Abstract Details

Title Predicting Stroke Mimics in Telestroke Consultations

Topic Cerebrovascular Disease and Interventional Neurology

Presentation(s) Cerebrovascular Disease and Interventional Neurology Posters (7:00 AM-5:00 PM)

Poster/Presentation
Number 166

Objective NA

Background Telestroke consultation is increasingly used to provide stroke care. Much like in-person stroke consults, stroke mimics are common. This study sought to identify patient and hospital characteristics more likely to be associated with a Telestroke diagnosis of stroke mimic.

Design/Methods We analyzed 2 years of video consults by the VA National Telestroke Program (NTSP). Stroke mimic was defined as a Telestroke consult coded as a diagnosis of “other.” Text responses for “other” diagnoses were grouped into clinical categories. We used Chi-squared and t-tests analysis to compare characteristics of patients with a stroke mimic diagnosis and those without. Co-variates included age, gender, co-morbid conditions, NIHSS, time of consult (night/weekend vs day), location of consult (emergency department vs inpatient), hospital rurality; hospital consult volume, and duration of institutional participation in NTSP at time of consult. Variables with a p-value < 0.25 in bivariate analysis were included in the multivariate model.

Results There were 561 stroke mimics. The most common mimics were toxic metabolic encephalopathy (19%) and seizure (12%). Variables significantly associated with stroke mimic in bivariate analyses were age, gender, history of alcohol abuse, history of atrial fibrillation, history of dementia, NIHSS, nights/weekend consults, and hospital rurality. In multivariate analyses, female sex [OR=1.63, p=0.001], inpatient consultations [OR=1.55, p= 0.019], history of dementia [OR=1.85, p=0.0002], and alcohol abuse [OR=1.42, p=0.002] were associated with a stroke mimic. Consults during nights/weekends [OR=0.76, p=0.001], and patients with atrial fibrillation [OR=0.81, p=0.031], increasing age [OR=0.90, p=0.019], and increasing NIHSS [OR=0.97, p=0.0042] were less likely to be a mimic.

Conclusions Patient and consult characteristics influenced the likelihood of a stroke mimic diagnosis. Medical history may reflect conditions likely to cause neurologic symptoms, and patients hospitalized with new symptoms represent a challenging subset to accurately distinguish stroke from mimics. Awareness of these factors may alert providers to diagnoses other than stroke.

Authors/Disclosures
Jennifer Siriwardane, MD (Yuma Regional Medical Center) PRESENTER	Dr. Siriwardane has nothing to disclose.
Linda S. Williams, MD, F�鶹��ýӳ�� (Roudebush VAMC)	The institution of Dr. Williams has received research support from VA HSR&D.
Holly Martin	Holly Martin has nothing to disclose.
	No disclosure on file
	No disclosure on file
	No disclosure on file
Laura Myers	The institution of Laura Myers has received research support from VA.
Glenn D. Graham, MD, PhD, F�鶹��ýӳ��	Dr. Graham has received personal compensation in the range of $10,000-$49,999 for serving as a Speaker with MER (non-profit CME provider). Dr. Graham has received personal compensation in the range of $500,000-$999,999 for serving as a Employee with Department of Veterans Affairs. An immediate family member of Dr. Graham has received personal compensation in the range of $100,000-$499,999 for serving as a Exployee with Department of Veterans Affairs.
Sharyl R. Martini, MD, PhD (VHA Neurology)	Dr. Martini has nothing to disclose.
Jane Anderson (Veterans Health Administration)	The institution of Jane Anderson has received research support from UT Systems.

�鶹��ýӳ��

�鶹��ýӳ��