Stroke mimic prediction scales for initial diagnosis in the emergency department

Summary

The article presents a current and comprehensive review of stroke mimics (SM), which represent a challenge for differential diagnosis due to their wide range of similar symptoms with strokes. It delves into the brief epidemiology, clinical features, and four predictive scales for SM diagnosis, which were identified on the basis of a literature search: TeleStroke Mimic Score (TSM), FABS, simplified FABS (sFABS), and Khan score. These validated tools might support rapid and efficient decision-making regarding treatment in an emergency department setting with the goal of minimizing delays in providing adequate care to patients with stroke. We would like to highlight the importance of correct identification of SM given the time sensitivity of revascularization treatment, with a focus on optimizing treatment and management of patients with acute onset of neurological symptoms.

Keywords:

intravenous thrombolysis – stroke mimics – ischemic strokes

This is an unauthorised machine translation into English made using the DeepL Translate Pro translator. The editors do not guarantee that the content of the article corresponds fully to the original language version.

 

 

Introduction

Stroke mimics (SM) include a wide range of disorders that mimic CMP in their symptoms. The time-sensitivity of recanalization treatment for acute CMP emphasizes early and sufficiently accurate symptom recognition, making the diagnosis of SM, both in prehospital and early hospital care, particularly difficult.

 

Epidemiology and risk factors

The percentage of patients with SM treated in the emergency department according to the current findings is 25-30% [1,2]. In studies that included only patients treated with intravenous thrombolysis (IVT), the proportion of patients with SM was 4.1-6.6% [3,4]. The results of a recently published systematic review suggest that patients with a final diagnosis of SM are more likely to be younger, with a higher proportion of women. Risk factors for stroke (arterial hypertension, smoking, dyslipidemia, diabetes mellitus, atrial fibrillation, ischemic heart disease or peripheral vascular disease) are less frequent, but in contrast, migraine or cognitive impairment are more common in the personal history of patients with SM [2]. Surprisingly, the incidence of psychiatric illness and epilepsy does not differ significantly between patients with SM and CMP. However, there are other studies indicating a higher incidence of CMP in patients with a history of psychiatric disorder, which may be explained by the increased risk of CMP in patients taking antipsychotic or antidepressant medications [5,6]. On the objective findings and vital signs side, patients with SM have been shown to have significantly lower National Institutes of Health Stroke Scale (NIHSS) scores as well as lower blood pressure (measured in the emergency department or pre-hospital phase) compared to patients with documented stroke.

 

Classification of stroke mimics

Stroke mimics can be divided into two main categories -⁠ medical and functional (Table 1).

The representation of individual SMs varies across studies. The most recent pooled review shows the following frequencies of each type: peripheral vestibular syndrome 23%, intoxication and metabolic disorders 13%, epileptic seizures 13%, migraine 7%, collapse/presyncope 6%, cranial (mono) neuropathy 5%, brain tumours 4%, amygdala 2%, dementia 1.2% and spinal cord lesions 0.7%.

 

Clinical picture

The onset of SM symptoms is not always sudden. Fluctuations in the severity of the condition and the occurrence of systemic symptoms such as fluctuations in state of consciousness, confusion, agitation or fever are also common [8]. Among the symptoms mimicking CMP, dizziness, vertigo, quantitative and qualitative disturbances of consciousness, paresthesia of the limbs or face, numbness of the limbs, monoplegia, ataxia of the limbs, headache or visual disturbances are the most common [7]. In contrast, a strong association has been shown between the absence of a drooping nook and SM, suggesting that SM patients often do not have a drooping nook [9].

 

Diagnosis of stroke mimics

For most patients with suspected stroke, ambulance crews are the first point of contact. Diagnosis of CMP or SM in the prehospital setting is often difficult because of the overlapping clinical picture and time constraints. Thus, correct recognition of symptoms, transfer of structured information to the ictal center physician during the teleconsultation and subsequent decision on referral play a significant role in reducing delays in the field of prehospital emergency care [10].

Some predictive scales validated for pre-hospital diagnosis of stroke (Los Angeles Prehospital Stroke Screen [LAPSS], Melbourne Ambulance Stroke Screen [MASS], Medic Prehospital Assessment for Code Stroke [MedPACS]) contain items (history of epilepsy or glucose values) whose presence increases suspicion of a possible diagnosis of SM [11-13]. A separate SM screening containing four items -⁠ comatose/near comatose state, history of epileptic seizure immediately preceding symptom development, glycaemia < 2.8 mmol/l, diagnosed (and active) malignant brain tumour -⁠ is part of the Ambulance Clinical Triage For Acute Stroke Treatment (ACT-FAST) pre-hospital large artery occlusion prediction scale [14].

Predictive scales for the diagnosis of SM are most commonly used in emergency departments. Based on a literature search conducted in Medline and Ovid using the Boolean operator ["stroke" AND "mimic*" AND ("predict*" OR "diagnos*") AND ("tool" OR "scale" OR "score") ], we identified five scales, with only those that were externally validated included in the summary table. The first of these is the TeleStroke Mimic Score (TSM) scale, which was validated back in 2014 for use through a telemedicine network in a hub-and-spoke model. The TSM includes six variables (age, atrial fibrillation, arterial hypertension, epileptic seizure, facial asymmetry and NIHSS > 14) [9]. The Recognition of Stroke in the Emergency Room (ROSIER) includes both symptoms of stroke (upper or lower limb weakness, speech impairment, corner drop or visual field loss) and parameters whose presence raises suspicion for an alternative diagnosis (epileptic seizure, initial impairment of consciousness or syncope) [15]. Another scoring system is the FABS. This consists of six items (absence of a dip in the carotid artery, negative history of atrial fibrillation, age < 50 years, initial systolic blood pressure < 150 mmHg, history of epilepsy or isolated sensory impairment). The best cutoff value for SM prediction was FABS ≥ 3 with 90% sensitivity and 91% specificity [16]. The FABS was subsequently simplified to sFABS (simplified FABS), which includes only four items (absence of angle drop, negative history of atrial fibrillation, age < 50 years, systolic blood pressure < 150 mmHg at baseline field measurement) [17]. The last of the scales mentioned is the Khan score. The Khan score was derived from SM risk factors including age, absence of risk factors for stroke (arterial hypertension, hyperlipidemia, diabetes mellitus, atrial fibrillation), history of migraine, epilepsy or psychiatric disease) [18]. In 2020, external validation of the above four SM prediction tools (TMS, FABS, sFABS, Khan score) was performed. In identifying possible SM, all four prediction tools had similar yields (Table 2) [4].

 

Conclusion

Stroke mimics include a wide range of disorders that mimic a stroke in their symptoms, although they are not cerebrovascular events. In prehospital and early hospital stratification, early and accurate symptom recognition is crucial with respect to the time sensitivity of recanalization treatment for acute ischemic CMP.

The organisation of ictal care in the Czech Republic demonstrates a high level of professionalism and specialisation, with patient triage playing a significant role in the rapid and effective provision of treatment. A patient with clinical signs of rapidly progressive brain lesions is a potential candidate for recanalization treatment and should therefore be evaluated as triage positive and transported to a highly specialized ict care centre or a highly specialized cerebrovascular care centre as soon as possible. The most common SMs include conditions following a previous epileptic seizure, such as Todd's hemiparesis or postictal phakic disorder. Another nosologic entity mimicking CMP is migrainous aura -⁠ especially in primomanifestation. The situation in which a person with migraine headache may have an ongoing cerebrovascular accident cannot be overlooked, as this group of patients is at higher risk of developing an ictus [2]. We speak of a migrainous infarction when the symptoms meet the characteristics of migraine with aura, last longer than 60 min and neuroimaging methods demonstrate ischemia that corresponds in localization to a focal deficit.

Given the above information, it is necessary to cite a large multicentre study for the overall context, which pointed out that IVT in patients with SM, especially those with functional neurological symptoms, is relatively safe even with respect to possible symptomatic intracranial haemorrhage [3]. Although efforts should be made to identify SM before IVT, the American Heart Association/American Stroke Association (AHA/ASA) recommendations suggest that IVT should not be abandoned because of concerns about SM treatment.

In order to have an effective conversation with the ambulance crew, it is advisable to ask about the following data, among others: history of epilepsy, migraine, psychiatric illness and presence or absence of atrial fibrillation.

Physicians (neurologists) or emergency physicians can use the scales presented by us for suspicion of SM in the initial differential diagnosis (before neuroimaging), but multimodal CT (CT, CTA, CT perfusion), possibly MRI incl. MRA.

 

Grant support

Supported by the grant of the Ministry of Health -⁠ AZV ČR reg. no. NU23-0400336; Ministry of Health -⁠ RVO (08/RVO-FNOs/2022), SGS19/LF/2024 and the National Research Ict Network STROCZECH within the research infrastructure CZECRIN (project no. LM2023049) financed by the state budget of the Czech Republic.

 

Conflict of interest

The authors declare that they have no conflict of interest in relation to the subject of the study.

 

Table 1. Classification of stroke mimics.

Medical SM		Functional SM
Organic brain damage	Migraine focal cranial neuropathy PRES (posterior reversible encephalopathy syndrome) Epilepsy bacterial endocarditis epidural/subdural haematomas hypertensive crisis brain neoplasm encephalitis posthypoxic/postanoxic states	depression stress anxiety disorder chronic pain/fatigue somatization and psychiatric complications of neurological diseases functional impairment of mobility, speech or perception
Systemic diseases	ionic imbalance metabolic disorders Intoxication

SM -⁠ stroke mimics

 

Table 2. Components of predictive scales for diagnosis of stroke mimics, translated from [4].

Scale	Clinical items	Score	Indicator	Sensitivity (%) 95% CI
				Specificity (%) 95% CI
TSM 0,75 (95% CI 0.63-0.87)	Age	+ 0.2 per year	a higher score means a lower probability of SM	91,3 (86,9-94,5)
	presence of atrial fibrillation	+ 6		58,8 (32,9-81,6)
	arterial hypertension	+ 3
	history of epilepsy	-⁠ 6
	decrease in the corner of the mouth	+ 9
	NIHSS > 14	+ 5
		Minimum: -⁠ 6 maximum: no upper limit
FABS 0,61 (95% CI 0.49-0.74)	absence of corner drop age < 50 years absence of atrial fibrillation systolic blood pressure < 150 mmHg the presence of an isolated sensory deficit history of epilepsy	1 point for each variable minimum: 0 maximum: 6	a higher score means a higher probability that it is SM	24,6 (19,3-30,5)
				52,9 (27,8-77,0)
Simplified FABS (sFABS) 0,61 (95% CI = 0.48-0.73)	absence of corner drop age <50 years absence of atrial fibrillation systolic blood pressure < 150 mmHg	1 point for each variable. minimum: 0 maximum: 4	a higher score means a higher probability that it is SM	25,4 (20-31,4)
				52,9 (27,8-77,0)
Khan score 0,60 (95% CI 0.52-0.69)	Age		a higher score means a higher probability that it may be SM	32,1 (26,2-38,4)
	50	2		88,2 (63,6-98,5)
	50-70	1
	> 70	0
	presence of hypertension/hyperlipidaemia/diabetes mellitus/FiSi
	no disease	3
	1 without FiSi	2
	2 or 3 without FiSi	1
	FiSi	0
	history of migraine	2
	history of epilepsy	1
	history of psychiatric illness	1
		minimum: 0 maximum: 9

 

CI -⁠ confidence interval; FiSi -⁠ atrial fibrillation; NIHSS -⁠ National Institutes of Health Stroke Scale; SM -⁠ stroke mimics; TSM -⁠ TeleStroke Mimic Score