Pediatric Stroke and Its Mimics: Limitations of a Pediatric Stroke Clinical Pathway

Accepted Manuscript Title: Pediatric Stroke and Its Mimics: Limitations of a Pediatric Stroke Clinical Pathway Author: Amy M. DeLaroche, Lalitha Sivaswamy, Ahmad Farooqi, Nirupama Kannikeswaran PII: DOI: Reference:

S0887-8994(17)30697-5 https://doi.org/doi:10.1016/j.pediatrneurol.2017.10.005 PNU 9246

To appear in:

Pediatric Neurology

Received date: Revised date: Accepted date:

30-6-2017 29-9-2017 6-10-2017

Please cite this article as: Amy M. DeLaroche, Lalitha Sivaswamy, Ahmad Farooqi, Nirupama Kannikeswaran, Pediatric Stroke and Its Mimics: Limitations of a Pediatric Stroke Clinical Pathway, Pediatric Neurology (2017), https://doi.org/doi:10.1016/j.pediatrneurol.2017.10.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

PEDIATRIC STROKE AND MIMICS Pediatric stroke and its mimics: Limitations of a Pediatric Stroke Clinical Pathway Amy M. DeLaroche, MBBS1, Lalitha Sivaswamy MD1, Ahmad Farooqi PhD2, Nirupama Kannikeswaran, MD1 Affiliations: 1Department of Pediatrics, Division of Emergency Medicine, Children’s Hospital of Michigan, 3901 Beaubien St, Detroit, Michigan, 48201; 2School of Medicine, Wayne State University, 540 E Canfield St, Detroit, Michigan, 48201 Email Addresses:

AD – [email protected] LS – [email protected] AF – [email protected] NK – [email protected]

Address correspondence to: Amy DeLaroche, Department of Pediatrics, Division of Emergency Medicine, Children’s Hospital of Michigan, 3901 Beaubien Street, Detroit, Michigan, 48201, 313-745-5260 (phone), 313-993-7166 (fax), [email protected] Word Count:

2756

Funding Source:

No funding was secured for this study.

Financial Disclosure:

The authors have no financial relationships relevant to this article to disclose.

Conflicts of interest:

The authors have no conflicts of interest to disclose.

Keywords:

pediatrics; stroke; clinical pathway; stroke mimic

1

Page 1 of 22

PEDIATRIC STROKE AND MIMICS Abstract

Background: Acute stroke protocols improve delivery of care but it is unclear whether these resource intensive protocols are able to differentiate stroke from mimics in children. The aim of this study is to describe our institution's experience with stroke mimics identified through our pediatric stroke clinical pathway (PSCP). Methods: The PSCP was implemented in our level 1 pediatric emergency department in June 2014 for children aged 1 month to 18 years. For patients managed using the PSCP from June 2014 to December 2016 demographic and clinical data were compared for patients diagnosed with stroke or a stroke mimic. Results: A total of 59 children were evaluated with the PSCP. Fourteen children were identified as having a stroke and 45 children had stroke mimics. The most common stroke mimics were functional neurologic disorders (20.0%), transient neurological deficits (17.8%), migraine (15.6%) and seizure (11.1%). Patient demographics and time to neuroimaging did not differ between those with and without stroke. Vomiting was commonly reported by patients with stroke (OR: 4.00, 95%CI: 1.12 to 14.35) whereas weakness was not (OR: 0.7, 95%CI: 0.07 to 0.90), but the physical examination did not differ between patients with and without stroke. Conclusion: The PSCP ensures timely evaluation of patients presenting with neurological deficits, but fails to reliably differentiate between patients with stroke and those with stroke mimics. Further multi-centered studies are needed to develop a “stroke screen” that reliably distinguishes pediatric stroke from its mimics.

2

Page 2 of 22

PEDIATRIC STROKE AND MIMICS Introduction Diagnostic delays in the identification and management of pediatric stroke remain an important challenge given the high personal and societal costs associated with this diagnosis.1 An important initiative which streamlines and improves the timely delivery of care to pediatric patients with stroke has been the development and implementation of acute stroke protocols.2-4 These protocols enhance provider awareness and are an important step toward remedying diagnostic delays due to under-recognition of pediatric stroke.5,6 However, it is unclear whether these protocols address the challenges associated with the non-specific clinical presentation of pediatric stroke.7

Differentiating pediatric stroke from stroke mimics is difficult due to the considerable overlap in the presenting signs and symptoms.7,8 In one prospective observational study, headache, vomiting, and focal weakness were common clinical features for children presenting to the emergency department with focal neurological symptoms yet, 93% were diagnosed with a stroke mimic.8 More recent data suggests that being well the week prior to presentation, an inability to walk, and face and arm weakness helps to differentiate pediatric stroke from mimics, but providers continue to have difficulties making a clinical diagnosis of pediatric stroke.9,10 Unfortunately, bedside screening tools that utilize presenting signs and symptoms to initiate an acute stroke protocol,3,4 do not reliably distinguish pediatric stroke from stroke mimics.11,12 As a result, only one quarter to one third of patients with acute neurological deficits managed through a stroke protocol were ultimately diagnosed with stroke.2,3

3

Page 3 of 22

PEDIATRIC STROKE AND MIMICS Acute stroke protocols are resource intensive, and in an era of cost containment with an emphasis on the rational allocation of resources, it is unclear how acute stroke protocols can be modified to improve their diagnostic accuracy as children with stroke mimics were not well characterized in the aforementioned studies.2,3 Furthermore, stroke mimics are not necessarily benign, with up to two thirds of patients in one series being diagnosed with a clinically significant structural abnormality on neuroimaging requiring urgent intervention.7 Thus, the aim of this paper is to describe our institution's experience with stroke mimics identified through our pediatric stroke clinical pathway (PSCP).

Methods This study was conducted at a free-standing children’s hospital in Detroit, Michigan with approximately 90,000 emergency department visits and 10,000 hospital admissions annually. Patients included in this study were those who presented to the pediatric emergency department from June 2014 to December 2016 with a focal neurologic deficit, for whom the PSCP was activated. Patients were identified through retrospective review of the PSCP pager log. The pager log is accessed monthly by the institution’s communications specialist and emailed to the principal investigator. For the purpose of this study, a stroke mimic was defined as a clinical entity characterized by a neurological deficit of sudden onset, reminiscent of a stroke by history and physical examination, but with normal findings or findings suggestive of a neurological disorder other than stroke by magnetic resonance imaging (MRI).

As previously reported and briefly summarized here, the PSCP resulted from a multidisciplinary collaboration between hospital administration and the departments of anesthesia, critical care,

4

Page 4 of 22

PEDIATRIC STROKE AND MIMICS emergency medicine, hematology, neurology, pharmacy, and radiology.3 Implemented in June 2014, the PSCP can be activated for patients aged 1 month to 18 years presenting to the pediatric emergency department with a positive “stroke screen” within 72 hours of symptom onset. A positive stroke screen for the protocol was defined as one or more of the following: (1) sudden onset of a focal neurological deficit, which may be motor (gait disturbance, hemiparesis, facial weakness) or non-motor (visual field impairment, sensory symptoms) in nature; (2) new onset seizure followed or preceded by a persistent focal neurological deficit; or (3) unexplained alteration of mental status in the presence of a risk factor for stroke. Risk factors were itemized on the PSCP for clinician reference.3 Patients with a positive stroke screen were coded as an emergency severity index 1 or 2 for immediate evaluation by the physician. The emergency severity index is a five level tool used to triage emergency department patients from highest acuity (level 1) through to lowest acuity (level 5).

Clinicians activate the PSCP using a single fan out page to alert attending physicians from the departments of anesthesia, critical care, neurology, and radiology in order to expedite imaging and hospital admission. With activation of the PSCP, a standardized protocol is initiated with an emphasis on urgent neuroimaging. The protocol also delineates requisite laboratory studies, interventions to be performed in the pediatric emergency department, guidelines for neuroprotective strategies and nursing care. Patients with confirmed stroke are admitted to the intensive care unit.

Demographic and clinical data, including presenting signs and symptoms, physical examination findings, imaging results, and hospital length of stay were abstracted using a standardized form.

5

Page 5 of 22

PEDIATRIC STROKE AND MIMICS This study was approved by the Institutional Review Board of Wayne State University School of Medicine.

Statistical Analyses Categorical variables are reported by numbers and percentages. Normality of the continuous variables is tested by Kolmogorov-Smirnov test. Due to the non-normality of the data in continuous variables, we reported median and inter-quartile range (IQR) as a descriptive statistic. Pearson’s Chi-squared test is used to analyze the distribution of categorical variable by groups, provided no expected frequency less than 1, and no more than 20% of the cell should have an expected frequency less than 5, otherwise Fisher-Exact test is used for the analysis. We determine odds ratio with a 95% confidence interval to compare the difference in groups. Two group comparisons for non-normally distributed continuous variables were compared using the Wilcoxon rank sum test. Data are analyzed by statistical package SAS (version 9.4, SAS Institute Inc. Cary, North Carolina). Significance level was set at 0.05.

Results Demographics The PSCP was activated for 59 patients during the study period and 14 (23.7%) patients were diagnosed with stroke. The most common diagnoses among the 45 children with stroke mimics were functional neurological disorders (FND) (9/45, 20.0%), transient neurological deficits (8/45, 17.8%), migraine (7/45, 15.6%), and seizure (5/45, 11.1%) (Table 1). The demographics of children with stroke and those with a stroke mimic were not significantly different. A higher proportion of patients in the stroke group were admitted to the pediatric intensive care unit

6

Page 6 of 22

PEDIATRIC STROKE AND MIMICS compared to those with a stroke mimic, although this difference was not statistically significant. Therapy was initiated in a significantly higher proportion of stroke patients compared to stroke mimics (Table 2).

Timing to Activation of the PSCP, Neuroimaging and Emergency Department Disposition Patients in both groups were evaluated within a median time of 20 minutes after arrival to the emergency department. Although not statistically significant, the PSCP was activated later for patients with a stroke mimic compared to those with a stroke. There was no significant difference in the proportion of patients who underwent computed tomography (CT) brain [stroke 9/14 (64.3%) versus non-stroke 23/45 (51.1%); p = 0.59] and MRI brain [stroke 14/14 (100%) versus non-stroke 39/45 (86.7%); p = 0.32]. The timing of these studies did not differ between the two groups and the emergency department length of stay was also similar in both groups. Patients with a stroke mimic had a shorter hospital length of stay compared to patients with a stroke, but this difference did not achieve statistical significance (Table 3).

Clinical presentation of Stroke and Stroke Mimics Vomiting was more commonly reported by patients with stroke (OR: 4.00, 95%CI: 1.12 to 14.35) than those with a stroke mimic (Figure 1). Weakness was less commonly reported by patients with stroke (OR: 0.7, 95%CI: 0.07 to 0.90) compared to those with a stroke mimic (Figure 1). These differences persisted when patients with stroke were compared to those with a final diagnosis of a FND (Table 4). Despite differences in patient self reported symptoms, the physical examination did not differ between the groups (Figure 2, Table 4).

7

Page 7 of 22

PEDIATRIC STROKE AND MIMICS Discussion In our study, only one quarter of children for whom the PSCP was activated were eventually diagnosed with a stroke. At first glance, the PSCP in its current form appears to have a low yield; however, it is in keeping with prior series examining pediatric stroke presentations to the emergency department.8,10 This low yield is in marked contrast, however, to adult series where approximately three-quarters of individuals with a preliminary diagnosis of stroke have a final discharge diagnosis of stroke.13,14 Stroke continues to be under diagnosed in children, with marked delays in the time to imaging and initiation of treatment due to poor recognition by pediatric providers.6 Paradoxically, in the presence of a stroke protocol, most activations turn out to have a final diagnosis that is not cerebrovascular disease.

The most common conditions that mimicked childhood stroke in our study were functional neurological disorders (FND), transient neurologic deficits, migraine, and seizures. In our study, the category of FND includes patients where the final discharge diagnosis was a somatic or conversion disorder. The category of transient neurologic deficits includes those patients with neurological symptoms and/or signs that resolved or improved either in the emergency department or during their admission coupled with a normal MRI.

FND are exceedingly more common than pediatric stroke, accounting for up to 50% of primary care visits.15 In the emergency department setting, FND are an important differential diagnosis for pediatric stroke as previous authors have reported that stroke like attacks can be a manifestation of FND.16 Distinguishing the two, however, can be problematic. In contrast to the series by Mackay et al. where focal limb weakness and an inability to walk predicted the

8

Page 8 of 22

PEDIATRIC STROKE AND MIMICS presence of a true stroke,9 we noted that half of our cohort with FND had motor weakness of an extremity, with an abnormal gait found in few patients with or without stroke. However, facial weakness, ophthalmoplegia and depressed level of consciousness were notably absent in all children with FND.

Recurrent emergency department visits coupled with certain key historical factors (“mixed” symptomatology) and positive physical signs on examination (Hoover’s sign, midline splitting, astasia-abasia, and “collapsing gait”), can help an astute clinician distinguish organic disease from functional disease processes.15,17 It is important to note, however, that the validity of “la belle indifference” – thought to be the sine qua non of FND – has been called into question.18 Moreover, in contrast to popular belief, individuals who present with a FND may not have a profound stressor, such as physical or sexual abuse, immediately prior to the onset of symptoms; instead, family conflict is a relatively common precipitant.19 Similarly, children with FND need not have a major psychiatric co-morbidity that accounts for their presentation, as quantitative personality scales and categorical psychiatric diagnoses were not noted to be significantly different between healthy controls and adults with certain forms of FND.20 However, an important confounding factor is that up to half of children with a FND may have a co-morbid organic neurological disease,21 resulting in a complex clinical picture. Unfortunately, the need for rapid triage and assessment in the emergency department, and a lack of access to a pediatric neurologist, may limit the time and clinical expertise available for obtaining the nuanced history and detailed neurological examination needed to attempt to differentiate FND from stroke on clinical grounds.

9

Page 9 of 22

PEDIATRIC STROKE AND MIMICS In our study, transient neurological deficits accounted for a rather surprising number of children noted to have a stroke mimic. As mentioned, this category included children who had a transient neurological deficit with a normal MRI. True transient ischemic attacks, caused by small vessel pathology, account for less than 5% of all “brain attacks” in children, but the exact prevalence in the pediatric age group is unclear.2,8 Distinguishing transient neurological deficits due to a migraine variant, Todd’s paralysis or a transient vasculopathy from a retrospective chart review is difficult; thus, this category of children likely includes patients with a variety of disease states causing symptoms to wax and wane. However, children with transient ischemic attacks of a vascular etiology are at risk for an arterial ischemic stroke and one third of children with an ischemic stroke have a recent diagnosis of a transient ischemic attack.22 Therefore, it appears prudent that children presenting with transient neurological deficits be admitted to the hospital and evaluated for stroke risk factors.

Headache is a common presenting feature of several forms of childhood stroke; 23,24 thus, migraine is an important differential diagnosis for pediatric patients with stroke like symptoms. In fact, migraine was the most common final diagnosis in one series, found in 28% of pediatric patients presenting to the emergency department with a “brain attack”.8 In contrast, in our study, migraine was noted in 16% of patients, being less common than both FND and transient neurological deficits. In particular, migraine with brain stem aura can present with symptoms suggestive of a stroke, such as ataxia, altered level of consciousness, vertigo, and slurred speech. Vertigo and dizziness are frequently encountered in other forms of childhood migraine as well, making the distinction between stroke and migraine challenging for the emergency department physician.25 Furthermore, migraine can be a predisposing factor for childhood stroke, further

10

Page 10 of 22

PEDIATRIC STROKE AND MIMICS complicating the clinical picture for a child with a history of migraine presenting with a neurological deficit.26 However, in most instances, migraine can be distinguished from stroke by the relatively gradual onset of symptoms and absence of focal motor weakness on physical examination.

Finally, seizures are a presenting feature of arterial ischemic stroke in 11% to 52% of children.27,28 Seizure is particularly common with stroke in younger children. Importantly, seizures in the pediatric population are the presenting symptom for a wide range of disorders of the nervous system, including acute demyelinating syndromes, vascular brain malformations and neurocutaneous syndromes, as well as systemic disease states such as sepsis and drug toxicity. In other words, seizures are a non-specific feature of many conditions in children and often warrant urgent evaluation. In the case of seizure followed by unresponsiveness or focal weakness, a clinical pathway can serve to expedite care for these patients, as our results demonstrate that even those with a mimic received care within the same timeframe as those patients with stroke.

Overall, our results demonstrate that there is a wide spectrum of disease states that can present with stroke like features in children. Although children with stroke like symptoms are more often diagnosed with a benign entity, up to 41% of patients have a clinically significant diagnosis.8 In our series, none of the patients with a stroke mimic required neurosurgical intervention, but 20% were started on urgent medical therapy, such as steroids. Thus, stroke protocols do play a role in not only reducing the time to a diagnosis of stroke,2-4 but also serve to quickly identify clinically important etiologies of “brain attacks” in children, enabling prompt institution of therapy.8 The ongoing challenge, however, it to further refine these protocols so as to improve their

11

Page 11 of 22

PEDIATRIC STROKE AND MIMICS identification of children with organic disease, who require urgent imaging as outlined by International Pediatric Stroke Study Neuroimaging Subgroup29 while limiting the overuse of resources for patients with FND.15

Limitations This study has several limitations. The sample size is small and the retrospective nature of the study could have limited the clinical information that was available for analysis. In addition, we included only those patients for whom the PSCP was activated. Thus, it is possible that the study missed patients presenting with acute neurological deficits whose ultimate diagnosis was stroke or a stroke mimic. The study was also performed at a single institution with 24/7 pediatric neurology and radiology coverage, which limits generalizability to other practice settings where these services may not be available around the clock. Finally, the PSCP could be activated for patients with symptom onset within 72 hours of presentation to the emergency department. This extended timeframe with inconsistencies in documentation of the time of symptom onset may account for the large number of patients with both FND and transient neurological deficits. As previous studies demonstrate that when the exact time of symptom onset can be pin-pointed, a diagnosis of stroke is more likely, a narrow window between symptom onset and protocol activation may improve the predictive value of the protocol.13

Conclusions A standardized pediatric stroke clinical pathway ensures timely evaluation of patients presenting with neurological deficits, but fails to reliably differentiate between patients with stroke and those with stroke mimics. Since acute neurological deficits in children can signify both benign

12

Page 12 of 22

PEDIATRIC STROKE AND MIMICS and non-benign entities, further multi-centered studies are needed to develop a “stroke screen” that reliably distinguishes pediatric stroke from its mimics, with the aim to provide timely care to all children with clinically significant etiologies for their acute neurological symptoms. At the present time, algorithms that focus upon “brain attacks” as opposed to stroke alone may be best suited for pediatric emergency departments.

13

Page 13 of 22

PEDIATRIC STROKE AND MIMICS References 1. Jordan LC, Hillis AE. Challenges in the diagnosis and management of pediatric stroke. Nat Rev Neurol. 2011;7:199-208. 2. Ladner TR, Mahdi J, Gindville MC, Gordon A, Harris ZL, Crossman K, Pruthi S, Abramo TJ, Jordan LC. Pediatric acute stroke protocol activation in a children’s hospital emergency department. Stroke. 2015;46:2328-2331. 3. DeLaroche AM, Sivaswamy L, Farooqi A, et al. Pediatric stroke clinical pathway improves the time to diagnosis in an emergency department. Pediatr Neurol. 2016;65:3944. 4. Shack M, Andrade A, Shah-Basak PP, et al. A pediatric institutional acute stroke protocol improves timely access to stroke treatment. Dev Med Child Neurol. 2017;59:31-37. 5. Rafay MF, Pontigon AM, Chiang J, et al. Delay to diagnosis in acute pediatric arterial ischemic stroke. Stoke. 2009;40:58-64. 6. Srinivasan J, Miller SP, Phan TG, et al. Delayed recognition of initial stroke in children: need for increased awareness. Pediatrics. 2009;124(2):e227. 7. Shellhaas RA, Smith SE, O’Tool E, et al. Mimics of childhood stroke: characteristics of a prospective cohort. Pediatrics. 2006;118(2):704-709. 8. Mackay MT, Chua AK, Lee M, et al. Stroke and nonstroke brain attacks in children. Neurology. 2014;82:1434-1440. 9. Mackay MT, Yock-Corrales A, Churilov L, et al. Differentiating childhood stroke from mimics in the emergency department. Stroke. 2016;47:2476-2481. 10. Mackay MT, Yock-Corrales A, Churilov L, et al. Accuracy and reliability of stroke diagnosis in the pediatric emergency department. Stroke. 2017;48:1198-1202.

14

Page 14 of 22

PEDIATRIC STROKE AND MIMICS 11. Neville K, Lo W. Sensitivity and specificity of an adult stroke screening tool in childhood ischemic stroke. Pediatr Neurol. 2016;58:53-56. 12. Mackay MT, Churilov L, Donnan GA, et al. Performance of bedside stroke recognition tools in discriminating childhood stroke from mimics. Neurology. 2016;86:2154-2161. 13. Hand PJ, Kwan J, Lindley RI, et al. Distinguishing between stroke and mimic at the bedside: the brain attack study. Stroke. 2006;37(3):769-75. 14. Libman RB, Wirkowski E, Alvir J, et al. Conditions that mimic stroke in the emergency department. Implications for acute stroke trials. Arch Neurol. 1995;52:1119–1122. 15. de Gusmão CM, Guerriero RM, Bernson-Leung ME, et al. Functional neurological symptom disorders in a pediatric emergency room: diagnostic accuracy, features, and outcome. Pediatr Neurol. 2014;51(2):233-238. 16. Vilela P. Acute stroke differential diagnosis: Stroke mimics. Eur J Radiol. Published online: May 5, 2017. DOI: 10.1016/j.ejrad.2017.05.008. [Epub ahead of print]. 17. Tremolizzo L, Susani E, Riva MA, et al. Positive signs of functional weakness. J Neurol Sci. 2014;340(1-2):13-18. 18. Stone J, Warlow C, Sharpe M. The symptom of functional weakness: a controlled study of 107 patients. Brain. 2010;133(Pt 5):1537-1551. 19. Kozlowska K, Nunn KP, Rose D, et al. Conversion disorder in Australian pediatric practice. J Am Acad Child Adolesc Psychiatry. 2007;46(1):68-75. 20. Kranick S, Ekanayake V, Martinez V, et al. Psychopathology and psychogenic movement disorders. Mov Disord. 2011;26(10):1844-1850. 21. Baird G, Santosh PJ. Interface between neurology and psychiatry in childhood. J Neurol Neurosurg Psychiatry. 2003;74(Suppl 1):i17-i22.

15

Page 15 of 22

PEDIATRIC STROKE AND MIMICS 22. Adil MM, Qureshi AI, Beslow LA, et al. Transient ischemic attack requiring hospitalization of children in the United States: kids’ inpatient database 2003 to 2009. Stroke. 2014;45(3):887-888. 23. Schwartz NE, Vertinsky AT, Hirsch KG, et al. Clinical and radiographic natural history of cervical artery dissections. J Stroke Cerebrovasc Dis. 2009;18(6):416-423. 24. Guo R, Ma L, Shrestha BK, et al. A retrospective clinical study of 98 adult idiopathic primary intraventricular hemorrhage cases. Medicine. 2016;95(42):e5089. 25. Jahn K, Langhagen T, Heinen F. Vertigo and dizziness in children. Curr Opin Neurol. 2015;28(1):78-82. 26. Eikermann-Haerter K, Lee JH, Yuzawa I, et al. Migraine mutations increase stroke vulnerability by facilitating ischemic depolarizations. Circulation. 2012;125(2):225-245. 27. Gabis LV, Yangala R, Lenn NJ. Time lag to diagnosis of stroke in children. Pediatrics. 2002;110:924–928. 28. Zimmer JA, Garg BP, Williams LS, et al. Age-related variation in presenting signs of childhood arterial ischemic stroke. Pediatr Neurol. 2007;37:171–175. 29. Mirsky DM, Beslow LA, Amlie-Lefond C, et al. Pathways for neuroimaging of childhood stroke. Pediatr Neurol. 2017;69:11-23.

16

Page 16 of 22

PEDIATRIC STROKE AND MIMICS Figure 1. Symptoms in patients with and without stroke after implementation of the PSCP

17

Page 17 of 22

PEDIATRIC STROKE AND MIMICS Figure 2. Signs in patients with and without stroke after implementation of the PSCP

18

Page 18 of 22

PEDIATRIC STROKE AND MIMICS Table 1. Final diagnoses for all patients (n = 59) for whom the PSCP was activated

Diagnosis Stroke Functional neurological disorder Transient neurological deficit Migraine Seizure Central nervous system metastases Demyelinating disease

Mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes Drug toxicity Moya Moya without stroke Sepsis / meningitis Atypical thrombotic thrombocytopenic purpura and hemolytic-uremic syndrome Bell’s palsy

N, (%) 14, (23.7%) 9, (15.3%) 8, (13.6%) 7, (11.9%) 5, (8.5%) 5, (8.5%) 4 (6.8%) 2, (3.4%)

1, (1.7%) 1, (1.7%) 1, (1.7%) 1, (1.7%)

1, (1.7%)

19

Page 19 of 22

PEDIATRIC STROKE AND MIMICS Table 3. Median timing of care for patients with and without stroke after implementation of the PSCP

Time

Arrival to physician time (min) Arrival to PSCP time (min) Physician to PSCP time (min) Arrival to CT time (min) Arrival to MRI time (hours) Emergency department length of stay (hours) Hospital length of stay (hours)

Stroke (n = 14) (Median ± IQR) 17.50 ± 16.00

Stroke Mimic (n = 45) (Median ± IQR) 16.00 ± 14.00

P value

29.50 ± 42.00

52.00 ± 54.00

0.08

23.00 ± 30.00

34.00 ± 51.00

0.12

61.00 ± 64.00

56.00 ± 55.00

0.60

4.00 ± 10.00

4.00 ± 7.00

0.79

4.39 ± 2.39

4.46 ± 2.36

0.39

101. 00 ± 107.00

44.00 ± 109.00

0.09

0.71

Table 2. Demographic characteristics of all patients for whom the PSCP was activated

Demographic Characteristic Age (years) Median ± IQR Male Race African American White Asian Afternoon / overnight presentation Emergency severity index code 1

Stroke (n = 14) 11.5 ± 7.0

Stroke Mimic (n = 45) 13.0 ± 6.0

p value

8 (57%)

19 (42%)

0.37

7 (50%) 7 (50%) 0 (0%) 8 (57%)

28 (62%) 16 (36%) 1 (2%) 31 (69%)

0.64

8 (57%)

20 (44%)

0.26

20

0.11

0.48

Page 20 of 22

PEDIATRIC STROKE AND MIMICS Intensive care admission (n, %) Therapy (n, %)

8 (57%)

13 (29%)

0.23

6 (43%)

3 (7%)

0.003

Table 4. Clinical characteristics of all patients with stroke, a stroke mimic or a FND after implementation of the PSCP

Clinical Feature & Symptoms Risk factor Sudden symptom onset Baseline state of health Headache Vomiting Weakness Numbness / tingling Blurry vision Seizure Abnormal mental status Dizziness Slurred speech Abnormal gait Pain** Signs Facial weakness Extremity weakness Dysarthria Gait ataxia Abnormal extraocular movements

Stroke (n = 14)

FND ( n = 9)

P value

6/14 (43%) 9/14 (64%)

Stroke Mimics* (n = 36) 18/45 (40%) 29/45 (64%)

2/9 (22%) 6/9 (67%)

0.52 1.00

11/14 (79%)

36/45 (80%)

8/9 (89%)

0.90

5/14 (36%) 7/14 (50%) 4/14 (29%)

19/45 (42%) 9/45 (20%) 28/45 (62%)

3/9 (33%) 0/9 (0%) 7/9 (78%)

0.75

6/14 (43%) 5/14 (36%) 2/14 (14%) 2/14 (14%)

19/45 (42%) 9/45 (20%) 6/45 (13%) 8/45 (18%)

3/9 (33%) 2/9 (22%) 0/9 (0%) 1/9 (11%)

0.83 0.46 0.49 1.00

0/14 (0%) 2/14 (14%) 3/14 (21%)

3/45 (7%) 9/45 (20%) 6/45 (13%) 16/45 (36%)

2/9 (22%)

0.09

2/9 (22%) 2/9 (22%)

1.00 0.51

3/9 (33%)

0.67

8/45 (18%)

0/9 (0%)

0.22

19/45 (42%) 6/45 (13%) 6/45 (13%) 5/45 (11%)

5/9 (56%) 2/9 (22%) 2/9 (22%)

0.58 0.26

0/9 (0%)

0.72

3/14 (21%) 1/14 (7%) 7/14 (50%) 0/14 (0%) 4/14 (29%) 2/14 (14%)

21

0.03 0.05

0.25

Page 21 of 22

PEDIATRIC STROKE AND MIMICS Visual defects Facial sensory changes Extremity sensory changes Altered responsiveness Pupillary abnormalities Other neurological signs

0/14 (0%) 1/14 (7%) 3/14 (21%) 5/14 (36%) 3/14 (21%) 4/14 (29%)

3/45 (7%) 3/45 (7%) 9/45 (20%) 10/45 (22%) 3/45 (7%) 9/45 (20%)

1/9 (11%)

0.50

0/9 (0%)

1.00

3/9 (33%)

0.47

0/9 (0%) 0/9 (0%) 2/9 (22%)

0.15 0.31 0.75

*Excludes patients with a final diagnosis of a FND. **Pain sites include: chest, abdomen, back, extremities.

22

Page 22 of 22