 2018; 33(23):e163. https://doi.org/10.3346/jkms.2018.33.e163 PMID: 29853819 18. Reilly PA, Lehr T, Haertter S, Connolly SJ, Yusuf S, Eikelboom JW, et al. The effect of dabigatran plasma concentrations and patient characteristics on the frequency of ischemic stroke and major bleed- ing in atrial fibrillation patients: the RE-LY Trial (Randomized Evaluation of Long-Term Anticoagulation Therapy). J Am Coll Cardiol. 2014; 63(4):321–8. https://doi.org/10.1016/j.jacc.2013.07.104 PMID: 19. Barra ME, Fanikos J, Connors JM, Sylvester KW, Piazza G, Goldhaber SZ. Evaluation of Dose- Reduced Direct Oral Anticoagulant Therapy. Am J Med. 2016; 129(11):1198–204. https://doi.org/10. 1016/j.amjmed.2016.05.041 PMID: 27341955. 20. Howard M, Lipshutz A, Roess B, Hawes E, Deyo Z, Burkhart JI, et al. Identification of risk factors for inappropriate and suboptimal initiation of direct oral anticoagulants. J Thromb Thrombolysis. 2017; 43:149–56. https://doi.org/10.1007/s11239-016-1435-3 PMID: 27757787. 21. Zhou M, Chang HY, Segal JB, Alexander GC, Singh S. Adherence to a Novel Oral Anticoagulant Among Patients with Atrial Fibrillation. J Manag Care Spec Pharm. 2015; 21(11):1054–62. https://doi. org/10.18553/jmcp.2015.21.11.1054 PMID: 26521117. 22. Carter AA, Leblanc K, Woods A, Lowe D. Utilization of Dabigatran for Atrial Fibrillation at 3 Tertiary Care Centres. Can J Hosp Pharm. 2015; 68(5):369–77. PMID: 26478582; PubMed Central PMCID: PMCPMC4605460. 23. Beyer-Westendorf J, Ehlken B, Evers T. Real-world persistence and adherence to oral anticoagulation for stroke risk reduction in patients with atrial fibrillation. Europace. 2016; 18(8):1150–7. https://doi.org/ 10.1093/europace/euv421 PMID: 26830891. 24. Harper P, Pollock D, Stephens M. Dabigatran persistence and adherence in New Zealand: a nationwide retrospective observational study. BMJ Open. 2018; 8(4):e020212. https://doi.org/10.1136/bmjopen- 2017-020212 PMID: 29626048 PLOS ONE | https://doi.org/10.1371/journal.pone.0214666 April 1, 2019 13 / 13

Journal

PLoS ONEPublic Library of Science (PLoS) Journal

Published: Apr 1, 2019

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create folders to
organize your research

Export folders, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off