SEARCH

SEARCH BY CITATION

Keywords:

  • atrial fibrillation;
  • clinical trial;
  • Holter ECG;
  • ischaemic stroke;
  • paroxysmal atrial fibrillation;
  • stroke unit

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosure of conflicts of interest
  9. References

Background and purpose

Detection of paroxysmal atrial fibrillation (pAF) after an ischaemic cerebrovascular event is of imminent interest, because oral anticoagulation as a highly effective secondary preventive treatment is available. Whereas permanent atrial fibrillation (AF) can be detected during routine electrocardiogram (ECG), longer detection duration will detect more pAF but might be resource consuming. The current study tried to identify clinical predictors for pAF detected during long-term Holter ECG and clinical follow-up.

Methods

Patients with acute ischaemic stroke were prospectively investigated with an intensified algorithm to detect pAF (7-day Holter ECG, follow-up investigations after 90 days and 1 year).

Results

Two hundred and eighty-one patients were included, 44 of whom had to be excluded since they presented with permanent AF and another 13 patients had to be excluded due to other causes leaving 224 patients (mean age 68.5 years, 58.5% male). Twenty-nine (12.9%) patients could be identified to have pAF during prolonged Holter monitoring, an additional 13 (5.8%) after follow-up investigations. Multivariate analysis identified advanced age [odds ratio (OR) 1.05, 95% confidence interval (CI) 1.01–1.08] as well as clinical symptoms >24 h (OR 5.17, 95% CI 1.73–15.48) and a history of coronary artery disease (OR 3.14, 95% CI 1.35–7.28) to be predictive for the detection of pAF.

Conclusions

In acute stroke patients with advanced age, history of coronary artery disease and clinical symptoms >24 h, a prolonged Holter ECG monitoring and follow-up is warranted to identify pAF. This could increase the detection rate of patients requiring anticoagulation and may be able to reduce the risk of recurrent stroke in the case of successful anticoagulation of these patients.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosure of conflicts of interest
  9. References

Atrial fibrillation (AF) with subsequent cardioembolic stroke constitutes one of the most frequent ischaemic stroke etiologies and accounts for approximately every fourth cerebrovascular ischaemic event [1]. Detection of AF after an ischaemic cerebrovascular event is of imminent interest, because oral anticoagulation as a highly effective secondary preventive treatment is available [2]. Moreover, new direct oral anticoagulants might be even more effective and convenient compared with conventional anticoagulation with warfarin [3]. This may result in an even higher proportion of patients being anticoagulated after cardioembolic stroke thereby reducing subsequent stroke events. Whereas the detection of permanent AF can be easily managed by means of a routine 12-channel surface electrocardiogram (ECG), detection of paroxysmal atrial fibrillation (pAF) is challenging. Currently, monitoring for 24 h is frequently being used as the method of choice during routine stroke work-up to identify pAF [4]. Unfortunately, specific recommendations about the type and duration of monitoring are controversial and definite recommendations are lacking. It was recently shown that longer detection duration increases the yield of pAF detection [5-9]. To date, the best detection strategy for pAF is still unknown, and considering the fact that longer detection duration may be inconvenient for the patients as well as resource consuming, future patient stratification appears expedient. In this context, the identification of variables available from routine diagnostic stroke work-up could be useful to identify those patients in whom an intensified pAF screening algorithm might result in a higher detection rate of pAF with subsequent anticoagulation.

The aim of the current study was to investigate the potential value of clinical factors obtained during routine stroke work-up diagnostics to predict pAF as diagnosed during an intensified screening algorithm which included a prolonged 7-day Holter ECG and subsequent follow-up investigations up to 1 year after acute ischaemic stroke.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosure of conflicts of interest
  9. References

Trial Registration Information: http://controlled-trials.com/. Unique identifier: ISRCTN46104198.

Study design

In this prospective single-center trial [6] patients with symptoms of stroke starting <24 h ago were asked to give written informed consent for participation in the study. Participants were included after presenting in the emergency department of the University of Göttingen between March 2009 and February 2010 after signs of a haemorrhagic stroke had been ruled out by cerebral imaging. All patients received routine stroke care (cerebral imaging: computer tomography or magnetic resonance imaging), cardiovascular work-up including 12-channel surface ECG, transthoracic echocardiography and Holter ECG, ultrasound of the brain-supplying arteries, and routine blood analysis including, for example, lipid profile, HbA1c, C-reactive protein and creatinine, on a certified stroke unit. In addition to the routine work-up, all patients who presented in sinus rhythm received a prolonged 7-day Holter monitoring (CardioMem® CM 3000; getemed Medizin- und Informationstechnik, Teltow, Germany) [6]. All patients were scheduled to receive follow-up investigations after 3 months (telephone contact) and 1 year (physician contact with a structured interview and 12-channel surface ECG).

Data collection

Traditional cardiovascular risk factors and patients' history were recorded in a predefined data sheet. Classification of stroke etiology was done by two experienced stroke clinicians (J. W., K. G.) using the TOAST (Trial of Org 10172 in Acute Stroke Treatment) classification scheme [10] combined with the results from a 24-h Holter ECG measurement (result of the analysis of day 4 of the long-term Holter monitoring). Neuroimaging data (CT or MRI) were evaluated by board-certified neuroradiologists and study personnel assessing the occurrence of acute or old vascular lesions with respect to their vascular territory [11]. When both imaging modalities were available for a patient, MRI results were preferably used for interpretation. Transthoracic echo cardiography was offered to all patients and standard measurements such as left atrial dimension (LAD), left ventricular ejection fraction (LV EF), left ventricular end-diastolic dimension (LV EDD) and left ventricular end-systolic dimension (LV ESD) were recorded.

During follow-up investigations, detailed information regarding hospitalization or ambulant physician contact due to the occurrence of new pAF in the meantime was obtained. In the case of occurrence of pAF, source documentation (12-channel surface ECG) was retrieved and evaluated by two experienced cardiologists. The primary outcome for the current analysis was the new occurrence of pAF and was defined as a documented period longer than 30 s [12] of AF up to the 1-year follow-up investigation in patients who presented in sinus rhythm in the admission 12-channel surface ECG.

This study is in accordance with the Declaration of Helsinki and current ICH/GCP guidelines. Written informed consent was obtained from all participants or relatives. The study was approved by the Ethics Committee of the University of Göttingen.

Statistical analysis

Continuous values are expressed as mean ± standard deviation (SD) and nominal variables as count and percentages. Median values with the corresponding interquartile range (IQR) were computed for non-normally distributed variables. A two-sided t-test was used for comparison of normally distributed variables and the non-parametric Kruskal–Wallis test for not normally distributed values. For comparisons of categorical data two-tailed chi-squared statistics with Yates's correction or the Fisher exact test were used as applicable. Multiple stepwise binomial regression analyses were conducted for those variables with P < 0.01 on a univariate level to estimate a potential effect on the prediction of stroke outcome (P to enter 0.05, P to leave 0.1) using additive and multiple interaction terms. Receiver operator characteristic (ROC) curves and area under the curve (AUC) were calculated for estimation of incremental prognostic information on pAF prediction of the variables which remained significant after multiple regression analysis. P < 0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed using SPSS (Version 19; SPSS Inc., Chicago, IL, USA) unless otherwise stated.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosure of conflicts of interest
  9. References

Two hundred and eighty-one patients (47.5% of all admitted patients eligible for the study by fulfilling all study criteria) gave written informed consent to participate in the study and received a 12-channel surface ECG; of these 44 presented with atrial fibrillation and were excluded from further analysis. For the present study, an additional seven patients had to be excluded because the diagnosis of suspected ischaemic stroke was not confirmed during hospitalization, five patients had incomplete follow-up data and one patient withdrew consent (Fig. 1). The patient characteristics of the remaining 224 patients (mean age 68.5 years, 58.5% male) included in the final analysis and divided into two groups according to the occurrence of pAF are presented in Table 1. In 29 (12.9%) patients, pAF was diagnosed during initial prolonged (median 6.7 days, IQR 4.4–7.0) Holter monitoring. The median duration of the detected pAF was 12.9 h and ranged from 4.4 to 19.7 h (IQR). An additional eight (3.6%) and five (2.2%) pAF episodes were detected in follow-up investigations after 90 days and 1 year, respectively. In 182 patients, no AF could be detected across the study period.

Table 1. Baseline characteristics of the study population
Variable No AFParoxysmal AFP-value
  1. Final study population, n = 224. AF, atrial fibrillation; BP, blood pressure; TIA, transient ischaemic attack; LDL, low density lipoprotein; HDL, high density lipoprotein; CRP, C-reactive protein; CCT, cranial computer tomography; MRI, magnetic resonance imaging; SAE, subcortical arteriosclerotic encephalopathy; LAD, left atrial diameter; LV EF, left ventricular ejection fraction; LV EDD, left ventricular end-diastolic diameter; LV ESD, left ventricular end-systolic diameter; TOAST, Trial of Org 10172 in Acute Stroke Treatment. *Variables included in multivariate analysis.

N 18242 
Male sex (%)108 (59.3)23 (54.8)0.606
Age67 (±13.0)74 (±10.5)<0.001*
Symptoms >24 h (%)113 (62.1)38 (90.5)<0.001*
Thrombolysis (%)8 (4.4)7 (16.7)0.010
BPsys mmHg143 (±22.1)149 (±25.7)0.119
BPdia mmHg79 (±12.4)79 (±13.4)0.885
Heart rate/min73 (±12.8)69 (±15.6)0.069
Arterial hypertension (%)130 (71.4)34 (81.0)0.249
Diabetes mellitus (%)41 (22.5)9 (21.4)1.000
Heart failure (%)9 (4.9)3 (7.1)0.702
Tobacco use (active or former) (%)80 (44)12 (28.6)0.082
Hyperlipidemia (%)58 (31.9)20 (47.6)0.072
Coronary artery disease (%)19 (10.4)15 (35.7)<0.001*
History of TIA/stroke (%)46 (25.3)8 (19.0)0.548
Blood analysis
Cholesterine (mM/l)5.1 (±1.2)5.0 (±1.2)0.586
LDL (mM/l)3.4 (±1.0)3.3 (±1.0)0.810
HDL (mM/l)1.3 (±0.3)1.3 (±0.4)0.928
Triglycerides (mM/l)1.5 (±1.7)1.3 (±1.2)0.116
HbA1c (%)5.9 (5.7–6.4)6.9 (5.7–6.4)0.405
Creatinine (mg/dl)0.94 (±0.4)1.14 (±0.9)0.032
CRP (mg/l)2 (2–4.3)2.6 (2–8.7)0.046
Ancillary diagnostic
CCT/MRI with an old visible lesion according to the qualifying clinical event (%)
Same vascular territory24 (13.2)8 (19.0)0.180
Additional vascular territory22 (12.1)7 (16.7)
No lesion visible86 (47.3)12 (28.6)
Only SAE visible50 (27.5)15 (35.7)
Extracranial duplex-sonography (%)
Carotid stenosis >50%20 (11.0)4 (9.5)0.183
Plaques89 (48.9)27 (64.3)
Unremarkable73 (40.1)11 (26.2)
Transthoracic echo cardiography
LAD (mm)40.9 (±6.4)44.0 (±6.6)0.008*
LV EF (%)61.7 (±9.3)59.8 (±9.9)0.272
LV EDD (mm)46.5 (±6.7)46.0 (±8.2)0.697
LV ESD (mm)46.5 (±6.7)46.0 (±8.2)0.996
TOAST classification scheme (%)
Large-artery disease27 (14.8)8 (19)0.009*
Small-vessel occlusion61 (33.5)10 (23.8)
Cardioembolism19 (10.4)13 (31.0)
Other etiology5 (2.7)1 (2.4)
Undetermined etiology70 (38.5)10 (23.8)
image

Figure 1. Patient selection flow chart.

Download figure to PowerPoint

Differences between the two study populations are presented in Table 1. Patients with pAF were older (74 ± 10.5 vs. 67 ± 13.0; P < 0.001), presented more frequently with symptoms of an ischaemic stroke lasting longer than 24 h (90.5% vs. 62.5%, P < 0.001), received recombinant tissue plasminogen activator more frequently (16.7% vs. 4.4%; P = 0.01) and had a manifold history of coronary artery disease compared with those without pAF. During routine stroke work-up, baseline blood analysis showed elevated creatinine (1.14 ± 0.9 vs. 0.94 ± 0.4 mg/dl; P = 0.032) and C-reactive protein values [median 2.6 (IQR 2–8.7) vs. median 2 (IQR 2–4.3) mg/l; P = 0.046] in the group with pAF. Routine transthoracic echocardiography was available for 90.2% of the patients and showed a higher LAD (44.0 ± 6.6 vs. 40.9 ± 6.4 mm; P = 0.008) in patients with pAF. There was a difference in stroke etiology as classified by the TOAST classification scheme, with a significantly higher proportion of suspected cardioembolic strokes in the group with pAF (31% pAF vs. 10.4% no AF; P = 0.009) but a lower proportion of pAF in patients with stroke of undetermined etiology.

After applying binomial multivariate analysis with variables which were imbalanced (P < 0.01) on the univariate level (age, symptoms >24 h, coronary artery disease, LAD, TOAST classification), the variables age [odds ratio (OR) 1.05, 95% confidence interval (CI) 1.01–1.08, P < 0.011], symptoms >24 h (OR 5.17, 95% CI 1.76–15.48, P = 0.003) and history of coronary heart disease (OR 3.14, 95% CI 1.35–7.28, P = 0.008) remained significant for predicting pAF during prolonged Holter ECG and 1-year follow-up (Table 2). ROC curves for all significant variables after multiple logistic regression analysis resulted in an AUC of 0.618 (95% CI 0.515–0.721; P = 0.018), 0.666 (95% CI 0.579–0.754; P = 0.001) and 0.643 (95% CI 0.559–0.727; P = 0.004) for history of coronary heart disease, age and symptoms lasting >24 h as the qualifying event, respectively. The combination of all variables yielded an AUC of the ROC analysis of 0.774. The increasing detection rate of pAF with ongoing age is displayed in Fig. 2.

Table 2. Results of the multivariate model
VariableOR95% CIP-value
  1. OR, odds ratio; CI, confidence interval.

Age (per year)1.051.01–1.080.011
Symptoms >24 h5.171.76–15.490.003
Coronary artery disease3.141.35–7.280.008
image

Figure 2. Percentage (of the total study population) of paticipants with paroxysmal atrial fibrillation (black) and free of atrial fibrillation (grey) in different age quartiles.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosure of conflicts of interest
  9. References

The aim of our study was to evaluate the value of clinical factors which might be useful to predict those patients in whom pAF could be detected after extensive diagnostic work-up. Age, symptoms >24 h (compared with transient symptoms as a qualifying event) and a history of coronary artery disease were found to be predictive for pAF. With this clinical information it may be possible to induce a more extensive search for pAF, because even very short and intermittent episodes of AF may be associated with an elevated risk for subsequent stroke [13, 14].

Our results are in line with previous studies in which advanced age (Fig. 2) could be shown to be associated with a higher incidence of AF and pAF after stroke [1, 11, 15-19]. In the retrospective study of Alhadramy et al. [11], all patients in whom pAF could be detected after 24-h Holter monitoring were aged above 55 years, and Doliwa Sobocinski et al. [15] reported similar results with an additional twice-daily ECG recording for up to 30 days: the prevalence of pAF in all patients was 6.8% and increased in stroke patients aged ≥75 years to 11.8%. This argues for a structural abnormality of the atria with age which may lead to electrical instability thus causing pAF. Clinical diseases such as coronary artery disease or insufficiency are frequently observed comorbidities, which is concordant with an analysis of Bugnicourt et al. [18] and our results showing a history of coronary artery disease as a relevant predictor for pAF. Routine echocardiography might be suitable to reveal these early structural changes, because they are commonly clinically inconspicuous and challenging to diagnose. Because an echocardiography is frequently indicated and carried out during routine stroke work-up to identify structural changes, potential embolic sources or as a prognostic parameter, subclinical alterations may thereby be detected and could be predictive for subsequent pAF [18, 20]. Left atrium dilatation as one parameter easy to determine could be identified to be useful to predict pAF within our cohort on the univariate level. This is in line with previously published investigations, showing a dilated left atrium as an independent predictor for the development of AF within the common population [21] as well as in stroke subgroups [16-18]. However, after multivariate analysis this parameter did not contribute significantly to our final predictive model. An explanation might be that pAF frequently converts into permanent AF with ongoing disease duration, which then leads to a further dilatation of the atrium. Because we excluded all patients with a history of permanent AF, most of our patients might have been identified with pAF during a very early stage of disease and therefore with only beginning left atrium dilatation, as this is expected to become more pronounced with AF disease progression. However, a more elaborate analysis of echocardiographic measurements, which is currently not yet routinely incorporated in most of the routine stroke work-up diagnostics, has recently been shown to be useful for the prediction of pAF and may thus be worthwhile integrated into future routine diagnostic measurements [22].

Cardioembolic strokes are frequently larger than those due to other sources according to the TOAST classification scheme [1], which is supported within our cohort: symptoms >24 h as a qualifying event compared with only transient and frequently unremarkable symptoms showed the strongest association with subsequent pAF detection. A further explanation might be that, despite the fact that the diagnosis was determined by experienced vascular neurologists, stroke mimics are more frequent in patients with transient ischaemic attack, and this might have biased this subgroup of patients. Of interest and contrary to earlier results by Alhadramy et al. [11, 19] and other research groups, cerebral neuroradiological imaging obtained from routine diagnostics was not found to be useful to predict pAF in our cohort. In the above mentioned study, acute and chronic infarcts, especially as seen on MRI, were associated with pAF [11]. Unfortunately, MRI was not done for the majority of our patients and in most cases only acute cranial CT imaging was done without confirmatory cranial CT on a routine basis which might explain the low incidence of acute visible brain lesions in our sample and the missing predictive value of this routine diagnostic imaging parameter.

The strengths of the current analysis are the extensive and prolonged Holter monitoring, the long patient follow-up and the fact that our model was restricted only to pAF patients. Recently, efforts have been made to identify predictors for pAF but many of these studies also included patients with permanent AF, which could easily be diagnosed by a 12-channel surface ECG and therefore usually does not require any additional effort due to extensive work-up [16, 23]. Data with patients presenting in sinus rhythm are still sparse and especially thorough and long monitoring has not been done in the studies assessing this patient subgroup [24, 25]. The challenging question of whether patients with subclinical AF and especially those with only short episodes of pAF have to be anticoagulated remains unanswered for the time being. Whether the number of subclinical AF episodes, the burden of AF (percentage of time spent in AF divided by total time) or the duration of the longest AF episode is the best predictor for subsequent stroke is currently unknown because those patients have only rarely been included in previous anticoagulation trials [26]. Moreover, anticoagulation in older patients with severe disabling strokes may be of limited therapeutic benefit and therefore has to be further evaluated in future trials.

Despite the strengths, there are some limitations: because our analysis focused on the usefulness of diagnostics as gathered during routine stroke work-up, a more detailed diagnostic including, for example, additional biomarkers [27] or consecutive cerebral MRI could have further improved the predictive yield. Due to the retrospective analysis of the prospective data and the participation of 47.5% of all possibly eligible patients a slight bias cannot be excluded.

In conclusion, our data support the growing knowledge of clinical variables which can easily be obtained during standardized routine stroke unit work-up to identify those patients who might have pAF and in whom a more intensified cardiac monitoring could be initiated to identify this condition. A prospective study to validate these results is recommended and especially an assessment of whether a higher detection rate of pAF is eventually associated with a lower incidence of stroke re-occurrence due to an increased use of oral anticoagulants.

Acknowledgement

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosure of conflicts of interest
  9. References

This study was supported by an unrestricted grant from Roche Diagnostics. This work was supported by grants from the German Federal Ministry of Education and Research [German Heart Failure Network, TP 7 (FKZ 01GI0205)].

Disclosure of conflicts of interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosure of conflicts of interest
  9. References

The authors declare no financial or other conflicts of interest

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Disclosure of conflicts of interest
  9. References
  • 1
    Grau AJ, Weimar C, Buggle F, et al. Risk factors, outcome, and treatment in subtypes of ischemic stroke: the German stroke data bank. Stroke 2001; 32: 25592566.
  • 2
    Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med 2007; 146: 857867.
  • 3
    Alberts MJ, Eikelboom JW, Hankey GJ. Antithrombotic therapy for stroke prevention in non-valvular atrial fibrillation. Lancet Neurol 2012; 11: 10661081.
  • 4
    Adams HP Jr, del Zoppo G, Alberts MJ, et al. Guidelines for the early management of adults with ischemic stroke: a guideline from the American Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups: the American Academy of Neurology affirms the value of this guideline as an educational tool for neurologists. Stroke 2007; 38: 16551711.
  • 5
    Liao J, Khalid Z, Scallan C, et al. Noninvasive cardiac monitoring for detecting paroxysmal atrial fibrillation or flutter after acute ischemic stroke: a systematic review. Stroke 2007; 38: 29352940.
  • 6
    Stahrenberg R, Weber-Krüger M, Seegers J, et al. Enhanced detection of paroxysmal atrial fibrillation by early and prolonged continuous Holter monitoring in patients with cerebral ischemia presenting in sinus rhythm. Stroke 2010; 41: 28842888.
  • 7
    Gumbinger C, Krumsdorf U, Veltkamp R, et al. Continuous monitoring versus Holter ECG for detection of atrial fibrillation in patients with stroke. Eur J Neurol 2012; 19: 253257.
  • 8
    Ziegler PD, Glotzer TV, Daoud EG, et al. Incidence of newly detected atrial arrhythmias via implantable devices in patients with a history of thromboembolic events. Stroke 2010; 41: 256260.
  • 9
    Elijovich L, Josephson SA, Fung GL, et al. Intermittent atrial fibrillation may account for a large proportion of otherwise cryptogenic stroke: a study of 30-day cardiac event monitors. J Stroke Cerebrovasc Dis 2009; 18: 185189.
  • 10
    Adams HP Jr, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993; 24: 3541.
  • 11
    Alhadramy O, Jeerakathil TJ, Majumdar SR, et al. Prevalence and predictors of paroxysmal atrial fibrillation on Holter monitor in patients with stroke or transient ischemic attack. Stroke 2010; 41: 25962600.
  • 12
    Wann LS, Curtis AB, January CT, et al. 2011 ACCF/AHA/HRS focused update on the management of patients with atrial fibrillation (updating the 2006 guideline): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2011; 123: 104123.
  • 13
    Healey JS, Connolly SJ, Gold MR, et al. Subclinical atrial fibrillation and the risk of stroke. N Engl J Med 2012; 366: 120129.
  • 14
    Hohnloser SH, Pajitnev D, Pogue J, et al. Incidence of stroke in paroxysmal versus sustained atrial fibrillation in patients taking oral anticoagulation or combined antiplatelet therapy: an ACTIVE W Substudy. J Am Coll Cardiol 2007; 50: 21562161.
  • 15
    Doliwa Sobocinski P, Anggardh Rooth E, Frykman Kull V, et al. Improved screening for silent atrial fibrillation after ischaemic stroke. Europace 2012; 14: 11121116.
  • 16
    Suissa L, Bertora D, Lachaud S, et al. Score for the targeting of atrial fibrillation (STAF): a new approach to the detection of atrial fibrillation in the secondary prevention of ischemic stroke. Stroke 2009; 40: 28662868.
  • 17
    Malik S, Hicks WJ, Schultz L, et al. Development of a scoring system for atrial fibrillation in acute stroke and transient ischemic attack patients: the LADS scoring system. J Neurol Sci 2011; 301: 2730.
  • 18
    Bugnicourt JM, Flament M, Guillaumont MP, et al. Predictors of newly diagnosed atrial fibrillation in cryptogenic stroke: a cohort study. Eur J Neurol 2012; 20: 13521359.
  • 19
    Lazzaro MA, Krishnan K, Prabhakaran S. Detection of atrial fibrillation with concurrent Holter monitoring and continuous cardiac telemetry following ischemic stroke and transient ischemic attack. J Stroke Cerebrovasc Dis 2012; 21: 8993.
  • 20
    Vaziri SM, Larson MG, Benjamin EJ, et al. Echocardiographic predictors of nonrheumatic atrial fibrillation. The Framingham Heart Study. Circulation 1994; 89: 724730.
  • 21
    Kannel WB, Wolf PA, Benjamin EJ, et al. Prevalence, incidence, prognosis, and predisposing conditions for atrial fibrillation: population-based estimates. Am J Cardiol 1998; 82: 2N9N.
  • 22
    Stahrenberg R, Edelmann F, Haase B, et al. Transthoracic echocardiography to rule out paroxysmal atrial fibrillation as a cause of stroke or transient ischemic attack. Stroke 2011; 42: 36433645.
  • 23
    Kobayashi A, Iguchi M, Shimizu S, et al. Silent cerebral infarcts and cerebral white matter lesions in patients with nonvalvular atrial fibrillation. J Stroke Cerebrovasc Dis 2012; 21: 310317.
  • 24
    Rizos T, Guntner J, Jenetzky E, et al. Continuous stroke unit electrocardiographic monitoring versus 24-hour Holter electrocardiography for detection of paroxysmal atrial fibrillation after stroke. Stroke 2012; 43: 26892694.
  • 25
    Kallmunzer B, Breuer L, Hering C, et al. A structured reading algorithm improves telemetric detection of atrial fibrillation after acute ischemic stroke. Stroke 2012; 43: 994999.
  • 26
    Wachter R, Stahrenberg R, Gröschel K. Subclinical atrial fibrillation: how hard should we look? Heart 2013; 99: 151153.
  • 27
    Wachter R, Lahno R, Haase B, et al. Natriuretic peptides for the detection of paroxysmal atrial fibrillation in patients with cerebral ischemia – the Find-AF study. PLoS ONE 2012; 7: e34351.