Abstracts

Diagnostic yield of continuous video electroencephalography for paroxysmal vital sign changes in children

Abstract number : 1.020
Submission category : 3. Neurophysiology
Year : 2015
Submission ID : 2325762
Source : www.aesnet.org
Presentation date : 12/5/2015 12:00:00 AM
Published date : Nov 13, 2015, 12:43 PM

Authors :
Louis Dang, Renee Shellhaas

Rationale: American Clinical Neurophysiology Society guidelines suggest that continuous video EEG monitoring (cVEEG) is indicated to determine whether paroxysmal vital sign changes (PVSC) are seizures. However, the diagnostic yield of cVEEG for this purpose is unknown. We hypothesized that PVSCs are rarely due to seizures, especially in the absence of associated non-autonomic symptoms (NAS). We also hypothesized that semiologic features of the index events and the patient’s clinical characteristics could help determine a priori whether the index events are likely to be seizures.Methods: We performed a single-center chart review of all pediatric cVEEGs requested over 10 years for the purpose of determining whether PVSCs were seizures. The types of PVSCs, associated NAS, and relevant patient characteristics were tabulated. Univariate and multivariate analyses were employed to determine which semiologic or patient factors were associated with the index events being seizures.Results: From 1/2005 to 5/2015, 324 cVEEG studies were requested at our center for differential diagnosis of PVSCs (median cVEEG duration 2 calendar days, IQR=1). Most patients were neonates and infants (n=253, 78%). The PVSCs included apnea (n=157, 48%), oxygen desaturation (n=155, 48%), hypertension (n=45, 14%), bradycardia (n=43, 13%), hypotension (n=26, 8%), and tachycardia (n=23, 7%). The index events were recorded in 169/324 studies (53%), and the events were associated with electrographic seizures in 30/169 (18%). Apnea without NAS was recorded in 41 cVEEGs, and only 3 were found to be seizures (7%), while apneic events with NAS were frequently confirmed to be electroclinical seizures (19/42, 45%). Episodic hypertension without NAS (n=12) and hypotensive events (n=12) were never found to be seizures. Semiologic features that increased the odds that the index events were seizures included apnea (odds ratio, OR=3.5; 95% Confidence interval 1.5-8.5; p=0.005), apnea with NAS (OR=8.7; 3.7-20.8; p<0.0001), decreased tone/limpness (OR=8.1; 2.4-27.8; p=0.0008), and eye movements/eye deviation (OR=3.1; 1.7-7.8; p=0.018). Patients with PVSCs who were on an anti-epileptic drug (AED) also had increased odds of the events being seizures (OR=2.9; 1.3-6.5; p=0.009). On multivariate analysis, apnea with associated NAS (OR=7.6; 3.1-18.5; p<0.0001) and decreased/limp tone during the event (OR=5.5; 1.4-21.7; p=0.016) were independently associated with seizures.Conclusions: Since cVEEG is very resource intensive, monitoring strategies to target those at highest risk for seizures are crucial. Apneic events, especially with associated NAS, are more likely than other paroxysmal vital sign changes to be electroclinical seizures. Pediatric patients with apnea, NAS, and decreased tone during their events are quite likely to have seizures and warrant cVEEG. Conversely, among infants and children who do not have a history of epilepsy and are not on AEDs, PVSCs without associated NAS are very unlikely to be due to seizures; such patients might not all require cVEEG.
Neurophysiology