Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 https://doi.org/10.1186/s12872-021-02019-0 RESEARCH ARTICLE Open Access Demographic, clinical, and functional determinants of antithrombotic treatment in patients with nonvalvular atrial fibrillation Jose María Mostaza1* , Carmen Suarez2, Jose María Cepeda3, Luis Manzano4, Demetrio Sánchez5 and on behalf of the PERFILAR study investigators Abstract Background: This study assessed the sociodemographic, functional, and clinical determinants of antithrombotic treatment in patients with nonvalvular atrial fibrillation (NVAF) attended in the internal medicine setting. Methods: A multicenter, cross-sectional study was conducted in NVAF patients who attended internal medicine departments for either a routine visit (outpatients) or hospitalization (inpatients). Results: A total of 961 patients were evaluated. Their antithrombotic management included: no treatment (4.7%), vitamin K antagonists (VKAs) (59.6%), direct oral anticoagulants (DOACs) (21.6%), antiplatelets (6.6%), and antiplatelets plus anticoagulants (7.5%). Permanent NVAF and congestive heart failure were associated with preferential use of oral anticoagulation over antiplatelets, while intermediate-to high-mortality risk according to the PROFUND index was associated with a higher likelihood of using antiplatelet therapy instead of oral anticoagulation. Longer disease dura- tion and institutionalization were identified as determinants of VKA use over DOACs. Female gender, higher educa- tion, and having suffered a stroke determined a preferential use of DOACs. Conclusions: This real-world study showed that most elderly NVAF patients received oral anticoagulation, mainly VKAs, while DOACs remained underused. Antiplatelets were still offered to a proportion of patients. Longer duration of NVAF and institutionalization were identified as determinants of VKA use over DOACs. A poor prognosis according to the PROFUND index was identified as a factor preventing the use of oral anticoagulation. Keywords: Antithrombotic treatment, Direct-acting oral anticoagulants (DOACs), Nonvalvular atrial fibrillation (NVAF), Vitamin K antagonists (VKAs) Background ischemic stroke and systemic thromboembolism [2, 3]. Atrial fibrillation (AF) is the most common type of sus- The risk of stroke in AF patients is about fivefold higher tained cardiac arrhythmia, and its prevalence rises with than in the non-AF population [4], and AF-related strokes age, with about 18% of patients older than 80 years being are generally more severe, with increased risk of death affected [1]. Nonvalvular AF (NVAF) is strongly associ- and disability compared to strokes from other causes ated with increased morbidity and mortality related to [5]. Elderly patients with AF are at higher risk of stroke than younger AF patients [6, 7]. Indeed, age ≥ 75 years is a significant risk factor comparable to a history of stroke *Correspondence: josemaria.mostaza@salud.madrid.org for the assessment of stroke risk by the ­CHA2DS2-VASc 1 Department of Internal Medicine, Hospital Carlos III, Calle Sinesio score [8]. Prevention of stroke is therefore imperative in Delgado, 10, 28029 Madrid, Spain AF patients, particularly in elderly patients. Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 2 of 12 Oral anticoagulation (OAC) with vitamin K antagonists Based on this background, we conducted a cross-sec- (VKA) has traditionally been the mainstay for stroke pre- tional study to describe the demographic, functional, and vention in AF based on the robust clinical evidence of clinical characteristics of patients with NVAF attend- their efficacy in preventing stroke or systemic embolism ing internal medicine departments in Spanish hospitals and reducing mortality [9]. However, VKAs have several and the potential factors associated with antithrombotic known limitations, including the risk of major bleeding treatment patterns. complications, especially intracranial hemorrhage, many food and drug interactions, and the need for frequent Methods coagulation monitoring due to their narrow therapeu- Study design and patients tic window. Direct-acting oral anticoagulants (DOACs) A multicenter, cross-sectional observational study was targeting thrombin or factor Xa emerged as a welcome conducted in internal medicine departments from 93 addition for stroke prevention in AF. These agents have hospitals distributed across Spain. predictable pharmacodynamic effects, allowing fixed Eligible patients were consecutive adult patients dosing without the need for anticoagulation monitoring (aged ≥ 18 years) diagnosed with NVAF (defined as the [10]. DOACs, such as rivaroxaban, dabigatran, apixaban rhythm disturbance occurring in the absence of rheu- and edoxaban, have demonstrated to be noninferior to matic mitral stenosis or a prosthetic heart valve) who warfarin in stroke prevention without an increased risk attended the internal medicine departments either for a of major bleeding [11–15]. Based on their favorable effi- routine visit (outpatients) or hospitalization (inpatients) cacy, safety profile and convenience of use, DOACs are for any reason during the 9-month enrolment period. recommended over VKAs for stroke prevention in most Patients were selected on the basis of the information patients with NVAF [16]. recorded in the medical charts. The patients must have Adequate selection of antithrombotic therapy for been diagnosed with NVAF before the study inclusion, stroke prevention is critical for improving the clinical and disease-related data (type of NVAF, disease duration, outcome of patients with NVAF. Several clinical practice etc.) was collected from medical charts. Patients cur- guidelines have been developed to guide the manage- rently receiving anticoagulant therapy for venous throm- ment of AF patients, providing clinicians with recom- boembolism and patients participating in a clinical trial mendations on individualization of treatment based on with anticoagulant or antiplatelet agents in the previous the patient’s characteristics [16, 17]. However, the imple- six months were excluded. mentation of guideline recommendations in routine clin- A cross-sectional chart review was performed to collect ical practice may be suboptimal. OAC is still underused patients’ sociodemographic, functional and clinical data, in AF patients who are at high risk of stroke, and many as well as treatment-related data. The social, functional, patients are instead treated with antiplatelet agents or do and cognitive status of patients was also assessed through not receive antithrombotic treatment [18, 19]. Accord- face-to-face interviews with patients at the time of the ingly, despite the higher risk of stroke and bleeding in study visit (cross-sectional evaluation). Cognitive impair- elderly NVAF patients [20], anticoagulation has been ment was evaluated using the Pfeiffer questionnaire [22, traditionally underused in this population mainly due to 23], and data on functional disability assessed using the the high frequency of associated comorbidities, including Barthel index [24] was collected. Patient comorbidity cardiovascular and kidney disease, multiple drug therapy, was measured by using the Charlson comorbidity index and concerns about cognitive impairment and risk of falls [25] (absent of comorbidity: score = 0–1; low comor- and bleeding [21]. bidity: score = 2; severe comorbidity: score ≥ 3). The To date, limited data are available on the clinical man- PROFUND index score [26] was calculated to estimate agement of NVAF patients, particularly in those patients one-year mortality risk as follows: low-risk (12.1–14.6%) attended in the internal medicine setting, where these for a score of 0–2, low-intermediate risk (21.5–31.5%) patients are typically managed. There is therefore a need for a score of 3–6, intermediate-high risk (45–50%) for a to identify current therapies used for stroke prevention score of 7–10, and high-risk (61.3–68%) for a score ≥ 11. and the factors that may guide the selection of treatment Physician’s estimation of the patient’s life expectancy (< 6 in the real-world setting. A better understanding of treat- or ≥ 6 months) was also assessed. The risk of ischemic ment patterns and factors potentially influencing treat- stroke and bleeding was assessed by ­ CHA2DS2-VASc ment strategy is crucial to know whether clinical practice score [8] and HAS-BLED score [27], respectively. is in line with treatment recommendations of the current The independent ethics committee of La Princesa Univer- guidelines. Knowledge about the clinical management sity Hospital (Madrid, Spain) approved the study. All patients of patients with NVAF in clinical practice may improve gave their written informed consent before their inclusion in OAC utilization for stroke prevention and outcomes. the study. The study was carried out in accordance with the Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 3 of 12 Declaration of Helsinki and Good Clinical Practice Guide- Most patients received antithrombotic treatment lines and applicable regulatory requirements. (> 95%). Overall, 88.7% of patients received OAC (VKAs, DOACs, or any OAC plus antiplatelet therapy). The Statistical analysis majority of patients were treated with VKAs (nearly 60% In order to describe the demographic, functional, and of patients), and DOACs were received by less than one- clinical characteristics of patients on each antithrom- quarter of patients. Antiplatelet therapy alone was used botic treatment approach, patients included in the in less than 7% of patients, and OAC plus antiplatelet study were categorized into five groups according to the therapy was received by nearly 8% of patients (Fig. 1). antithrombotic therapy used for their clinical manage- ment as follows: no treatment, VKAs (acenocumarol or Demographic and functional characteristics, morbidity warfarin), DOACs (apixaban, rivaroxaban or dabigatran), and life expectancy antiplatelet drugs (aspirin and/or other antiplatelets), The demographic and functional characteristics, mor- and antiplatelet plus anticoagulant drugs. Only patients bidity and life expectancy of patients according to the who fulfilled the eligibility criteria and could be catego- antithrombotic treatment strategy are shown in Table 1. rized in any of these treatment groups were considered The median age of patients was 81 years, with similar as evaluable for the study analysis. A descriptive analysis proportions of male and female patients. Most patients of the characteristics of each treatment group was per- were not institutionalized (> 90%) and had no cogni- formed using measures of central tendency and disper- tive impairment according to the Pfeiffer questionnaire sion (mean [± standard deviation], median [interquartile (> 60% of patients in all groups). About 65% of patients range]) for quantitative variables, and counts and per- were not dependent on assistance for activities of daily centages for qualitative variables. The Kolmogorov– living (ADLs). The median Charlson comorbidity index Smirnov test and the Shapiro–Wilk test were used for score was 7 (high comorbidity). Most patients (90%) had checking normality of data. The Kruskal–Wallis test was a life expectancy ≥ 6 months according to the investiga- used to compare continuous non-parametric variables tor’s estimation. and the ANOVA to compare continuous parametric vari- Compared with autonomous patients, non-autono- ables. The Chi-square or Fisher’s exact test was used for mous patients were older (84 [80–88] vs. 78 [72–83] comparison of categorical variables. years; p < 0.001), predominantly female (62.3% vs. 43.7%; Stepwise binary logistic regression analyses were con- p < 0.001), and with higher risk of stroke and bleed- ducted to assess the potential factors associated with the ing according to C ­ HA2DS2-VASc (5.6 ± 1.5 vs. 4.4 ± 1.5; use of specific treatment strategies whose comparison is p < 0.001) and HAS-BLED (3.4 ± 1.1 vs. 2.8 ± 1.1; of clinical relevance. Thus, factors associated with OAC p < 0.001) scores. Similarly, patients with cognitive (VKAs, DOACs, or any OAC plus antiplatelet therapy) impairment had a more advanced age (83 [79–88] vs. versus antiplatelet therapy and with VKAs versus DOACs 79.0 [73–84] years; p < 0.001), were mostly women (61.2% were assessed. Clinically relevant variables and those vs. 45.1%; p < 0.001), and had higher ­ CHA2DS2-VASc with a p < 0.2 were included in each multivariate model (5.2 ± 1.6 vs. 4.6 ± 1.6; p < 0.001) and HAS-BLED with stepwise selection to determine independent factors (3.1 ± 1.2 vs. 2.9 ± 1.1; p < 0.001) scores compared with associated with each treatment strategy used for clinical patients without cognitive impairment. Additionally, the management. Odds ratio (OR) and 95% confidence inter- median age of patients with < 4 comorbidities was 64 (54– val (95% CI) were calculated. 67) years while it was 81 (75–86) years in patients with ≥ 4 All analyses were performed using the Statistical Pack- comorbidities (p < 0.001) in whom ­CHA2DS2-VASc and age for the Social Sciences (SPSS) version 18.0 (SPSS Inc., HAS-BLED scores were higher (vs. < 4 comorbidities) Chicago, IL, USA). ­(CHA2DS2-VASc: 4.9 ± 1.5 vs. 1.9 ± 1.0; p < 0.001; HAS- BLED: 3.1 ± 1.1 vs. 1.2 ± 1.0; p < 0.001). Patients with a Results life expectancy < 6 months (vs. ≥ 6 months) were older Patients and treatment (86 [81–91] vs. 80 [74–85] years; p < 0.001) and with A total of 1000 patients from 93 hospitals distributed higher ­CHA2DS2-VASc (5.3 ± 1.6 vs. 4.7 ± 1.6; p < 0.001) across Spain were enrolled in the study from March to and HAS-BLED (3.5 ± 1.2 vs. 2.9 ± 1.1; p < 0.001) scores. October 2015. Thirteen patients were excluded from the The differences in sociodemographic and functional analysis, as they did not meet at least one inclusion crite- characteristics, morbidity and life expectancy between ria. Therefore, a total of 987 patients were included in the the patients undergoing different antithrombotic treat- study analysis. A total of 26 patients could not be catego- ment strategies are shown in Table 1. There were sta- rized in any of the five treatment groups. As a result, 961 tistically significant differences in terms of age and patients were evaluable for the study analysis (Fig. 1). educational level between the different treatment groups. Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 4 of 12 N = 1,000 Patients enrolled N = 13 Did not fullfil at least one inclusion criteria N = 987 Patients evaluable N = 487 outpatients N = 500 inpatients N = 26 20 patients received other anticoagulant (heparin) 6 patients received two anticoagulants Heparin and acenocumarol (n=4) Apixaban and acenocumarol (n=1) Rivaroxaban and acenocumarol (n=1) N = 961 Patients categorized according to the antithrombotic treatment strategy No treatment VKAs DOACs Antiplatelets Antiplatelets plus anticoagulants N = 45 N = 573 N = 208 N = 63 (4.7%) (59.6%) (21.6%) (6.6%) N = 72 (7.5%) Fig. 1 Study patient flow chart Significant differences in functional characteristics (insti- common comorbidities were hypertension (90%), con- tutionalization, dependence in ADLs, and functional dis- gestive heart failure (CHF) (54%), renal disease (43%), ability measured by the Barthel index) were also observed and diabetes mellitus (40%). A history of bleeding was between the different treatment groups except in terms present in less than 15% of patients. of cognitive impairment. In addition, there were statisti- The vast majority of patients had high thromboembolic cally significant differences in estimated prognostic char- risk according to the C ­ HA2DS2-VASc score (≥ 2), with acteristics, including comorbidity (Charlson comorbidity more than 45% of patients with a score ≥ 5 in all treat- index score), 1-year mortality risk (PROFUND index) ment groups. Most patients (> 60%) had a HAS-BLED and life expectancy (investigator’s clinical judgment) score ≥ 3. among the patients treated with the different antithrom- The differences in clinical characteristics between botic treatment strategies. the patients undergoing different antithrombotic treat- ment strategies are shown in Table 2. There were sta- Clinical characteristics tistically significant differences in the type of NVAF Clinical characteristics according to the antithrombotic and the median time since diagnosis among treatment treatment strategy are shown in Table 2. Most patients groups. Cardiovascular history also differed among were diagnosed with permanent NVAF (70%), with a the patients treated with each antithrombotic therapy median time since diagnosis of over four years. The most except for CHF. No significant differences were found Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 5 of 12 Table 1 Patient’s sociodemographic and functional characteristics, morbidity and life expectancy, according to antithrombotic treatment strategy Characteristics No treatment VKAs DOACs Antiplatelets Antiplatelets + Overall p value anticoagulants No. (%) 45 (4.7) 573 (59.6) 208 (21.6) 63 (6.6) 72 (7.5) 961 (100.0) – Age, median (IQR), years a 83.0 (74.5–88.0) 81.0 (75.0–86.0) 80.0 (74.0–85.0) 83.0 (76.0–88.0) 78.0 (70.0–82.8) 81.0 (75.0–86.0) 0.001 ≥ 85 years, n (%) 19 (42.2) 175 (30.5) 56 (26.9) 27 (42.9) 11 (15.3) 288 (30.0) 0.002 Gender, male, n (%) b 24 (53.3) 285 (49.7) 91 (43.8) 30 (47.6) 46 (63.9) 476 (49.5) 0.059 BMI, median (IQR), Kg/m2 a 24.6 (22.7–27.7) 28.0 (24.9–31.7) 27.8 (24.2–31.4) 26.4 (23.5–30.7) 29.2 (25.3–33.6) 27.7 (24.8–31.6) 0.001 Educational level, n (%) b 0.047 Did not complete com- 10 (23.8) 114 (20.9) 36 (17.9) 10 (16.9) 14 (20.0) 184 (20.0) pulsory education Primary education 22 (52.4) 348 (63.7) 113 (56.2) 33 (55.9) 45 (64.3) 561 (61.1) Secondary education 4 (9.5) 61 (11.2) 33 (16.4) 10 (16.9) 8 (11.4) 116 (12.6) University or higher 6 (14.3) 23 (4.2) 19 (9.5) 6 (10.2) 3 (4.3) 57 (6.2) Institutionalized, n (%) c 4 (8.9) 41 (7.2) 2 (1.0) 9 (14.3) 2 (2.8) 58 (6.1) < 0.001 Dependence in ADLs, n 18 (40.9) 187 (32.7) 63 (30.3) 34 (54.0) 25 (34.7) 327 (34.1) 0.008 (%) b Barthel index score < 60 15 (33.3) 84 (14.7) 33 (15.9) 23 (36.5) 7 (9.7) 162 (16.9) < 0.001 (severe/total depend- ence), n (%) b Charlson index, median 7.0 (5.5–9.0) 7.04 (6.0–9.0) 7.0 (5.0–8.0) 8.0 (6.0–9.0) 8.0 (6.0–10.0) 7.04 (6.0–9.0) 0.009 (IQR) c Cognitive impairment 13 (28.9) 172 (30.2) 65 (31.3) 23 (36.5) 18 (25.0) 291 (30.4) 0.692 (SPMSQ score), n (%) b Score > 7 (severe) 3 (6.7) 86 (15.0) 37 (17.8) 12 (19.0) 7 (9.7) 145 (15.1) 0.194 PROFUND index score, 6.0 (1.0–10.5) 3.0 (0.0–6.0) 3.0  (0.0–5.8) 5.0 (0.0–10.0) 2.0 (0.0–5.0) 3.0 (0.0–6.0) < 0.001 median (IQR)a   1-year mortality risk, n < 0.001 (%) b Low (12.1–14.6%) 14 (31.1) 258 (45.0) 99 (47.6) 17 (27.0) 37 (51.4) 425 (44.2) Low-intermediate 10 (22.2) 191 (33.3) 72 (34.6) 17 (27.0) 24 (33.3) 314 (32.7) (21.5–31.5%) Intermediate-high 10 (22.2) 72 (12.6) 19 (9.1) 17 (27.0) 9 (12.5) 127 (13.2) (45.0–50.0%) High (61.3–68.0%) 11 (24.4) 52 (9.1) 18 (8.7) 12 (19.0) 2 (2.8) 95 (9.9) Life expectancy, n (%) b < 0.001 ≥ 6 months 34 (75.6) 522 (91.1) 196 (94.2) 53 (84.1) 68 (94.4) 873 (90.8) < 6 months 11 (24.4) 51 (8.9) 12 (5.8) 10 (15.9) 4 (5.6) 88 (9.2) ADL activities of daily living, BMI body mass index, DOAC direct-acting oral anticoagulant, IQR interquartile range, SD standard deviation, SPMSQ Pfeiffer Short Portable Mental Status Questionnaire, VKA vitamin K antagonist For normally distributed data, mean and standard deviation are used, and for data not normally distributed, median with interquartile range are used. aNon- parametric Kruskal–Wallis test, bChi-square test, cFisher’s exact test in the percentage of patients with high thromboem- Factors associated with the use of OAC versus antiplatelet bolic risk (­CHA2DS2-VASc ≥ 2) who were treated with therapy each antithrombotic therapy. However, the percentage The presence of CHF (OR, 2.084) and permanent NVAF of patients with high hemorrhagic risk (HAS-BLED) dif- (OR, 4.122) was associated with preferential use of OAC fered significantly by treatment strategy (p < 0.001). over antiplatelets, while intermediate- to high-mortality risk according to the PROFUND index (vs. low mortal- Factors associated with antithrombotic treatment patterns ity risk) (OR, 0.455; 0.144, and 0.133 respectively) was The results of the multivariate analysis performed to associated with a higher likelihood of using antiplatelet identify the independent factors associated with the use therapy alone instead of OAC. of OAC versus antiplatelet therapy and of DOACs versus VKAs are summarized in Table 3. Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 6 of 12 Table 2 Clinical characteristics of the overall population and according to antithrombotic treatment strategy Characteristics No treatment VKAs DOACs Antiplatelets Antiplatelets + Overall p value anticoagulants No. (%) 45 (4.7) 573 (59.6) 208 (21.6) 63 (6.6) 72 (7.5) 961 (100.0) –- Inpatient/Outpatient, n (%) 21 (46.7)/ 24 (53.3) 301(52.5)/ 272 70 (33.7)/ 138 46 (73.0)/ 17 (27.0) 43 (59.7)/ 29 (40.3) 481(50.1)/ 480 –- (47.5) (66.3) (49.9) Type of NVAF, n (%) b < 0.001 Paroxysmal 21 (46.7) 90 (15.7) 52 (25.0) 26 (41.3) 14 (19.4) 203 (21.1) Persistent 6 (13.3) 47 (8.2) 19 (9.1) 6 (9.5) 7 (9.7) 85 (8.8) Permanent 18 (40.0) 436 (76.1) 137 (65.9) 31 (49.2) 51 (70.8) 673 (70.0) Time since diagnosis, 2.5 (0.1–7.2) 4.7 (2.3–9.0) 3.0 (1.3–8.0) 5.0 (1.0–8.2) 3.9 (1.4–8.3) 4.2 (1.7–8.6) < 0.001 median (IQR), years a Comorbidities, n (%) b Active neoplasia 11 (24.4) 76 (13.3) 24 (11.5) 12 (19.0) 10 (13.9) 133 (13.8) 0.149 COPD 9 (20.0) 164 (28.6) 53 (25.5) 15 (23.8) 21 (29.2) 262 (27.3) 0.631 Renal disease/CKD 17 (37.8) 180 (31.4) 54 (26.0) 23 (36.5) 24 (33.3) 298 (31.0) 0.336 Cardiovascular history, n (%) Congestive heart failure 21 (46.7) 323 (56.4) 106 (51.0) 28 (44.4) 44 (61.1) 522 (54.3) 0.147 Ischaemic disease d 1 (2.2) 101 (17.6) 28 (13.5) 12 (19.0) 37 (51.4) 179 (18.6) < 0.001 Peripheral artery disease 1 (2.2) 43 (7.5) 17 (8.2) 9 (14.3) 18 (25.0) 88 (9.2) < 0.001 Stroke/TIA 8 (17.8) 95 (16.6) 47 (22.6) 17 (27.0) 22 (30.6) 189 (19.7) 0.016 Arterial thromboembo- 2 (4.4) 29 (5.1) 9 (4.3) 2 (3.2) 2 (2.8) 44 (4.6) 0.885 lism/venous thrombo- embolic disease, n (%) Hypertension, n (%) 35 (77.8) 519 (90.6) 191 (91.8) 54 (85.7) 65 (90.3) 864 (89.9) 0.047 Diabetes mellitus, n (%) 10 (22.2) 237 (41.4) 72 (34.6) 31 (49.2) 36 (50.0) 386 (40.2) 0.007 Abnormal hepatic func- 2 (4.4) 37 (6.5) 14 (6.7) 6 (9.5) 2 (2.8) 61 (6.3) 0.590 tion e, n (%) Abnormal renal function 20 (44.4) 249 (43.5) 82 (39.4) 31 (49.2) 32 (44.4) 414 (43.1) 0.695 f , n (%) Prior bleeding, n (%) b 8 (17.8) 71 (12.4) 33 (15.9) 9 (14.3) 10 (13.9) 131 (13.6) 0.678 CHA2DS2-VASc score, 4.0 (4.0–5.5) 5.0 (4.0–6.0) 5.0 (4.0–6.0) 5.0 (4.0–6.0) 5.0 (4.0–6.8)   5.0 (4.0–6.0) 0.097 median (IQR) a Risk categories, n (%) c < 0.001 0 4 (8.9) 0 (0.0) 0 (0.0) 1 (1.6) 0 (0.0) 5 (0.5) 1 0 (0.0) 6 (1.0) 1 (0.5) 0 (0.0) 1 (1.4) 8 (0.8) ≥2 41 (91.1) 567 (99.0) 207 (99.5) 62 (98.4) 71 (98.6) 948 (98.6) Score ≥ 5, n (%) 19 (46.4) 330 (58.2) 112 (54.6) 39 (62.9) 47 (66.0) 547 (57.6) HAS-BLED score, median 3.0 (2.0–3.0) 3.0 (2.0–4.0) 3.0 (2.0–4.0) 4.0 (3.0–4.0) 4.0 (3.0–5.0) 3.0 (2.0–4.0) < 0.001 (IQR) a Risk categories, n (%) b < 0.001 <3 22 (48.9) 228 (39.8) 75 (36.1) 7 (11.1) 5 (6.9) 337 (35.1) ≥3 23 (51.1) 345 (60.2) 133 (63.9) 56 (88.9) 67 (93.1) 624 (64.9) NVAF non-valvular atrial fibrillation, COPD chronic obstructive pulmonary disease, CKD chronic kidney disease, DOAC direct-acting oral anticoagulant, IQR interquartile range, SD standard deviation, TIA transient ischemic attack, VKA vitamin K antagonist For normally distributed data, mean and standard deviation are used, and for data not normally distributed, median with interquartile range are used. aNon- parametric Kruskal–Wallis test, bChi-square test, cFisher’s exact test. dIncludes myocardial infarction and stable coronary artery disease, echronic hepatic disease (e.g. cirrhosis) or biochemical data indicative of significant hepatic damage (e.g. bilirubin > 2 × the upper normal limit, associated with AST/ALT > 3 × the upper normal limit, etc.), fchronic dialysis, renal transplant or serum creatinine ≥ 200 μmol/l) (yes/no) and renal disease staging based on the glomerular filtration rate (GFR) according to the Kidney Disease Improving Global Outcomes (KDIGO) guidelines G3a, 45–59 mL/min/1,73 m ­ 2; G3b, 30–44 mL/min/1.73 m ­ 2; G4, 15–29 mL/min/1.73 ­m2; or G5, < 15 mL/min/1.73 m ­ 2 Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 7 of 12 Factors associated with the use of DOACs versus VKAs evidence demonstrating its benefit in this population [20, Factors significantly associated with the use of VKAs over 34]. DOACs were longer time since diagnosis (OR, 1.045) and Of note, the percentage of patients who received no institutionalization (OR, 7.744). However, female gender antithrombotic treatment in our series (4.7%) was similar (vs. male), secondary or university or higher studies, and to that previously reported in the European population of having suffered a stroke/transient ischemic attack (OR, the GLORIA-AF registry (4.1%) [35]. However, compared 0.591) were factors associated with preferential use of to other European studies, our analysis showed a signifi- DOACs. cantly lower proportion of untreated patients [36, 37]. A higher rate of untreated patients has also been reported Discussion among elderly patients in the internal medicine setting in The present study assessed the demographic, functional, Spain (13.8%) [38]. and clinical characteristics associated with antithrom- Although a history of prior bleeding has traditionally botic treatment patterns in a large real-life population been associated with anticoagulation under-prescription of patients with NVAF. The median age of patients was in elderly patients [39], bleeding history was not identi- about 80 years, and most patients had permanent NVAF, fied as a factor contributing to anticoagulation underuse with a median disease duration of over four years. As in our study. Indeed, a high HAS-BLED score should not expected, this elderly population had a high comorbid- be a reason for non-prescription of anticoagulation but ity burden, particularly cardiovascular risk factors and rather a tool to identify and control modifiable bleeding disease, a very high thromboembolic risk and a moder- risk factors to reduce the risk of bleeding [40, 41]. ate to high risk of bleeding. Most patients were autono- Antiplatelet therapy has traditionally been used instead mous and had no cognitive impairment. Patients with of anticoagulation to prevent the risk of bleeding despite worse functional status and worse prognosis in terms of the demonstrated greater efficacy of anticoagulant treat- survival were those with more advanced age and with a ment over antiplatelet agents without increasing major higher risk of stroke and bleeding. bleeding [42]. Nevertheless, we found that the use of anti- Despite being a population comprised of elderly platelet therapy alone (7%) was notably lower than that patients with a high comorbidity burden, we found previously reported in primary care (20%) [43]. Interest- that the vast majority of patients (95%) were receiving ingly, antiplatelet therapy users were less frequent com- antithrombotic treatment due to their high thrombo- pared to prior national studies in elderly inpatients (about embolic risk. Most patients received VKA treatment 18%) [38, 44]. Indeed, we found that compared to outpa- (60%) while DOACs were given in less than one-quarter tients, the proportion of hospitalized patients receiving of patients. Of note, antiplatelet therapy alone was pre- antiplatelet therapy was higher, probably because these scribed in less than 7% of patients with NVAF in routine patients are more likely to have frailty and more comor- clinical practice. bidity compared to outpatients and thus may not be con- Older age is a common reason given for not pre- sidered optimal candidates for anticoagulation [28]. scribing anticoagulation [20, 28]. Thus, the high use of Of note, institutionalized patients and those with total OAC shown in this real-world analysis (88%) was espe- or partial dependence for daily activities, who were older cially interesting, since 75% of patients were older than and with higher thromboembolic and bleeding risk, were 75 years and 30% of patients were aged ≥ 85 years. This primarily treated with antiplatelets, which were also pref- study has therefore highlighted that age alone was not erentially used in patients with functional disability and considered as a contraindication for anticoagulation. worse prognosis. These findings, therefore, suggest that The underuse of OAC in elderly patients is mainly due there might be functional and prognostic factors driv- to overestimation of the risk of bleeding in these patients ing the selection of antiplatelet therapy alone over OAC. [29]. However, the benefit of stroke prevention outweighs Indeed, the univariate analysis showed that institution- the potential increased risk of bleeding in these patients alization, dependence for daily activities, 1-year mortality [30, 31], in whom a higher net clinical benefit has been risk assessed by the PROFUND index, and life expec- demonstrated compared to the younger population due tancy (physician’s clinical judgment) were factors sig- to their higher thromboembolic risk [32]. We found a nificantly associated with preferential use of antiplatelet substantial increase in the use of OAC in elderly patients therapy over OAC. However, among these factors, only compared to prior studies conducted in octogenarian AF 1-year mortality risk was retained in the multivariate patients in Spain, where only half of the patients received model, which showed that poor prognosis (intermediate anticoagulation [33]. This difference may suggest changes or high risk) was a factor contributing to OAC non-pre- in physicians’ perceptions of the benefits and risks of scription in favor of antiplatelet therapy alone. OAC for older NVAF patients based on the available Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 8 of 12 Table 3 Factors associated with treatment strategy for NVAF by multivariate analysis Endpointa OR (95% CI) p value Anticoagulants versus antiplatelet therapyb PROFUND Index (1-year mortality risk) (Referral category: low) < 0.0001 Low-intermediate 0.455 (0.221–0.938) 0.033 Intermediate-high 0.144 (0.067–0.312) < 0.0001 High 0.133 (0.056–0.316) < 0.0001 Type of NVAF (Referral category: paroxysmal) < 0.0001 Persistent 2.077 (0.796–5.422) 0.135 Permanent 4.122 (2.281–7.450) < 0.0001 Congestive heart failure (yes vs. no) 2.084 (1.173–3.703) 0.012 VKAs versus DOACsc Gender (male vs. female) 0.677 (0.478–0.959) 0.028 Educational level (Referral category: did not complete compulsory educa- 0.004 tion) Primary education 0.959 (0.617–1.491) 0.853 Secondary education 0.534 (0.297–0.961) 0.036 University or higher 0.337 (0.159–0.715) 0.023 Institutionalization (yes vs. no) 7.744 (1.816–33.027) 0.006 Time since diagnosis of NVAF 1.045 (1.011–1.080) 0.009 Prior stroke/TIA (yes vs. no) 0.591 (0.386–0.903) 0.015 CI: confidence interval, DOAC: direct-acting oral anticoagulant, NVAF non-valvular atrial fibrillation, OR: odds ratio, TIA: transient ischemic attack, VKA: vitamin K antagonist a The covariates included in the univariate models were as follows: age, gender, educational level, institutionalization, dependence in activities of daily living (ADLs), PROFUND index, Charlson comorbidity index, cognitive impairment (SPMSQ score), life expectancy (physician’s criteria), type of NAVF, time since diagnosis of NVAF, active neoplasia, dementia, diabetes mellitus, hypertension, congestive heart failure (CHF), ischaemic disease (myocardial infarction, and/or stable coronary arterial disease), peripheral artery disease (PAD), cerebrovascular disease (prior stroke/transient ischemic attack [TIA]), venous thromboembolic disease, prior bleeding, predisposition to bleeding, abnormal hepatic function, abnormal renal function, and thromboembolic risk (­ CHA2DS2-VASc score) b The variables included in the multivariate analysis were institutionalization (p = 0.005), dependence in ADLs (p = 0.001), 1-year risk mortality risk (PROFUND index) (p < 0.0001), Charlson comorbidity index (p = 0.012), life expectancy (p =0.031 ), type of NVAF (p < 0.001), and dementia (p = 0.007). Age (p =0.098), diabetes mellitus (p = 0.174), arterial hypertension (p =0.183), CHF (p = 0.092), PAD (p = 0.183), and prior stroke/TIA (p = 0.138) were also considered in the multivariate model (p < 0.2) c The variables included in the multivariate analysis were educational level (p = 0.008), institutionalization (p = 0.005), Charlson comorbidity index (p = 0.029), type of NVAF (p = 0.009), and time since diagnosis of NVAF (p < 0.008). Gender (p = 0.139), life expectancy (p = 0.159), diabetes mellitus (p = 0.089), CHF (p = 0.180), ischemic disease (p = 0.167), and prior stroke/TIA (p = 0.055) were also considered in the multivariate model (p < 0.2). Prior bleeding (p = 0.208) and abnormal renal function (0.314) were also included as relevant determinants of anticoagulant treatment This real-world assessment has also highlighted that DOACs among patients receiving OAC (24.4%) was simi- the type of NVAF appears to be a factor that has played lar to that reported in Spanish studies conducted at the an essential role in the underuse of OAC in NVAF. In this regional level during 2015 (24%) and 2011–2014 (25%) regard, we found that patients with paroxysmal NVAF [49] and the nationwide FANTASIIA registry (22%) [50]. were more likely to receive antiplatelet therapy alone Thus, a substantial proportion of Spanish NVAF patients instead of OAC compared to patients with permanent who could benefit from DOACs did not receive them, as NVAF in line with prior reports [1, 35, 45]. The under- recognized by experts [51]. In addition, the preference for use of OAC in favor of antiplatelet therapy in patients VKAs over DOACs may be due to the lack of available with paroxysmal NVAF could be due to the perception targeted DOAC reversal agents for patients with bleed- of a lower thromboembolic risk associated with this type ing at the time of the study. Moreover, the prescription of NVAF [46]. However, contrary findings regarding its of DOACs can be challenging in Spain due to the restric- impact on thromboembolic risk have been reported [47, tions for their use stated in the national Therapeutic 48]. Nevertheless, according to current guidelines [16], Positioning Report (TPR) [52] and the different admin- the decision concerning antithrombotic therapy should istrative requirements of the Spanish Autonomous Com- be based on the patient’s risk of stroke and bleeding, munities. This study also found that the use of DOACs in regardless of the type of NVAF. Spain is still far behind other European countries, where Regarding the type of OAC, we found that a significant an overall rate of DOAC prescription of approximately proportion of patients still received VKAs over DOACs 50% [35] has been reported. However, geographical despite guideline recommendations [16]. The use of Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 9 of 12 differences have been found in prescription patterns [36, we should note that these groups were created according 53, 54]. to clinical practice, given that patients could be receiv- We found that neither the functional characteristics ing any treatment for NVAF when they were included nor the prognosis (PROFUND index) or morbidity were in the study. Indeed, this study did not aim to assess the associated with the use of DOACs. Nevertheless, pref- differences between these groups, and the comparisons erential use of VKAs over DOACs was found among made between them are exploratory and descriptive. institutionalized patients, probably due to a more con- Despite these limitations, this study provides valuable trolled environment which may enhance VKA treatment real-world data on the profile and treatment patterns of adherence and might avoid unacceptable risks derived NVAF patients and grounds for discussion on whether from poor anticoagulation control. There appear to be antithrombotic treatment strategies are in line with sociodemographic factors driving DOAC selection, such guideline recommendations. Thus, the findings obtained as gender and educational level. Thus, educational level in this analysis may be used to identify critical issues was identified as an independent factor associated with that should be improved in the management of NVAF the use of DOACs, with a preference of these agents over patients, which might optimize patient care and out- VKAs in more educated patients, probably also with a comes. The strengths of this study include a highly repre- higher income, in line with prior reports [55]. This may sentative population of non-selected elderly patients with be explained by better access of patients with a high edu- NVAF in clinical practice, including inpatients and out- cational level to the available information on new oral patients, attended in the internal medicine departments anticoagulants, which may influence treatment decision of more than 90 hospitals distributed homogeneously for stroke prevention. In addition, patients with longer throughout the country. To our knowledge, this is the disease duration were also more likely to receive VKA first and the most extensive study to assess the real-world therapy over DOACs probably because prescription of characteristics of NVAF patients according to antithrom- DOACs may have only increased over the past few years botic treatment after approval of DOACs in the internal due to current guidelines recommendation [16]. A prior medicine setting in Spain. history of stroke was also associated with a higher likeli- hood of receiving DOACs according to the European rec- Conclusion ommendations [16]. This study showed that the vast majority of elderly DOAC prescription has been traditionally limited by patients with NVAF, with high comorbidity burden and renal function, especially in elderly patients whose renal high thromboembolic risk, received OAC in the real- function is often unstable and affected by concomitant world setting, with VKAs as the most frequently pre- comorbidities and hospitalizations [56, 57]. However, scribed treatment, while DOACs remained underused. abnormal renal function was not identified as a factor Longer duration of AF and institutionalization were contributing to DOAC non-prescription, in contrast to identified as determinants of VKA use over DOACs. other reports in the primary care setting [49]. Indeed, Antiplatelet therapy was still offered to a proportion of renal impairment should not be a limitation to DOAC patients. A poor prognosis according to the PROFUND prescription and reduced dose regimens of DOACs are index was identified as a factor preventing the use of approved for use in moderate and severe chronic kidney OAC. disease. The main limitations of this study are due to the Abbreviations research design. This was a cross-sectional study which ADL: Activities of daily living; AF: Atrial fibrillation; CHF: Congestive heart included a mixed incident and prevalent NVAF popula- failure; CI: Confidence interval; DOAC: Direct oral anticoagulants; NVAF: tion; new starters and switchers from antithrombotic Non-valvular atrial fibrillation; OAC: Oral anticoagulation; OR: Odds ratio; VKA: Vitamin K antagonist. therapy, who were selected from a large sample of hospi- tals in Spain. However, it is noteworthy that this study did Acknowledgements not aim to focus on treatment patterns, such as switch- The authors would like to acknowledge the investigators of the PERFILAR study for its contribution to the study. We also thank Cristina Vidal and Antonio ing or discontinuation. In addition, we must take into Torres from Dynamic Science (Spain) for their medical writing support, funded account the obvious limitations of a retrospective chart by Alliance Bristol-Myers Squibb/Pfizer, Spain. The authors would also like review that uses patients’ data already recorded in the to thank Alliance Bristol-Myers Squibb/Pfizer for sponsoring/supporting the study. medical charts for reasons other than research, includ- Investigators of the PERFILAR study group (collaborators): Fernando Javier ing incomplete or unrecorded information. A further Sánchez Lora, Hospital Clínico Universitario Virgen de la Victoria, Málaga limitation that should be acknowledged is the unequal (Spain); Francisco Ibañez Bermúdez, Hospital Comarcal Infanta Margarita, Cór- doba (Spain); Ana María Jurado Porcel, Complejo Hospitalario Regional Reina number of patients in each group defined by the thera- Sofía, Córdoba (Spain); Fernando Salgado Ordoñez, Complejo Hospitalario peutic strategy used for NVAF management. However, Regional de Málaga (Hospital Carlos Haya), Málaga (Spain); Francisco Rivera Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 10 of 12 Civico, Complejo Hospitalario de Poniente, El Ejido, Almería (Spain); Luis Felipe Tascón, Hospital Universitario Infanta Leonor, Madrid (Spain); Antonio Muiño Díez García, Hospital Torrecárdenas, Almeria (Spain); Fernando Jaén Águila, Miguez, Hospitalario General Universitario Gregorio Marañón, Madrid (Spain); Hospital Universitario Virgen de las Nieves, Granada (Spain); Manuel Geraldía Jorge Marrero Francés, Hospital de Fuenlabrada, Fuenlabrada, Madrid (Spain); Lama, Hospital Virgen de las Montañas, Cádiz (Spain); Enrique Peral Gutiérrez- Nicolas Ortega, Hospital Virgen de la Arrixaca, El Palmar, Murcia (Spain); Javier Ceballos, Complejo Hospitalario Virgen Macarena, Sevilla (Spain); Antonia Trujillo, Hospital Santa Lucía, Cartagena, Murcia (Spain); Julio Sánchez Álvarez, Domínguez, Hospital Quironsalud Sagrado Corazón, Sevilla (Spain); Francisco Complejo Hospitalario de Navarra, Pamplona, Navarra (Spain); Jose Ignacio Astudillo Martín, Hospital Infanta Luisa, Sevilla (Spain); Eduardo Aguilar, Catalán Ramos, Policlínica San José, Vitoria, Álava (Spain); Francisco Javier Hospital de Alcañiz, Teruel (Spain); Juan Ferrando Vela, Hospital Royo Villanova, Fresco Benito, Hospital Santa Marina, Bilbao, Vizcaya (Spain); Ainhoa Anuzita Zaragoza (Spain); Alfonso García Aranda, Hospital Universitario Miguel Servet, Alegría, Clínica Virgen Blanca, Bilbao, Vizcaya (Spain); Carlos Teruel, Hospital Zaragoza (Spain); Mercedes Sánchez Cembellín, Hospital San Agustín, Avilés, Gral. Castellón, Castellón de la Plana, Castellón (Spain); Arturo Artero Mora, Asturias (Spain); Juan Francisco López Caleya, Hospital de Cabueñes, Gijón Hospital Doctor Peset, Valencia (Spain); Pedro Moral, Hospital La Fe, Valencia (Spain); Sixto Ruiz, Hospital Comarcal de Inca, Islas Baleares (Spain); Melchor (Spain); José Miguel Seguí Ripoll, Hospital Universitario San Juan, Alicante Rodríguez Gaspar, Hospital Universitario de Canarias, Santa Cruz de Tenerife (Spain); Fernando Bonilla Rovira, Hospital General de Elche. Elche, Alicante (Spain); Alicia Conde Martel, Hospital Universitario de Gran Canaria Doctor (Spain); Ana Maestre Peiro, Hospital del Vinalopó, Elche, Alicante (Spain). Negrín, Las Palmas de Gran Canaria, Las Palmas (Spain); José Luis Hernández Hernández, Hospital Universitario Marqués de Valdecilla, Santander (Spain); Authors’ contributions Ismael Abascal Carrera, Hospital Comarcal de Laredo, Laredo (Spain); Alfonso JMM, CS, JMC, LM, and DS have contributed to the design, acquisition of data, Pérez del Molino Castellanos, Hospital Sierrallana, Torrelavega, Cantabria data analysis and interpretation, and manuscript preparation and review. All (Spain); Esther Fernández Pérez, Complejo Asistencial Universitario de León, authors read and approved the final version of the manuscript. León (Spain); Juan Carlos Martínez Acitores, Complejo Asistencial Universitario de Burgos, Burgos (Spain); Luis Miguel Seisdedos Cortes, Complejo Asistencial Funding de Zamora, Zamora (Spain); Laura Abad Manteca, Hospital Universitario Rio Alliance Bristol-Myers Squibb/Pfizer Spain supported this study. However, this Hortega, Valladolid (Spain); Marco Budiño Sánchez, Complejo Asistencial de company has not participated in data collection and analysis and interpreta- Ávila, Ávila (Spain); José Javier Moreno Palomares, Complejo Asistencial de tion of the data, and the preparation of the manuscript. Segovia, Segovia (Spain); Inmaculada Coca Prieto, Hospital Santa Bárbara, Ciudad Real (Spain); Ana Isabel Muñoz, Hospital Nuestra Señora del Prado, Availability of data and materials Talavera de la Reina, Toledo (Spain); Ángel Sánchez Castaño, Hospital Virgen The datasets used and/or analyzed during the current study are available from de la Salud, Toledo (Spain); Lola Ruiz Ribó, Hospital Virgen de la Luz, Cuenca the corresponding author on reasonable request. (Spain); Jordi Mascaró, Hospital de la Santa Creu i Sant Pau, Barcelona (Spain); César Morcillo Serra, Hospital Cima Sanitas, Barcelona (Spain); Teresa Auguet Quintillá, Hospital Universitari Joan XXIII, Tarragona (Spain); Francesz Marimón, Declarations Hospital Universitari de Sant Joan de Reus, Tarragona (Spain); Joaquín Fernán- dez Solá, Hospital Clinic i Provincial, Barcelona (Spain); José María Suriñach, Ethics approval and consent to participate Hospital Universitari Vall d’Hebron, Barcelona (Spain); Pablo Marchena, Parc The study was approved by the Independent Ethics Committee of La Princesa Sanitari Sant Joan de Deu, Barcelona (Spain); Antoni Riera-Mestre, Hospital University Hospital (Madrid, Spain), and written informed consent was Universitari de Bellvitge, Bellvitge Biomedical Research Institute (IDIBELL), obtained from all patients before they were included in the study. L’Hospitalet de Llobregat, Barcelona (Spain); Pedro Armario, Hospital de Sant Joan Despí Moisés Broggi, Sant Joan Despí, Barcelona (Spain); Ferrán García Consent for publication Bragado, Hospital Universitari de Girona Dr. Josep Trueta, Girona (Spain); Not applicable as this manuscript does not contain any person’s data. Fátima del Molino, Hospital Quirón Salud del Vallés, Sabadell, Barcelona (Spain); Oscar Sacristán, Hospital Santa María, Alcalá de Henares, Madrid Competing interests (Spain); Pere Almagro, Hospital Mutua de Terrassa, Terrassa, Barcelona (Spain); C. Suarez has received speaker honoraria or advisory board fees from BMS/ Conxita Falgà, Hospital de Mataró, Mataró, Barcelona (Spain); Francisco José Pfizer, Bayer, and Daiichi Sankyo. J.M. Cepeda has received speaker honoraria Muñoz Rodríguez, Hospital de Mollet, Mollet del Vallès, Barcelona (Spain); and advisory board fees from BMS/Pfizer, Bayer, and Daiichi Sankyo. L. Man- Riesco, Hospital del Vendrell, El Vendrell, Tarragona (Spain); Jorge Romero zano has received speaker honoraria or advisory board fees from BMS/Pfizer, Requena, Hospital Perpetuo Socorro, Badajoz (Spain); José Carlos Arévalo Bayer, Boehringer Ingelheim and Daiichi Sankyo. D. Sánchez has received Lorido, Complejo Hospitalario Llerena-Zafra, Llerena, Badajoz (Spain); Manuela payments from collaborations with Bayer, BMS/Pfizer, Boehringer Ingelheim Chiquero Palomo, Complejo Hospitalario de Cáceres, Cáceres (Spain); Ana and Daiichi Sankyo. J.M. Mostaza declares that he does not have any conflict Isabel de la Cruz, Hospital Virgen del Puerto, Plasencia, Cáceres (Spain), Agustín of interest. Pijerro, Hospital Infanta Cristina, Badajoz (Spain); Elena Fernández Bouza, Com- plexo Hospitalario Universitario Arquitecto Marcide-Novoa Santos, Ferrol, A Author details 1 Coruña (Spain); Juan José González Soler, Complexo Hospitalario Universitario Department of Internal Medicine, Hospital Carlos III, Calle Sinesio Delgado, 10, de Ourense, Ourense (Spain); Manuel Jesús Núñez Fernández, Complexo 28029 Madrid, Spain. 2 Department of Internal Medicine, Hospital Universitario Hospitalario Universitario de Pontevedra, Pontevedra (Spain); Javier De La de La Princesa, Madrid, Spain. 3 Department of Internal Medicine, Hospital Fuente Aguado, Hospital de Povisa, Vigo, Pontevedra (Spain); José Antonio Vega Baja, Orihuela, Alicante, Spain. 4 Department of Internal Medicine, Díaz Peromingo, Complejo Hospitalario Universitario de Santiago, Santiago de Hospital Ramón Y Cajal, Universidad de Alcalá, Ramón Y Cajal Health Research Compostela, A Coruña (Spain); Julián Fernández Martín, Complejo Hospitalario Institute (IRYCIS), Madrid, Spain. 5 Department of Internal Medicine, Hospital de Vigo, Vigo, Pontevedra (Spain); Rafael Daroca Pérez, Complejo Hospital Nuestra Señora De Sonsoles, Ávila, Spain. 6 Hospital Clínico Universitario San Pedro, Logroño, La Rioja (Spain); Dr. Jesús Castiella Herrero, Fundación Virgen de La Victoria, Málaga, Spain. 7 Hospital Comarcal Infanta Margarita, Hospital Calahorra, Calahorra, La Rioja (Spain); M. Cruz Carreño, Hospital Puerta Córdoba, Spain. 8 Complejo Hospitalario Regional Reina Sofía, Córdoba, Spain. 9 de Hierro, Madrid (Spain); Jorge Gómez Cerezo, Hospital Universitario Infanta Complejo Hospitalario Regional de Málaga (Hospital Carlos Haya), Málaga, Sofía, San Sebastián de los Reyes, Madrid (Spain); José Carlos Pontes Navarro, Spain. 10 Complejo Hospitalario de Poniente, El Ejido, Almería, Spain. 11 Hospital Hospital Clínico San Carlos, Madrid (Spain); José Felipe Varona Arche, Hospital Torrecárdenas, Almeria, Spain. 12 Hospital Universitario Virgen de Las Nieves, Universitario HM Montepríncipe, Boadilla del Monte, Madrid (Spain); Daniel Granada, Spain. 13 Hospital Virgen de Las Montañas, Cádiz, Spain. 14 Complejo Ferreiro López, Hospital Universitario 12 de octubre, Madrid (Spain); Benjamín Hospitalario Virgen Macarena, Sevilla, Spain. 15 Hospital Quironsalud Sagrado Muñoz Calvo, Hospital Universitario Príncipe de Asturias, Alcalá de Henares, Corazón, Sevilla, Spain. 16 Hospital Infanta Luisa, Sevilla, Spain. 17 Hospital de Madrid (Spain); Jesús Manuel Casado Cerrada, Hospital Universitario de Getafe, Alcañiz, Teruel, Spain. 18 Hospital Royo Villanova, Zaragoza, Spain. 19 Hospital Getafe, Madrid (Spain); María del Pilar Fidalgo Montero, Hospital Universi- Universitario Miguel Servet, Zaragoza, Spain. 20 Hospital San Agustín, Avilés, tario del Henares, Coslada, Madrid (Spain); José Manuel Casas Rojo, Hospital Asturias, Spain. 21 Hospital de Cabueñes, Gijón, Spain. 22 Hospital Comarcal de Universitario Infanta Cristina, Parla, Madrid (Spain); Benjamín Herreros, Hospital Inca, Islas Baleares, Spain. 23 Hospital Universitario de Canarias, Santa Cruz de Universitario Fundación Alcorcón, Alcorcón, Madrid (Spain); Guillermo Cuevas Tenerife, Spain. 24 Hospital Universitario de Gran Canaria Doctor Negrín, Las Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 11 of 12 Palmas de Gran Canaria, Las Palmas, Spain. 25 Hospital Universitario Marqués patients without: data from the Austrian Stroke registry. Eur Heart J. de Valdecilla, Santander, Spain. 26 Hospital Comarcal de Laredo, Laredo, Spain. 2004;25:1734–40. 27 Hospital Sierrallana, Torrelavega, Cantabria, Spain. 28 Complejo Asistencial 6. Lakshminarayan K, Solid CA, Collins AJ, Anderson DC, Herzog CA. Atrial Universitario de León, León, Spain. 29 Complejo Asistencial Universitario de fibrillation and stroke in the general medicare population: a 10-year Burgos, Burgos, Spain. 30 Complejo Asistencial de Zamora, Zamora, Spain. perspective (1992 to 2002). Stroke. 2006;37:1969–74. 31 Hospital Universitario Rio Hortega, Valladolid, Spain. 32 Complejo Asistencial 7. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation: a major contribu- de Ávila, Ávila, Spain. 33 Complejo Asistencial de Segovia, Segovia, Spain. 34 Hos- tor to stroke in the elderly. The Framingham Study. Arch Intern Med. pital Santa Bárbara, Ciudad Real, Spain. 35 Hospital Nuestra Señora del Prado, 1987;147:1561–4. Talavera de La Reina, Toledo, Spain. 36 Hospital Virgen de La Salud, Toledo, 8. Lip GY, Nieuwlaat R, Pisters R, Lane DA, Crijns HJ. Refining clinical risk Spain. 37 Hospital Virgen de La Luz, Cuenca, Spain. 38 Hospital de La Santa Creu stratification for predicting stroke and thromboembolism in atrial fibril- I Sant Pau, Barcelona, Spain. 39 Hospital Cima Sanitas, Barcelona, Spain. 40 Hos- lation using a novel risk factor-based approach: the euro heart survey on pital Universitari Joan XXIII, Tarragona, Spain. 41 Hospital Universitari de Sant atrial fibrillation. Chest. 2010;137:263–72. Joan de Reus, Tarragona, Spain. 42 Hospital Clinic I Provincial, Barcelona, Spain. 9. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to 43 Hospital Universitari Vall D’Hebron, Barcelona, Spain. 44 Parc Sanitari Sant prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Joan de Deu, Barcelona, Spain. 45 Hospital Universitari de Bellvitge, Bellvitge Intern Med. 2007;146:857–67. Biomedical Research Institute (IDIBELL), L’Hospitalet de Llobregat, Barcelona, 10. Mekaj YH, Mekaj AY, Duci SB, Miftari EI. New oral anticoagulants: their Spain. 46 Hospital de Sant Joan Despí Moisés Broggi, Sant Joan Despí, Barce- advantages and disadvantages compared with vitamin K antagonists in lona, Spain. 47 Hospital Universitari de Girona Dr. Josep Trueta, Girona, Spain. the prevention and treatment of patients with thromboembolic events. 48 Hospital Quirón Salud del Vallés, Sabadell, Barcelona, Spain. 49 Hospital Santa Ther Clin Risk Manag. 2015;11:967–77. María, Alcalá de Henares, Madrid, Spain. 50 Hospital Mutua de Terrassa, Terrassa, 11. Connolly SJ, Eikelboom J, Joyner C, Diener HC, Hart R, Golitsyn Barcelona, Spain. 51 Hospital de Mataró, Mataró, Barcelona, Spain. 52 Hospital de S, et al. Apixaban in patients with atrial fibrillation. N Engl J Med. Mollet, Mollet del Vallès, Barcelona, Spain. 53 Hospital del Vendrell, El Vendrell, 2011;364:806–17. Tarragona, Spain. 54 Hospital Perpetuo Socorro, Badajoz, Spain. 55 Complejo 12. Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, Hospitalario Llerena-Zafra, Llerena, Badajoz, Spain. 56 Complejo Hospitalario et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J de Cáceres, Cáceres, Spain. 57 Hospital Virgen del Puerto, Plasencia, Cáceres, Med. 2009;361:1139–51. Spain. 58 Hospital Infanta Cristina, Badajoz, Spain. 59 Complexo Hospitalario 13. Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, et al. Rivar- Universitario Arquitecto Marcide-Novoa Santos, Ferrol, A Coruña, Spain. oxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 60 Complexo Hospitalario Universitario de Ourense, Ourense, Spain. 61 Com- 2011;365:883–91. plexo Hospitalario Universitario de Pontevedra, Pontevedra, Spain. 62 Hospital 14. Ruff CT, Giugliano RP, Braunwald E, Hoffman EB, Deenadayalu N, Ezekow- de Povisa, Vigo, Pontevedra, Spain. 63 Complejo Hospitalario Universitario de itz MD, et al. Comparison of the efficacy and safety of new oral antico- Santiago, Santiago de Compostela, A Coruña, Spain. 64 Complejo Hospitalario agulants with warfarin in patients with atrial fibrillation: a meta-analysis of de Vigo, Vigo, Pontevedra, Spain. 65 Complejo Hospital San Pedro, Logroño, randomised trials. Lancet. 2014;383:955–62. La Rioja, Spain. 66 Fundación Hospital Calahorra, Calahorra, La Rioja, Spain. 15. Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, 67 Hospital Puerta de Hierro, Madrid, Spain. 68 Hospital Universitario Infanta et al. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Sofía, San Sebastián de Los Reyes, Madrid, Spain. 69 Hospital Clínico San Carlos, Med. 2013;369:2093–104. Madrid, Spain. 70 Hospital Universitario HM Montepríncipe, Boadilla del Monte, 16. Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, et al. 2016 Madrid, Spain. 71 Hospital Universitario 12 de octubre, Madrid, Spain. 72 Hospital ESC Guidelines for the management of atrial fibrillation developed in Universitario Príncipe de Asturias, Alcalá de Henares, Madrid, Spain. 73 Hospital collaboration with EACTS. Eur Heart J. 2016;37:2893–962. Universitario de Getafe, Getafe, Madrid, Spain. 74 Hospital Universitario del 17. Steffel J, Verhamme P, Potpara TS, Albaladejo P, Antz M, Desteghe L, et al. Henares, Coslada, Madrid, Spain. 75 Hospital Universitario Infanta Cristina, Parla, The 2018 European Heart Rhythm Association Practical Guide on the use Madrid, Spain. 76 Hospital Universitario Fundación Alcorcón, Alcorcón, Madrid, of non-vitamin K antagonist oral anticoagulants in patients with atrial Spain. 77 Hospital Universitario Infanta Leonor, Madrid, Spain. 78 Hospitalario fibrillation. Eur Heart J. 2018;39:1330–93. General Universitario Gregorio Marañón, Madrid, Spain. 79 Hospital de Fuen- 18. Le Heuzey JY, Ammentorp B, Darius H, De Caterina R, Schilling RJ, Schmitt labrada, Fuenlabrada, Madrid, Spain. 80 Hospital Virgen de La Arrixaca, El Palmar, J, et al. Differences among western European countries in anticoagu- Murcia, Spain. 81 Hospital Santa Lucía, Cartagena, Murcia, Spain. 82 Complejo lation management of atrial fibrillation. Data from the PREFER IN AF Hospitalario de Navarra, Pamplona, Navarra, Spain. 83 Policlínica San José, registry. Thromb Haemost. 2014;111:833–41. Vitoria, Álava, Spain. 84 Hospital Santa Marina, Bilbao, Vizcaya, Spain. 85 Clínica 19. Wilke T, Groth A, Mueller S, Pfannkuche M, Verheyen F, Linder R, et al. Oral Virgen Blanca, Bilbao, Vizcaya, Spain. 86 Hospital Gral. Castellón, Castellón de La anticoagulation use by patients with atrial fibrillation in Germany Adher- Plana, Castellón, Spain. 87 Hospital Doctor Peset, Valencia, Spain. 88 Hospital La ence to guidelines, causes of anticoagulation under-use and its clinical Fe, Valencia, Spain. 89 Hospital Universitario San Juan, Alicante, Spain. 90 Hos- outcomes, based on claims-data of 183,448 patients. Thromb Haemost. pital General de Elche. Elche, Alicante, Spain. 91 Hospital del Vinalopó, Elche, 2012;107:1053–65. Alicante, Spain. 20. Sharma M, Cornelius VR, Patel JP, Davies JG, Molokhia M. Efficacy and harms of direct oral anticoagulants in the elderly for stroke Received: 21 September 2020 Accepted: 15 April 2021 prevention in atrial fibrillation and secondary prevention of venous thromboembolism: systematic review and meta-analysis. Circulation. 2015;132:194–204. 21. Denoel P, Vanderstraeten J, Mols P, Pepersack T. Could some geriatric characteristics hinder the prescription of anticoagulants in atrial fibril- References lation in the elderly? J Aging Res. 2014;2014:693740. 1. Barrios V, Calderon A, Escobar C, de la Figuera M. Patients with atrial fibril- 22. Martinez de la Iglesia J, Duenas Herrero R, Onis Vilches MC, Aguado lation in a primary care setting: Val-FAAP study. Rev Esp Cardiol (Engl Ed). Taberne C, Albert Colomer C, Luque Luque R. Spanish language adap- 2012;65:47–53. tation and validation of the Pfeiffer’s questionnaire (SPMSQ) to detect 2. Lip GY. Stroke in atrial fibrillation: epidemiology and thromboprophylaxis. cognitive deterioration in people over 65 years of age [Adaptacion J Thromb Haemost. 2011;9(Suppl 1):344–51. y validacion al castellano del cuestionario de Pfeiffer (SPMSQ) para 3. Lip GY, Lim HS. Atrial fibrillation and stroke prevention. Lancet Neurol. detectar la existencia de deterioro cognitivo en personas mayores de 2007;6:981–93. 65 anos]. Med Clin (Barc). 2001;117:129–34. 4. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk 23. Pfeiffer E. A short portable mental status questionnaire for the assess- factor for stroke: the Framingham study. Stroke. 1991;22:983–8. ment of organic brain deficit in elderly patients. J Am Geriatr Soc. 5. Steger C, Pratter A, Martinek-Bregel M, Avanzini M, Valentin A, Slany J, 1975;23:433–41. et al. Stroke patients with atrial fibrillation have a worse prognosis than 24. Mahoney FI, Barthel DW. Functional evaluation: the barthel index. Md State Med J. 1965;14:61–5. Mostaza et al. BMC Cardiovasc Disord (2021) 21:384 Page 12 of 12 25. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of clas- 43. Polo Garcia J, Vargas Ortega D, Formiga F, Unzueta I, Fernandez de Cabo sifying prognostic comorbidity in longitudinal studies: development S, Chaves J. Profiling of patients with non-valvular atrial fibrillation and and validation. J Chronic Dis. 1987;40:373–83. moderate-to-high risk of stroke not receiving oral anticoagulation in 26. Bernabeu-Wittel M, Ollero-Baturone M, Moreno-Gavino L, Baron-Franco Spain. Semergen. 2018. B, Fuertes A, Murcia-Zaragoza J, et al. Development of a new predictive 44. Lopez Soto A, Formiga F, Bosch X, Garcia AJ. Prevalence of atrial fibril- model for polypathological patients. The PROFUND index. Eur J Intern lation and related factors in hospitalized old patients: ESFINGE study Med. 2011;22:311–7. [Prevalencia de la fibrilacion auricular y factores relacionados en 27. Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJ, Lip GY. A novel pacientes ancianos hospitalizados: estudio ESFINGE]. Med Clin (Barc). user-friendly score (HAS-BLED) to assess 1-year risk of major bleed- 2012;138:231–7. ing in patients with atrial fibrillation: the Euro Heart Survey. Chest. 45. Lobos-Bejarano JM, del Castillo-Rodriguez JC, Mena-Gonzalez A, 2010;138:1093–100. Aleman-Sanchez JJ, Cabrera de Leon A, Baron-Esquivias G, et al. Patients’ 28. Hylek EM, D’Antonio J, Evans-Molina C, Shea C, Henault LE, Regan S. characteristics and clinical management of atrial fibrillation in primary Translating the results of randomized trials into clinical practice: the healthcare in Spain: FIATE Study [Caracteristicas de los pacientes y challenge of warfarin candidacy among hospitalized elderly patients abordaje terapeutico de la fibrilacion auricular en atencion primaria en with atrial fibrillation. Stroke. 2006;37:1075–80. Espana: Estudio FIATE]. Med Clin (Barc). 2013;141:279–86. 29. Pugh D, Pugh J, Mead GE. Attitudes of physicians regarding anti- 46. Van Staa TP, Setakis E, Di Tanna GL, Lane DA, Lip GY. A comparison of risk coagulation for atrial fibrillation: a systematic review. Age Ageing. stratification schemes for stroke in 79,884 atrial fibrillation patients in 2011;40:675–83. general practice. J Thromb Haemost. 2011;9:39–48. 30. Gage BF, Birman-Deych E, Kerzner R, Radford MJ, Nilasena DS, Rich MW. 47. Boriani G, Laroche C, Diemberger I, Fantecchi E, Popescu MI, Rasmussen Incidence of intracranial hemorrhage in patients with atrial fibrillation LH, et al. ’Real-world’ management and outcomes of patients with par- who are prone to fall. Am J Med. 2005;118:612–7. oxysmal vs. non-paroxysmal atrial fibrillation in Europe: the EURObser- 31. Garwood CL, Corbett TL. Use of anticoagulation in elderly patients vational Research Programme-Atrial Fibrillation (EORP-AF) General Pilot with atrial fibrillation who are at risk for falls. Ann Pharmacother. Registry. Europace. 2016;18:648–57. 2008;42:523–32. 48. Ganesan AN, Chew DP, Hartshorne T, Selvanayagam JB, Aylward PE, Sand- 32. Singer DE, Chang Y, Fang MC, Borowsky LH, Pomernacki NK, Udaltsova ers P, et al. The impact of atrial fibrillation type on the risk of thromboem- N, et al. The net clinical benefit of warfarin anticoagulation in atrial bolism, mortality, and bleeding: a systematic review and meta-analysis. fibrillation. Ann Intern Med. 2009;151:297–305. Eur Heart J. 2016;37:1591–602. 33. Hernández VSJ, Mazoteras V, López MT. Underuse of oral anticoagula- 49. Rodriguez-Bernal CL, Hurtado I, Garcia-Sempere A, Peiro S, Sanfelix- tion in octogenarian patient with atrial fibrillation: a perspective from Gimeno G. Oral anticoagulants initiation in patients with atrial fibrillation: real clinical practice. Rev Colom Cardiol. 2015;22:231–4. real-world data from a population-based cohort. Front Pharmacol. 34. Bai Y, Guo SD, Deng H, Shantsila A, Fauchier L, Ma CS, et al. Effective- 2017;8:63. ness and safety of oral anticoagulants in older patients with atrial fibril- 50. Moreno-Arribas J, Bertomeu-Gonzalez V, Anguita-Sanchez M, Cequier A, lation: a systematic review and meta-regression analysis. Age Ageing. Muniz J, Castillo J, et al. Choice of new oral anticoagulant agents versus 2018;47:9–17. vitamin K antagonists in atrial fibrillation: FANTASIIA study. J Cardiovasc 35. Huisman MV, Rothman KJ, Paquette M, Teutsch C, Diener HC, Pharmacol Ther. 2016;21:150–6. Dubner SJ, et al. The changing landscape for stroke prevention in 51. Barrios V, Escobar C, Lobos JM, Polo J, Vargas D. Use of non-vitamin K AF: findings from the GLORIA-AF registry phase 2. J Am Coll Cardiol. antagonist oral anticoagulants in Primary Care: ACTUA study [Uso de los 2017;69:777–85. anticoagulantes orales de accion directa en atencion primaria: Estudio 36. Esposti LD, Briere JB, Bowrin K, Diego S, Perrone V, Pasquale GD. ACTUA]. SEMERGEN. 2017;43:477–85. Antithrombotic treatment patterns in patients with atrial fibrilla- 52. Agencia Española de Medicamento y Productos Sanitarios (AEMPS). Crite- tion in Italy pre- and post-DOACs: the REPAIR study. Future Cardiol. rios y recomendaciones para el uso de nuevos anticoagulantes orales 2019;15:109–18. (NACO) en la prevención del ictus y la embolia sistémica en pacientes 37. Lacoin L, Lumley M, Ridha E, Pereira M, McDonald L, Ramagopalan S, et al. con fibrilación auricular no valvular. Informe de Posicionamiento Tera- Evolving landscape of stroke prevention in atrial fibrillation within the UK péutico UT/V4/23122013. 2016. Available from: https://​www.​aemps.​gob.​ between 2012 and 2016: a cross-sectional analysis study using CPRD. BMJ es/​medic​ament​osUso​Humano/​infor​mesPu​blicos/​home.​htm#​antic​oagul​ Open. 2017;7:e015363. antes-​orales. 38. Gullon A, Suarez C, Diez-Manglano J, Formiga F, Cepeda JM, Pose A, et al. 53. Olesen JB, Sorensen R, Hansen ML, Lamberts M, Weeke P, Mikkelsen AP, Antithrombotic treatment and characteristics of elderly patients with et al. Non-vitamin K antagonist oral anticoagulation agents in anticoagu- non-valvular atrial fibrillation hospitalized at Internal Medicine depart- lant naive atrial fibrillation patients: Danish nationwide descriptive data ments. NONAVASC registry [Caracteristicas y tratamiento antitrombotico 2011–2013. Europace. 2015;17:187–93. de los pacientes ancianos con fibrilacion auricular no valvular hospi- 54. Urbaniak AM, Strom BO, Krontveit R, Svanqvist KH. Prescription patterns talizados en Medicina Interna. Registro NONAVASC]. Med Clin (Barc). of non-vitamin K oral anticoagulants across indications and factors 2017;148:204–10. associated with their increased prescribing in atrial fibrillation between 39. Poli D, Antonucci E, Grifoni E, Abbate R, Gensini GF, Prisco D. Bleeding 2012–2015: a study from the norwegian prescription database. Drugs risk during oral anticoagulation in atrial fibrillation patients older than 80 Aging. 2017;34:635–45. years. J Am Coll Cardiol. 2009;54:999–1002. 55. Steinberg BA, Shrader P, Thomas L, Ansell J, Fonarow GC, Gersh BJ, et al. 40. Friberg L, Rosenqvist M, Lip GY. Evaluation of risk stratification schemes Factors associated with non-vitamin K antagonist oral anticoagulants for for ischaemic stroke and bleeding in 182 678 patients with atrial stroke prevention in patients with new-onset atrial fibrillation: Results fibrillation: the Swedish Atrial Fibrillation cohort study. Eur Heart J. from the Outcomes Registry for Better Informed Treatment of Atrial Fibril- 2012;33:1500–10. lation II (ORBIT-AF II). Am Heart J. 2017;189:40–7. 41. Lip GY, Lane DA. Assessing bleeding risk in atrial fibrillation with the 56. Ho P, Brooy BL, Hayes L, Lim WK. Direct oral anticoagulants in frail older HAS-BLED and ORBIT scores: clinical application requires focus on the adults: a geriatric perspective. Semin Thromb Hemost. 2015;41:389–94. reversible bleeding risk factors. Eur Heart J. 2015;36:3265–7. 57. Pather M. The prevalence of renal impairment in the elderly hospitalized 42. Mant J, Hobbs FD, Fletcher K, Roalfe A, Fitzmaurice D, Lip GY, et al. population. Int J Gen Med. 2009;2:117–9. Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treat- ment of the Aged Study, BAFTA): a randomised controlled trial. Lancet. Publisher’s Note 2007;370:493–503. Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations.