Page 13 - BH-1-2

Brain & Heart Post-stroke atrial fibrillation and predictive scores

devices are considered the gold standard, detecting was detected, compared to 6.8% in patients without SAF
approximately 30% of SAF over a 3-year follow-up , the (P = 0.363) .
[18]
[20]
choice between non-invasive external devices and invasive One of the central controversies in this context revolves
implantable devices for prolonged ECG monitoring around how to define the duration of a SAF episode for
remains a topic of controversy. The limited availability, diagnostic purposes. Indeed, in five studies, SAF was
invasiveness, and cost of implantable devices restrict their
use in clinical practice. Therefore, the need for tools that defined as any episode, even if it lasted <30 s, while in six
can identify high-risk patients for SAF and prioritize their studies, SAF was defined as episodes lasting at least 30 s. This
prolonged ECG monitoring is evident. issue has sparked a vigorous debate between cardiologists
and stroke physicians. Cardiologists advocate for defining
In recent years, an increasing body of literature has diagnostic SAF episodes as those lasting at least 30 s [21,22] . In
addressed the development of clinical scores to predict the the NOR-FIB study, SAF was defined as an episode lasting
risk of SAF in stroke patients. In this paper, we reviewed at least 2 min . More recently, in the external validation
[20]
the literature focusing on clinical predictive scores for SAF of the AF-ESUS score, Kitsiou et al. defined SAF as an
detection in CS or ESUS. Our analysis identified 11 scores, episode lasting at least 6 min . However, it is important
[23]
with eight designed for CS patients and three for ESUS to acknowledge that the majority of SAF episodes detected
patients. It is worth noting that the majority of these scores following a stroke last for <30 s , underscoring a notable
[24]
were developed with relatively small sample sizes. With the gap in evidence of their clinical significance. Many stroke
exception of studies deriving the HAVOC and AF-ESUS physicians lean toward considering even these shorter SAF
scores [9,16] , the derivation cohorts typically included fewer episodes as diagnostic . As a result, it is not surprising
[25]
than 300 patients, with the scores ranging from 63 to that in approximately half of the studies deriving clinical
296 [7,8,10-15,17] .
predictive scores, SAF episodes of any duration are
Our analysis in this review underscores the robust considered diagnostic.
association between advanced age and the occurrence
of SAF, a factor consistently present in ten out of 11 The predictive power of the retrieved scores is strong,
scores. In addition, LAE, premature atrial beats, and CT with an AUC ranging from 0.72 to 0.94. However, there is a
characteristics such as cortical or subcortical infarcts, lack of prospective studies directly comparing these scores.
echocardiographic or laboratory signs of HF, and NIHSS Four out of the 11 scores were assessed against the predictive
scoring emerge as the most frequently represented variables power of the CHA DS -VASc score. Grifoni et al. conducted
2
2
within these clinical predictive scores. These findings are a comparative analysis, evaluating the performance of the
in agreement with a systematic review of the literature E AF score against CHA DS -VASc and four other scores,
2
2
2
performed by Noubiap et al., which identified age, female including AS5F and Brown ESUS-AF, which were selected
gender, left atrial size, LAE, and the CHA DS -VASc score in our search. Their findings indicated that the E AF score
2
2
2
as independent risk factors for post-stroke SAF . exhibited a significantly better predictive power than AS5F
[18]
and showed an improved predictive power compared to
Our research corroborates the positive relationship
between the duration of ECG monitoring and the rate of the CHA DS -VASc and Brown ESUS-AF scores, although
2
2
[17]
SAF detection. Specifically, the SAF detection rate was the latter difference was not statistically significant .
found to be up to 6% in patients monitored for at least More recently, Ratajczak-Tretel et al. conducted a similar
72 h and increased to 22% in patients monitored for more comparison, pitting the predictive power of the CHA DS -
2
2
than 2 weeks. However, it is important to note that only VASc score against seven clinical predictive scores, three
a minority of the patients (340 of 11904, 2.85%) received of which (HAVOC, AS5F, and Brown ESUS-AF) were
implantable ECG monitoring. selected in our study. Their analysis, performed on the
population enrolled in the NOR-FIB study, revealed that
SAF emerges as the prominent underlying etiology AS5F demonstrated the highest predictive power, with an
during the follow-up of patients with CS or ESUS. In a AUC of 0.741 (95% CI: 0.678 – 0.804) .
[26]
recent study, the NOR-FIB study, SAF was identified as
the probable cause of stroke in 43% of CS cases following We recognize that our review may have certain
a 12-month follow-up, while 57% of strokes remained limitations. The derivation studies vary in terms of design
cryptogenic . Furthermore, SAF was identified as and methodology, and their sample sizes are relatively
[19]
a possible cause in 29% of patients, contributing to a small. Furthermore, a prospective comparison among
substantial 67% of the identified etiologies . Recurrence them is absent. Therefore, it is important to exercise
[19]
rates at the 12-month follow-up were 5.8%, with a slightly caution when considering the implications of our findings
lower rate of 2.7% observed in patients in whom SAF for clinical practice.
Volume 1 Issue 2 (2023) 7 https://doi.org/10.36922/bh.0955