Rational:
Although patient management and treatment has greatly improved in recent years, atrial
fibrillation (AF) remains the most common arrhythmia related with stroke, heart failure,
sudden death and cardiovascular morbidity in the global population1. Considering the general
adult population, the prevalence of AF is less than 1%, while in people aged 75-84 it reaches
7%2,3, and is expected to increase in the elderly population and in patients with certain
predisposing risk factors4-10. Recent studies show that 20-30% of stroke survivors are likely
related to AF11-14 and, if not treated properly, it can lead to serious consequences such as
future thromboembolic complications, recurrence of stroke episodes or death15-17. The UK
National Clinical Guidelines for Stroke Patient Management, provided by the Royal College of
Physicians18, recommend cardiac monitoring for at least 12 hours for people with IS or TIA
who would be eligible for secondary prevention treatment for AF and at least 24 hours for
monitoring for patients with no other known cause of stroke.
The 2016 ESC guidelines for the treatment of AF instead recommend monitoring of at least 72
hours in patients with IS or TIA, for example 72 hours of Holter monitoring after the stroke
event12,19, or for longer periods20,21. Given the intermittent nature of AF and its
inclination to be asymptomatic, it would be useful to test the hypothesis that this last
approach could lead to a better diagnosis strategy of AF and consequently could clarify the
relationship between AF and stroke.
Detection of AF episodes has improved with the rapid development of technology in the medical
field, and recently there has been an increase in interest in evaluating whether brief atrial
fibrillation (BAF) episodes, less than 30 seconds, carry a similar risk to those of greater
length and whether these constitute an indication for treatment. Although new research is
focusing on detecting BAF episodes, uncertainty persists in determining whether or not these
episodes increase the risk of stroke, as well as the treatment needed for the aforementioned
range of patients.
A recent study in patients with first episode of IS or TIA and presence of AF detected by
standard ECG concluded that in 6.3% of 869 cases AF was paroxysmal22. The definition of
paroxysmal atrial fibrillation (PA AF) in this specific case was related to those episodes of
AF that spontaneously converted into sinus rhythm. Similarly, in another study from 2010,
previously undiagnosed PA AF was found in 9.2% of the population with IS or TIA23; in the
latter study, the mean monitoring time was 22.6 hours and the presence of only episodes of
BAF was detected in 72% of the people diagnosed with PA AF while the rest of the patients
also had longer episodes of 30 seconds. Similar results were obtained in an AF study in the
population with IS or TIA after mobile cardiac telemetry in which 85% of device-reported
episodes were found to be less than 30 seconds24. These results encourage researchers to
focus more on the subject, therefore in this study we want to investigate the real prevalence
of BAF. The results could guide clinicians to determine the most appropriate preventive and
curative interventions for this group of patients.
The international consortium "My-Atria" - which brings together universities, companies and
hospitals from five European countries including Italy, Spain, Germany, Sweden and the
Netherlands - aims to develop a multidisciplinary network to develop research and training
"without barriers" between academic, industrial and clinical entities in the area of atrial
diseases. Mortara Instrument EU company, as a member of the My-Atria Consortium, and the UOC
of Neurology and Stroke Network of the Ospedale Maggiore in Bologna, are both interested in
the relationship between cardiovascular risk and BAF. It therefore seems logical to join
forces and move together in the same direction by proposing a protocol with the aim of
studying whether BAF represents a biomarker of increased cardiovascular risk and consequently
investigating the clinical relevance and modality of a screening process.
New methodologies are needed to detect and predict even brief episodes of AF, so there is a
need to collect a new database of multiple, high resolution, long-term monitoring ECG
signals. The data will be used by the members of the My-Atria Consortium for research
purposes, for the development and validation of arrhythmia detection algorithms.
Objectives:
The main outcome of this study will be to compare the prevalence of PA AF episodes longer
than 30 seconds versus the prevalence of PA AF episodes including episodes shorter than 30
seconds, in adult patients who survived IS or TIA with no previous history of AF, after 7-day
Holter monitoring. Based on the previous assumptions, the secondary objectives of the project
will be:
To compare the prevalence of patients diagnosed with BAF after 24-hour, 72-hour, and
7-day Holter monitoring.
To study the feasibility of using information obtained from the ECG during sinus rhythm,
for example the characteristics of the P-wave, to predict the presence of PA AF and BAF.
To Collect a new database of multi-monitored, high-resolution, long-term ECG signals.
The data will be used by the My-Atria consortium for research purposes, for the
development and validation of arrhythmia detection algorithms.
The results of the study are intended to have an impact on how AF is screened, diagnosed, and
treated. The results will also contribute to the choice of the most appropriate secondary
stroke prevention modality, which therefore refers to the treatment of subjects who have
already had an episode of IS or TIA.
Study design:
Observational study with transversal design.
Population:
The study population concerns adult subjects admitted to the Neurology and Stroke Network
UOC, diagnosed with IS or TIA (confirmed through standardized clinical practice).
The enrollments will be made for a total number of about 660 patients. Considering that, in
the UOC of Neurology and Metropolitan Stroke Network at the Ospedale Maggiore in Bologna
there is an absolute number of patients with stroke or TIA episodes that is equivalent to
about 800 per year, the study considers an enrollment phase lasting about one year.
Main parameters / endpoints:
Primary endpoint
Absolute number of AF episodes (i.e. all AF episodes shorter than seven days) and BAF in
adult patients with recent IS or TIA with no previous history of AF, in order to identify and
compare prevalence rates to determine the clinical relevance of any BAF screening process.
Secondary endpoint
Absolute number of BAF episodes detected with Holter 24 hours, 72 hours and 7 days after
recent IS or TIA event, with the aim of comparing detected prevalence rates to support the
hypothesis that the prevalence of BAF will increase in patients subjected to a longer
continuous monitoring time.
Extent of the burden on clinical staff and risks and benefits associated with participation
in the study:
The study is non-interventional, involves the application of two Holter devices for a total
duration of 7 days in addition to routine clinical practice (which involves monitoring, but
without continuous recording via multi-parameter monitor).
Enrolled patients will be treated according to the normal diagnostic and therapeutic
procedure and will not directly benefit from participation in this study. Participation in
the study could be associated with the risk of mild adverse skin reactions due to the
presence of electrodes of Holter devices, to be applied to the upper chest of the subject or
to patient discomfort (the materials used are already commercially available, tested and used
in normal clinical practice). Nursing staff will be able to remove the devices if the
discomfort for the patient becomes intolerable. For the clinical staff, the burden will
mainly be due to the provision of an informed consent form (ICF), the compilation of the data
collection form (CRF) and the application, decontamination and data download of Holter
devices. The data collected will be made anonymous in accordance with the laws and
regulations on privacy (GDPR).