Abstract
Aims
Direct oral anticoagulant (DOAC) is widely used for the prevention of embolic stroke in non-valvular atrial fibrillation (NVAF) patients. However, the gastrointestinal bleeding risk in several DOAC regimens was higher than warfarin, especially in once-daily regimens.
Methods and results
We conducted a single-center prospective registry of patients with NVAF treated with DOACs: the DIRECT registry (N = 2216; follow-up duration 650 [IQR 103–1574] days, UMIN000033283). All patients were divided into 2 groups: the twice-daily (BID) regimen group (dabigatran and apixaban) versus the once-daily (QD) regimen group (rivaroxaban and edoxaban). Out of 2216 patients, we successfully matched 904 patients in the QD group and 904 patients in the BID group using propensity score. The primary endpoint was gastrointestinal bleeding defined as any bleeding in the gastrointestinal tract that was identified through medical records regardless of bleeding site or severity. The BID group showed a significantly lower gastrointestinal bleeding rate than the QD group (3.5/100 person-year vs. 6.2/100 person-year, log-rank P < 0.0001). The secondary endpoints were all death, stroke, major bleeding, and any bleeding. The rate of major bleeding was significantly lower in patients with BID regimen group (log-rank P = 0.040). In contrast, all death, stroke, and any bleeding did not differ between both groups (log-rank P = 0.280, 0.520 and 0.066, respectively).
Conclusions
The BID regimen as compared with the QD regimen was associated with reduced risk of gastrointestinal bleeding.
Keywords: Gastrointestinal bleeding, Direct oral anticoagulants, Atrial fibrillation, BID, QD
Graphical abstract
1. Introduction
In recent era, direct oral anticoagulants (DOACs) are widely used for non-valvular atrial fibrillation (NVAF). Four landmark phase 3 trials have shown that each DOAC was more or equally effective for stroke prevention and also had almost similar safety profile with respect to bleeding events in patients with NVAF compared with vitamin K antagonist (warfarin) [1], [2], [3], [4]. However, in these trials, rates of gastrointestinal bleeding were higher than warfarin in several DOAC regimens, especially in once-daily regimens. The peak level of drug concentration is thought to be associated with bleeding events [5], [6], [7]. The peak level would be higher in the once-daily [QD (quaque die)] regimens than in the twice-daily [BID (bis in die)] regimens. Furthermore, in contrast to warfarin, DOAC has active anticoagulation effect within the lumen of gastrointestinal tract after oral intake [8]. These points may explain the reason why rivaroxaban and edoxaban (both QD regimen) were associated with the higher risk of gastrointestinal bleeding compared with warfarin in the clinical trials [2], [4]. Twice-daily dosing (BID) rather than once-daily dosing (QD) may theoretically suppress the peak level of drug concentration in the gut lumen and blood plasma. The range of drug concentration can be narrower in the BID regimen than in the QD regimen (Fig. 1), which may theoretically result in a better safety profile [9].
Fig. 1.
Theoretical pharmacokinetic profile of a drug.
These schematic graphs illustrate the theoretical pharmacokinetic profile of a drug in serum and gastrointestinal lumen: a dose X administered once-daily (QD, black line), and a dose X/2 administered twice-daily (BID, red line). A) Taking proper regimen. B) Taking an extra tablet may result in a higher peak in the QD than in the BID regimen. C) A single miss of DOAC may deviate the concentration downward from the therapeutic range more severely in the QD regimens than in the BID regimens. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
It remains unclear whether BID regimen shows the theoretical superiority over QD regimen in the real-world clinical data. The purpose of this study was to investigate the impact of once- or twice-daily regimen on the gastrointestinal bleeding in the real-world patients treated with DOACs.
2. Methods
2.1. Study population
We conducted a single-center prospective observational registry of NVAF patients with DOACs: Safety and effectiveness of 4 Different dIrect oRal anticoagulants, dabigatran, rivaroxaban, apixaban and edoxaban in the rEal-world Clinical pracTice: [DIRECT registry (UMIN000033283)]. All serial adult patients (aged ≥18 years) in our institution with NVAF who were users of dabigatran, rivaroxaban, apixaban, or edoxaban from June 2011 to November 2017 were enrolled. If a patient ever used DOAC during the study period, the first fill of DOAC was defined as the index medication. The treatment period was defined as the time from the first administration of a drug to last follow-up or 2 days after the trial drugs were discontinued if the patient stopped taking the medication. In the present study, we divided all patients into 2 groups: BID regimen (dabigatran and apixaban) group and QD regimen (rivaroxaban and edoxaban) group.
2.2. Endpoints
The primary endpoint in the current study was gastrointestinal bleeding. Gastrointestinal bleeding was defined as any bleeding in the gastrointestinal tract that was identified through medical records regardless of bleeding site or severity.
The secondary endpoints were all death, stroke, major bleeding, and any bleeding. Stroke was defined as a neurologic deficit persisting ≥24 h attributed to an acute focal injury of the central nervous system by a vascular cause, including cerebral infarction, intracerebral hemorrhage, and subarachnoid hemorrhage. Major bleeding was defined according to the International Society on Thrombosis and Haemostasis (ISTH) criteria, as clinically overt bleeding accompanied by a decrease in the hemoglobin level of at least 2 g/dL or transfusion of at least 2 units of packed red cells, occurring at a critical site, or resulting in death [10]. Any bleeding was defined as a composite of major bleeding and clinically relevant non-major bleeding according to ISTH criteria. Clinical events were monitored by questioning, physical examination, and laboratory test at each outpatient visit every 2–4 months.
2.3. Statistical analysis
All analyses were performed using R software (version 4.1.1; R Foundation for Statistical Computing, Vienna, Austria). P value of <0.05 was considered statistically significant. Data are presented with listwise deletion. Categorical variables are expressed as counts (percentages) and compared with the chi-squared test or Fisher's exact test. Continuous variables are expressed as mean (standard deviation) or median [interquartile range] and compared using Student's t-test or the Mann–Whitney U test as appropriate. Because excluding missing data cases can cause bias in this analysis and loss of power for detecting a statistical difference, we performed random imputation using the “missForest” method. The method we used for imputation, “missForest”, is a package in the R statistical software. The function missForest is used to impute missing values particularly in the case of mixed-type data [11]. Because dose of DOAC was not randomly selected, potential confounding factors were eliminated through a propensity score-matching technique. Propensity scores for the estimated probability of the BID regimen of DOAC with reference to the QD regimen in each patient were generated using a multiple logistic regression model. The final model included following 14 variables based on the clinical consensus: age, sex, body weight, antiplatelet use, hemoglobin, creatinine, diabetes mellitus, hypertension, dyslipidemia, histories of heart failure, bleeding, stroke, atherosclerosis, liver dysfunction. The area under the receiver operating characteristics (ROC) curve (AUC) was 0.601 (95 % CI: 0.577–0.624) for this model. We successfully matched 904 patients treated with QD regimen and 904 patients with BID regimen. The balance between the groups was assessed with a standardized mean difference (SMD). The SMD within 10 % is considered a negligible imbalance between 2 groups [12]. Outcomes were assessed according to the 2 groups, BID regimen group and QD regimen group, in a time-to-first-event fashion with the Kaplan-Meier method and compared with log-rank test. Impact of the BID regimen as compared to the QD regimen was estimated with Cox proportional hazard model. Subgroup analysis was done in the following subpopulations: elderly patients (age ≥ 75), sex, diabetes mellitus, hypertension, and dyslipidemia.
3. Results
3.1. Subjects
Patient selection flowchart is shown in Fig. 2. A total of 2216 patients [dabigatran (N = 648), apixaban (N = 599), rivaroxaban (N = 538) and edoxaban (N = 431)] were enrolled in the present registry. BID regimen (dabigatran and apixaban) group and QD regimen (rivaroxaban and edoxaban) group consist of 1247 and 969 patients, respectively. Median follow-up duration in the whole population was 650 days (IQR, 103–1574 days). Baseline characteristics of the BID and QD regimen groups are presented in Table 1. We successfully matched 904 patients treated with QD regimen and 904 patients with BID regimen using propensity score. After propensity score matching the baseline characteristics were well-balanced (Table 2).
Fig. 2.
Patient selection flowchart.
*A propensity score was calculated by a logistic regression model with the parameters of age, sex, body weight, antiplatelet use, hemoglobin, creatinine, diabetes mellitus, hypertension, dyslipidemia, a history of heart failure, a history of bleeding, a history of stroke, atherosclerosis, and liver dysfunction. Abbreviations: NVAF = non-valvular atrial fibrillation; DOAC = direct oral anticoagulants; BID = twice-daily dosing; QD = once-daily dosing.
Table 1.
Overall cohort: Patient background.
| Variable | BID (N = 1247) | QD (N = 969) | SMD | Missing (%) |
|---|---|---|---|---|
| Age (years) | 72.00 [65.00, 79.00] | 73.00 [66.00, 80.00] | 0.076 | 0 |
| Body weight (kg) | 60.50 [51.20, 69.45] | 60.00 [50.00, 70.00] | 0.002 | 1.5 |
| Woman | 456 (36.6 %) | 350 (36.1 %) | 0.009 | 0 |
| Hypertension | 903 (72.4 %) | 725 (74.8 %) | 0.055 | 0 |
| Diabetes mellitus | 336 (26.9 %) | 283 (29.2 %) | 0.050 | 0 |
| Dyslipidemia | 820 (65.8 %) | 624 (64.4 %) | 0.030 | 0 |
| Antiplatelets use | 282 (22.6 %) | 195 (20.1 %) | 0.061 | 0 |
| History of heart failure | 289 (23.2 %) | 236 (24.4 %) | 0.028 | 0 |
| History of bleeding | 288 (23.1 %) | 313 (32.3 %) | 0.207 | 0 |
| History of atherosclerosis | 310 (24.9 %) | 256 (26.4 %) | 0.036 | 0 |
| History of stroke | 250 (20.0 %) | 197 (20.3 %) | 0.007 | 0 |
| Liver dysfunction | 355 (28.5 %) | 382 (39.4 %) | 0.233 | 0 |
| Creatinine (mg/dL) | 0.84 [0.69, 1.01] | 0.83 [0.70, 1.00] | 0.047 | 0 |
| Hemoglobin (g/dL) | 13.40 [12.10, 14.80] | 13.20 [11.90, 14.70] | 0.107 | 0 |
Data are expressed as n (%) or median [interquartile range]. BID = twice-daily dosing; QD = once-daily dosing; SMD = standard mean difference.
Table 2.
Patient background after propensity score matching.
| Variable | BID (N = 904) | QD (N = 904) | SMD |
|---|---|---|---|
| Age (years) | 73.00 [66.00, 79.00] | 73.00 [66.00, 80.00] | 0.026 |
| Body weight (kg) | 60.10 [50.86, 69.00] | 60.00 [50.00, 70.00] | 0.011 |
| Woman | 330 (36.5 %) | 330 (36.5 %) | <0.001 |
| Hypertension | 665 (73.6 %) | 669 (74.0 %) | 0.010 |
| Diabetes mellitus | 269 (29.8 %) | 260 (28.8 %) | 0.022 |
| Dyslipidemia | 600 (66.4 %) | 586 (64.8 %) | 0.033 |
| Antiplatelets use | 188 (20.8 %) | 187 (20.7 %) | 0.003 |
| History of heart failure | 209 (23.1 %) | 219 (24.2 %) | 0.026 |
| History of bleeding | 277 (30.6 %) | 269 (29.8 %) | 0.019 |
| History of atherosclerosis | 251 (27.8 %) | 233 (25.8 %) | 0.045 |
| History of stroke | 181 (20.0 %) | 180 (19.9 %) | 0.003 |
| Liver dysfunction | 309 (34.2 %) | 326 (36.4 %) | 0.039 |
| Creatinine (mg/dL) | 0.83 [0.68, 1.00] | 0.83 [0.70, 1.00] | 0.015 |
| Hemoglobin (g/dL) | 13.30 [12.00, 14.60] | 13.30 [12.00, 14.70] | 0.016 |
Data are expressed as n (%) or median [interquartile range]. BID = twice-daily dosing; QD = once-daily dosing; SMD = standard mean difference.
3.2. Clinical endpoints
Fig. 3, Fig. 4 depict Kaplan-Meier curves for the primary and secondary endpoints. Table 3 tabulates the outcomes in the BID and QD regimen groups. Event rate of gastrointestinal bleeding was significantly lower in the BID regimen group than in the QD regimen group (3.5/100 person-years vs. 6.2/100 person-years, log-rank P < 0.0001). Furthermore, the rate of major bleeding was also significantly lower in patients with BID regimen group (log-rank P = 0.040). On the other hand, all death, stroke and any bleeding did not differ between BID and QD regimen groups (log-rank P = 0.280, 0.520 and 0.066). In the overall population, the risk of gastrointestinal bleeding was lower in the BID group (hazard ratio [HR]: 0.556; 95 % CI 0.416–0.743, P < 0.001). The superiority of BID regimen was consistent across various subgroups (Table 4).
Fig. 3.
Kaplan-Meier estimates for gastrointestinal bleeding.
Kaplan-Meier analysis for gastrointestinal bleeding according to QD vs. BID regimens of DOAC in the propensity score-matched population. Patient with BID regimen group less frequently had gastrointestinal bleeding than those with QD regimen group (log-rank P < 0.0001). Abbreviations: GI = gastrointestinal; DOAC = direct oral anticoagulants; BID = twice-daily dosing; QD = once-daily dosing.
Fig. 4.
Kaplan-Meier estimates for secondary endpoints.
Kaplan-Meier analysis for secondary endpoints according to QD vs. BID regiments of DOAC in the propensity score-matched population. There was no significant difference between the BID regimen group and the QD regimen group in each endpoint in all death (A), stroke (B), and any bleeding (D). Patients with the QD regimens experienced major bleeding more frequently than those with the BID regimens (C) (log-rank P = 0.040). Abbreviations: DOAC = direct oral anticoagulants; BID = twice-daily dosing; QD = once-daily dosing.
Table 3.
Outcomes in the BID and QD groups.
| BID (N = 904) |
QD (N = 904) |
P value⁎ | |||||
|---|---|---|---|---|---|---|---|
| Patients with event, n | Follow-up, person-year | Rate/100 person-year | Patients with event, n | Follow-up, person-year | Rate/100 person-year | ||
| Gastrointestinal bleeding | 80 | 2286 | 3.5 | 110 | 1781 | 6.2 | <0.0001 |
| All death | 54 | 2400 | 2.3 | 56 | 1924 | 2.9 | 0.280 |
| Stroke | 36 | 2357 | 1.5 | 27 | 1903 | 1.4 | 0.520 |
| Major bleeding | 63 | 2321 | 2.7 | 74 | 1834 | 4 | 0.040 |
| Any bleeding | 350 | 1651 | 21.2 | 340 | 1342 | 25.3 | 0.066 |
BID = twice-daily dosing; QD = once-daily dosing.
P value for log rank test.
Table 4.
Subgroup analysis of the primary endpoint (gastrointestinal bleeding).
| HR (95 % CI) | P value | P value for interaction | |
|---|---|---|---|
| Overall | 0.556 (0.416, 0.743) | <0.001 | |
| Age | 0.292 | ||
| ≥75 years | 0.482 (0.319, 0.729) | <0.001 | |
| <75 years | 0.618 (0.411, 0.930) | 0.021 | |
| Sex | 0.561 | ||
| Men | 0.581 (0.403, 0.837) | 0.004 | |
| Women | 0.521 (0.323, 0.838) | 0.007 | |
| Diabetes mellitus | 0.548 | ||
| Present | 0.620 (0.381, 1.010) | 0.055 | |
| Absent | 0.520 (0.362, 0.747) | <0.001 | |
| Hypertension | 0.675 | ||
| Present | 0.540 (0.391, 0.745) | <0.001 | |
| Absent | 0.652 (0.334, 1.27) | 0.209 | |
| Dyslipidemia | 0.024 | ||
| Present | 0.684 (0.487, 0.960) | 0.028 | |
| Absent | 0.312 (0.171, 0.567) | <0.001 |
Hazard ratio for gastrointestinal bleeding was assessed in multiple subgroups. BID = twice-daily dosing; QD = once-daily dosing.
4. Discussion
The main findings of this study can be summarized as follows: 1) The BID regimen group had a significantly lower gastrointestinal bleeding risk than the QD regimen group; 2) the BID regimen group showed a lower rate of major bleeding than the QD regimen group; 3) both groups did not have a significant difference in all death, stroke, and any bleeding rates.
In the last several years, DOACs have been approved as a first-line therapy for stroke prevention in NVAF patients. Compared with warfarin, DOACs have rapid onset and offset action, predictable pharmacodynamics obviating regular therapeutic monitoring, and fewer food-drug or drug-drug interactions [8]. However, gastrointestinal bleeding remains major concern of DOAC use.
4.1. Pathophysiology of gastrointestinal bleeding caused by DOAC
The mechanisms of DOAC-related gastrointestinal bleeding could be summarized as following factors: (1) systemic anticoagulant effect; (2) focal anticoagulant effect; and (3) inhibition of mucosal healing and direct caustic effect [5], [8]. The systemic anticoagulant effect is not a specific issue of DOACs, whereas focal anticoagulant effect and focal effect of inhibition of mucosal healing are specific for DOAC. Warfarin is well absorbed and has a bioavailability of over 95 %. Non-absorbed warfarin within the gastrointestinal lumen has no anticoagulant activity. In patients with warfarin, the risk of gastrointestinal bleeding reflects the systemic anticoagulant activity of drug. In contrast, the absorption of DOACs ranges from 7 % to 66 % (Dabigatran; 7 %, Apixaban; 50 %, Edoxaban; 62 %, Rivaroxaban; 66 %) [13] and non-absorbed DOACs within the gastrointestinal lumen has focal anticoagulant effect. Even in healthy volunteers, gastric erosions are noted in 5–10 % [14] and small bowel erosions are noted in 10–15 % [15]. The focal effect of DOACs may promote the gastrointestinal bleeding in these asymptomatic patients. The mucosal healing inhibition of DOACs may directly increase gastrointestinal bleeding. Dabigatran targets the protease, trypsin, while Rivaroxaban and Apixaban target chymotrypsin. Rivaroxaban targets matriptase as well [16]. Trypsin and chymotrypsin, and matriptase are digestive enzymes found within the gastrointestinal tract and their secretion leads to mucosal vulnerability that makes the gastrointestinal tract more prone to gastric bleeding. Through binding interactions with the DOACs, these enzymes become trapped within the gastrointestinal tract and lead to increased mucosal vulnerability and thus increased rates of gastrointestinal bleeding. Dabigatran is the only DOAC to contain an acid and would be the only DOAC to lead to gastrointestinal bleeding through a direct caustic effect [16]. This pathophysiology may explain the reason why DOACs have higher gastrointestinal bleeding risk than warfarin. However, only edoxaban targets neither trypsin nor chymotrypsin [16]. Nevertheless, it showed the higher risk of gastrointestinal bleeding than warfarin [4]. This suggested that the first two factors, systemic anticoagulant effect and focal anticoagulant effect, may play a major role in the gastrointestinal bleeding in patients treated with DOACs. We hypothesized that the BID regimen rather than the QD regimen keeps the systemic anticoagulant effect and focal anticoagulant effect more stable and therefore provides better safety since it keeps drug concentration in the gut lumen and vessels more stable in the narrow range (Fig. 1).
4.2. BID versus QD regimen of DOAC
In the pivotal trials, gastrointestinal bleeding rates of DOACs were higher than those of warfarin especially in QD DOAC regimens: rivaroxaban (20 mg once-daily) [2] and edoxaban (60 mg once-daily) [4]. One of the possible reasons would be their relatively high peak level of plasma drug concentration which is reportedly associated with gastrointestinal bleeding [5], [6], [7]. Change in the focal drug concentration inside the gut lumen would also be similar to the plasma drug concentration. The BID regimen should theoretically minimize daily fluctuation in plasma and focal concentration, consequently lower the peak level of plasma and focal drug concentration. We depicted the theoretical pharmacokinetic profile of a drug: a dose X administered QD, and a dose X/2 administered BID (Fig. 1). Twice daily dosing may be safer than once daily dosing when a patient takes the medicine mistakenly. If a patient takes an extra dose incidentally, drug concentration level easily exceeds safety therapeutic range which could lead gastrointestinal bleeding in the QD regimen group. Alternatively, the concentration level of BID drug regimen could be sustained within the optimal therapeutic range. In contrast to clinical trials, adherence to anticoagulation therapy is one of the big issues in the real-world setting. BID regimen could be beneficial for maintaining the drug concentration in the therapeutic range even in patients with suboptimal adherence. These may explain the results that BID regimen decreased gastrointestinal bleeding risk compared with QD regimen [17]. A large-scale retrospective cohort study (N = 581,451) reported that the QD rivaroxaban was associated with a significantly increased risk of hemorrhagic events compared with the BID apixaban, including gastrointestinal bleeding (35.2/1000 person-years vs. 16.3/1000 person-years; rate difference, 19.0 [95 % CI, 17.9 to 20.1]; hazard ratio, 2.09 [95 % CI, 2.01 to 2.18]) [18]. This result also supports our findings. It is of note that high-dose dabigatran (150 mg twice-daily) was related to high risk of gastrointestinal bleeding compared with warfarin [1]. There seems to be a reason specific for dabigatran. The pellets of dabigatran contain a tartaric acid core coated by dabigatran etexilate. This construction generates an acidic microenvironment, then increases drug dissolution and gastric absorption [19]. However, the tartaric acid in dabigatran etexilate is postulated to cause direct caustic injury [5]. Hence, high-dose dabigatran might have shown a relatively high risk of gastrointestinal bleeding in spite of its BID regimen superiority.
4.3. Study strength and limitation
The present prospective observational registry would have, to date, relatively large Asian cohort, and long follow-up period. Data of the fourth DOAC, edoxaban, was uniquely available due to the geographical reason. A few limitations, however, need to be acknowledged. First, the present study is a single center prospective registry. The institution is located in the urban area in Japan. Generalizability to other regions is limited. Second, there might be unknown and unmeasured relevant factors such as frailty which were not integrated into the adjustment of propensity score matching. The unexpected selection bias could not be totally eliminated in our study setting. Lastly, we could not evaluate the drug adherence and its difference between the BID and QD regimens. Previous studies revealed that patients' adherence was significantly higher for drugs with the QD regimens than the BID regimens [9], [20], [21]. Readers should keep in their mind that the current results reflect a “real-world” clinical data of Asians population and are just hypothesis-generating. The present study suggests the necessity of further prospective large-scale randomized controlled trials.
5. Conclusions
The BID regimen of DOAC as compared with the QD regimen was associated with reduced risk of gastrointestinal bleeding. Further prospective randomized clinical trials are warranted to confirm the superiority of the BID regimen over the QD regimen in the real-world settings.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Y. Sotomi, A. Hirata, Y. Sakata, A. Hirayama, and Y. Higuchi received grants, travel expenses, and speaker honorarium from Daiichi-Sankyo, Bayer, Boehringer Ingelheim, and Bristol-Myers Squibb. The other authors have nothing to disclose.
Acknowledgements
The authors thank Ayaka Murakami, Tomoe Yamamoto, Ayako Fukao, and Yoko Inoue for their invaluable support in data collection and management.
References
- 1.Connolly S.J., Ezekowitz M.D., Yusuf S., Eikelboom J., Oldgren J., Parekh A., et al. Dabigatran versus warfarin in patients with atrial fibrillation. N. Engl. J. Med. 2009;361(12):1139–1151. doi: 10.1056/NEJMoa0905561. [DOI] [PubMed] [Google Scholar]
- 2.Patel M.R., Mahaffey K.W., Garg J., Pan G., Singer D.E., Hacke W., et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N. Engl. J. Med. 2011;365(10):883–891. doi: 10.1056/NEJMoa1009638. [DOI] [PubMed] [Google Scholar]
- 3.Granger C.B., Alexander J.H., McMurray J.J.V., Lopes R.D., Hylek E.M., Hanna M., et al. Apixaban versus warfarin in patients with atrial fibrillation. N. Engl. J. Med. 2011;365(11):981–992. doi: 10.1056/NEJMoa1107039. [DOI] [PubMed] [Google Scholar]
- 4.Giugliano R.P., Ruff C.T., Braunwald E., Murphy S.A., Wiviott S.D., Halperin J.L., et al. Edoxaban versus warfarin in patients with atrial fibrillation. N. Engl. J. Med. 2013;369(22):2093–2104. doi: 10.1056/NEJMoa1310907. [DOI] [PubMed] [Google Scholar]
- 5.Cheung K.S., Leung W.K. Gastrointestinal bleeding in patients on novel oral anticoagulants: risk, prevention and management. World J. Gastroenterol. 2017;23(11):1954–1963. doi: 10.3748/wjg.v23.i11.1954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Testa S., Legnani C., Antonucci E., Paoletti O., Dellanoce C., Cosmi B., et al. Drug levels and bleeding complications in atrial fibrillation patients treated with direct oral anticoagulants. J. Thromb. Haemost. 2019;17(7):1064–1072. doi: 10.1111/jth.14457. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Škorňová I., Samoš M., Bolek T., Kamenišťáková A., Stančiaková L., Galajda P., et al. Direct Oral anticoagulants plasma levels in patients with atrial fibrillation at the time of bleeding: a pilot prospective study. J. Cardiovasc. Pharmacol. 2021 Jul;78(1):e122–e127. doi: 10.1097/FJC.0000000000001038. [DOI] [PubMed] [Google Scholar]
- 8.Desai J., Kolb J.M., Weitz J.I., Aisenberg J. Gastrointestinal bleeding with the new oral anticoagulants - defining the issues and the management strategies. Thromb. Haemost. 2013;110(2):205–212. doi: 10.1160/TH13-02-0150. [DOI] [PubMed] [Google Scholar]
- 9.Vrijens B., Heidbuchel H. Non-vitamin K antagonist oral anticoagulants: considerations on once- vs. twice-daily regimens and their potential impact on medication adherence. Europace. 2015;17(4):514–523. doi: 10.1093/europace/euu311. [DOI] [PubMed] [Google Scholar]
- 10.Schulman S., Anger S.U., Bergqvist D., Eriksson B., Lassen M.R., Fisher W. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in surgical patients. J. Thromb. Haemost. 2010;8(1):202–204. doi: 10.1111/j.1538-7836.2009.03678.x. [DOI] [PubMed] [Google Scholar]
- 11.Stekhoven D.J., Bühlmann P. Missforest-non-parametric missing value imputation for mixed-type data. Bioinformatics. 2012;28(1):112–118. doi: 10.1093/bioinformatics/btr597. [DOI] [PubMed] [Google Scholar]
- 12.Zhang Z., Kim H.J., Lonjon G., Zhu Y. Balance diagnostics after propensity score matching. Ann. Transl. Med. 2019;7(1):16. doi: 10.21037/atm.2018.12.10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Heidbuchel H., Verhamme P., Alings M., Antz M., Diener H.C., Hacke W., et al. Updated european heart rhythm association practical guide on the use of non-vitamin K antagonist anticoagulants in patients with non-valvular atrial fibrillation. Europace. 2015;17(10):1467–1507. doi: 10.1093/europace/euv309. [DOI] [PubMed] [Google Scholar]
- 14.Laine L., Maller E.S., Yu C., Quan H., Simon T. Ulcer formation with low-dose enteric-coated aspirin and the effect of COX-2 selective inhibition: a double-blind trial. Gastroenterology. 2004;127(2):395–402. doi: 10.1053/j.gastro.2004.05.001. [DOI] [PubMed] [Google Scholar]
- 15.Goldstein J.L., Eisen G.M., Lewis B., Gralnek I.M., Aisenberg J., Bhadra P., et al. Small bowel mucosal injury is reduced in healthy subjects treated with celecoxib compared with ibuprofen plus omeprazole, as assessed by video capsule endoscopy. Aliment. Pharmacol. Ther. 2007;25(10):1211–1222. doi: 10.1111/j.1365-2036.2007.03312.x. [DOI] [PubMed] [Google Scholar]
- 16.Ferro C.J., Solkhon F., Jalal Z., Al-Hamid A.M., Jones A.M. Relevance of physicochemical properties and functional pharmacology data to predict the clinical safety profile of direct oral anticoagulants. Pharmacol. Res. Perspect. 2020;8(3):1–10. doi: 10.1002/prp2.603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Graham D.J., Reichman M.E., Wernecke M., Hsueh Y.H., Izem R., Southworth M.R., et al. Stroke, bleeding, and mortality risks in elderly medicare beneficiaries treated with dabigatran or rivaroxaban for nonvalvular atrial fibrillation. JAMA Intern. Med. 2016;176(11):1662–1671. doi: 10.1001/jamainternmed.2016.5954. [DOI] [PubMed] [Google Scholar]
- 18.Ray W.A., Chung C.P., Stein C.M., Smalley W., Zimmerman E., Dupont W.D., et al. Association of Rivaroxaban vs apixaban with major ischemic or hemorrhagic events in patients with atrial fibrillation. J. Am. Med. Assoc. 2021;326(23):2395–2404. doi: 10.1001/jama.2021.21222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Stangier J. Clinical pharmacokinetics and pharmacodynamics of the oral direct thrombin inhibitor dabigatran etexilate. Clin. Pharmacokinet. 2008;47(5):285–295. doi: 10.2165/00003088-200847050-00001. [DOI] [PubMed] [Google Scholar]
- 20.Laliberté F., Nelson W.W., Lefebvre P., Schein J.R., Rondeau-Leclaire J., Duh M.S. Impact of daily dosing frequency on adherence to chronic medications among nonvalvular atrial fibrillation patients. Adv. Ther. 2012;29(8):675–690. doi: 10.1007/s12325-012-0040-x. [DOI] [PubMed] [Google Scholar]
- 21.Emren S.V., Zoghi M., Berilgen R., Özdemir İ.H., Çelik O., Çetin N., et al. Safety of once-or twice-daily dosing of non-vitamin K antagonist oral anticoagulants (NOACs) in patients with nonvalvular atrial fibrillation: a NOAC-TR study. Bosn. J. Basic. Med. Sci. 2018;18(2):185–190. doi: 10.17305/bjbms.2017.2279. [DOI] [PMC free article] [PubMed] [Google Scholar]