ARTICLE IN PRESS
The TriAGe+ Score for Vertigo or Dizziness: A Diagnostic Model for Stroke in the Emergency Department R. Kuroda, MD,* T. Nakada, MD, PhD,† T. Ojima, MD, PhD,‡ M. Serizawa, MD, PhD,§ N. Imai, MD, PhD,§ N. Yagi, MD,§ A. Tasaki, MD,‖ M. Aoki, MD,† T. Oiwa, MD,† T. Ogane, MD,† K. Mochizuki, MD,† M. Kobari, MD, PhD,§ and H. Miyajima, MD, PhD*
Background: Vertigo or dizziness is a common occurrence, but it remains a challenging symptom when encountered in the emergency department (ED). A diagnostic score for stroke with high accuracy is therefore required. Methods: A singlecenter observational study (498 patients) was conducted. The predictor variables were derived from a multivariate logistic regression analysis with Akaike information criterion. The outcome was the occurrence of stroke. We evaluated the utility of a new diagnostic score (TriAGe+) and compared it with the ABCD2 score. Results: The cohorts included 498 patients (147 with stroke [29.4%]). Eight variables were included: triggers, atrial fibrillation, male gender, blood pressure ≥140/ 90 mm Hg, brainstem or cerebellar dysfunction, focal weakness or speech impairment, dizziness, and no history of vertigo or dizziness or labyrinth or vestibular disease. We derived the TriAGe+ score from these variables. In the cohort, the prevalence of stroke increased significantly using the diagnostic score: 5.9% for a score of 0-4; 9.1% for 5-7; 24.7% for 8-9; and 57.3% for 10-17. At a cutoff value of 10 points, the sensitivity of the score was 77.5%, the specificity was 72.1%, and the positive likelihood ratio was 3.2. When the cutoff was defined as 5 points, the score obtained a high sensitivity (96.6%) with a good negative likelihood ratio (.15). The new score outperformed the ABCD2 score for the occurrence of stroke (C statistic, .818 versus .726; P < .001). Conclusions: The TriAGe+ score can identify the occurrence of stroke in patients with vertigo or dizziness presenting to the ED. Key Words: Vertigo—dizziness—stroke—diagnosis—emergency departments— ABCD2. © 2017 National Stroke Association. Published by Elsevier Inc. All rights reserved.
From the *Department of Neurology, Hamamatsu University School of Medicine, Shizuoka, Japan; †Critical Care Medical Center, Japanese Red Cross Shizuoka Hospital, Shizuoka, Japan; ‡Department of Community Health and Preventive Medicine, Hamamatsu University School of Medicine, Shizuoka, Japan; §Department of Neurology, Japanese Red Cross Shizuoka Hospital, Shizuoka, Japan; and ‖Departments of Neurology, National Center of Neurology and Psychiatry, Tokyo, Japan. Received September 13, 2016; revision received December 13, 2016; accepted January 13, 2017. The authors declare no financial or other conflicts of interest in association with this study. Address correspondence to Ryo Kuroda, MD, Department of Neurology, Hamamatsu University School of Medicine, 1-20-1 Handayama Higashi-ku, Hamamatsu, 431-3192, Japan. E-mail: [email protected]. 1052-3057/$ - see front matter © 2017 National Stroke Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2017.01.009
Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2017: pp ■■–■■
1
ARTICLE IN PRESS R. KURODA ET AL.
2
Introduction Vertigo and dizziness are common symptoms requiring timely assessment and treatment in the emergency department (ED). A study for the epidemiology of vertigo and dizziness in the ED indicated that these symptoms were shown in approximately 4% of ED patients.1,2 According to several studies conducted in patients with dizziness or vertigo in the ED, the vertigo or dizziness in 3%-20% of these patients is due to cerebrovascular causes.1,3-8 Vertigo and dizziness are also challenging symptoms to classify, owing to the difficulty in the diagnosis and the wide range of diseases featuring these symptoms, such as disorders of the inner ears and stroke, especially in the ED.5,6 According to 1 international survey examining emergency physicians, an identification clinical tool for central vertigo or dizziness is in the second greatest need among various clinical priorities.9 However, the ability to accurately diagnose patients with vertigo or dizziness is limited.10 During the past decades, several studies have focused on identifying diagnostic factors for the prevalence of stroke.7,8,11-13 Navi et al demonstrated that the ABCD2 score, commonly used as a risk score of transient ischemic attack, was useful for identifying cases of cerebrovascular disease (CVD) among patients with dizziness.12 Newman-Toker et al also presented the accuracy of Head-Impulse—Nystagmus—Test-of-Skew (HINTS) and its variants in the diagnosis of acute vestibular syndrome (AVS).14-16 However, care should be taken before applying the eye movement test in routine ED practice, as the interpretations differ between experts and novices.17 Whether a diagnostic score specifically designed for vertigo and dizziness improves the diagnostic accuracy concerning stroke compared with the ABCD2 score is unknown. The prime objective in our study was to establish a more efficient diagnostic score to help identify the occurrence of stroke in patients with vertigo or dizziness presenting to the ED, and to compare its performance with that of the ABCD2 score.
Materials and Methods Study Setting We conducted a single-center observational study of patients with vertigo or dizziness visiting the ED of Japanese Red Cross Shizuoka Hospital, located in Shizuoka city, which has 700,000 residents. The hospital is a 500-bed community-based hospital that provides acute medical care. The annual numbers of patients visiting the ED and those with stroke on admission are approximately 17,000 and 400, respectively. The emergency medicine system in the city is covered by 5 hospitals, including the Japanese Red Cross Shizuoka Hospital, by rotation. The patients whose conditions deteriorate
after the index ED visit or after being discharged subsequently revisit the same hospital as an ED rule in Shizuoka city. In the ED setting, formatted neurological examinations were supposed to be recorded by physicians and double-checked by senior emergency physicians, neurologists, or both. Between August 1, 2008, and September 31, 2013, we retrospectively collected the clinical information of patients with vertigo or dizziness in the ED using a hospital-based database and electronic medical records. To identify patients with vertigo, dizziness, or both, we extracted patients with the hospital-based original ED diagnosis codes of “vertigo or dizziness,” “CVD,” “acute coronary syndromes,” “arrhythmia,” “central nervous system disorders (except CVD),” “hypoglycemia,” and “metabolic disease (except hypoglycemia)”, including a disease list by Newman-Toker et al. as diseases that have risk of stroke with vertigo or dizziness from all patients visiting the ED (Table 1).5 We then extracted eligible patients using a text search of ED visit records of these patients for the following terms: “vertigo,” “dizziness,” or “dizzy.” We excluded patients who had faintness or presyncope without any other symptoms, who had intestinal bleeding, who were aged
The TriAGe+ Score for Vertigo or Dizziness: A Diagnostic Model for Stroke in the Emergency Department R. Kuroda, MD,* T. Nakada, MD, PhD,† T. Ojima, MD, PhD,‡ M. Serizawa, MD, PhD,§ N. Imai, MD, PhD,§ N. Yagi, MD,§ A. Tasaki, MD,‖ M. Aoki, MD,† T. Oiwa, MD,† T. Ogane, MD,† K. Mochizuki, MD,† M. Kobari, MD, PhD,§ and H. Miyajima, MD, PhD*
Background: Vertigo or dizziness is a common occurrence, but it remains a challenging symptom when encountered in the emergency department (ED). A diagnostic score for stroke with high accuracy is therefore required. Methods: A singlecenter observational study (498 patients) was conducted. The predictor variables were derived from a multivariate logistic regression analysis with Akaike information criterion. The outcome was the occurrence of stroke. We evaluated the utility of a new diagnostic score (TriAGe+) and compared it with the ABCD2 score. Results: The cohorts included 498 patients (147 with stroke [29.4%]). Eight variables were included: triggers, atrial fibrillation, male gender, blood pressure ≥140/ 90 mm Hg, brainstem or cerebellar dysfunction, focal weakness or speech impairment, dizziness, and no history of vertigo or dizziness or labyrinth or vestibular disease. We derived the TriAGe+ score from these variables. In the cohort, the prevalence of stroke increased significantly using the diagnostic score: 5.9% for a score of 0-4; 9.1% for 5-7; 24.7% for 8-9; and 57.3% for 10-17. At a cutoff value of 10 points, the sensitivity of the score was 77.5%, the specificity was 72.1%, and the positive likelihood ratio was 3.2. When the cutoff was defined as 5 points, the score obtained a high sensitivity (96.6%) with a good negative likelihood ratio (.15). The new score outperformed the ABCD2 score for the occurrence of stroke (C statistic, .818 versus .726; P < .001). Conclusions: The TriAGe+ score can identify the occurrence of stroke in patients with vertigo or dizziness presenting to the ED. Key Words: Vertigo—dizziness—stroke—diagnosis—emergency departments— ABCD2. © 2017 National Stroke Association. Published by Elsevier Inc. All rights reserved.
From the *Department of Neurology, Hamamatsu University School of Medicine, Shizuoka, Japan; †Critical Care Medical Center, Japanese Red Cross Shizuoka Hospital, Shizuoka, Japan; ‡Department of Community Health and Preventive Medicine, Hamamatsu University School of Medicine, Shizuoka, Japan; §Department of Neurology, Japanese Red Cross Shizuoka Hospital, Shizuoka, Japan; and ‖Departments of Neurology, National Center of Neurology and Psychiatry, Tokyo, Japan. Received September 13, 2016; revision received December 13, 2016; accepted January 13, 2017. The authors declare no financial or other conflicts of interest in association with this study. Address correspondence to Ryo Kuroda, MD, Department of Neurology, Hamamatsu University School of Medicine, 1-20-1 Handayama Higashi-ku, Hamamatsu, 431-3192, Japan. E-mail: [email protected]. 1052-3057/$ - see front matter © 2017 National Stroke Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2017.01.009
Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2017: pp ■■–■■
1
ARTICLE IN PRESS R. KURODA ET AL.
2
Introduction Vertigo and dizziness are common symptoms requiring timely assessment and treatment in the emergency department (ED). A study for the epidemiology of vertigo and dizziness in the ED indicated that these symptoms were shown in approximately 4% of ED patients.1,2 According to several studies conducted in patients with dizziness or vertigo in the ED, the vertigo or dizziness in 3%-20% of these patients is due to cerebrovascular causes.1,3-8 Vertigo and dizziness are also challenging symptoms to classify, owing to the difficulty in the diagnosis and the wide range of diseases featuring these symptoms, such as disorders of the inner ears and stroke, especially in the ED.5,6 According to 1 international survey examining emergency physicians, an identification clinical tool for central vertigo or dizziness is in the second greatest need among various clinical priorities.9 However, the ability to accurately diagnose patients with vertigo or dizziness is limited.10 During the past decades, several studies have focused on identifying diagnostic factors for the prevalence of stroke.7,8,11-13 Navi et al demonstrated that the ABCD2 score, commonly used as a risk score of transient ischemic attack, was useful for identifying cases of cerebrovascular disease (CVD) among patients with dizziness.12 Newman-Toker et al also presented the accuracy of Head-Impulse—Nystagmus—Test-of-Skew (HINTS) and its variants in the diagnosis of acute vestibular syndrome (AVS).14-16 However, care should be taken before applying the eye movement test in routine ED practice, as the interpretations differ between experts and novices.17 Whether a diagnostic score specifically designed for vertigo and dizziness improves the diagnostic accuracy concerning stroke compared with the ABCD2 score is unknown. The prime objective in our study was to establish a more efficient diagnostic score to help identify the occurrence of stroke in patients with vertigo or dizziness presenting to the ED, and to compare its performance with that of the ABCD2 score.
Materials and Methods Study Setting We conducted a single-center observational study of patients with vertigo or dizziness visiting the ED of Japanese Red Cross Shizuoka Hospital, located in Shizuoka city, which has 700,000 residents. The hospital is a 500-bed community-based hospital that provides acute medical care. The annual numbers of patients visiting the ED and those with stroke on admission are approximately 17,000 and 400, respectively. The emergency medicine system in the city is covered by 5 hospitals, including the Japanese Red Cross Shizuoka Hospital, by rotation. The patients whose conditions deteriorate
after the index ED visit or after being discharged subsequently revisit the same hospital as an ED rule in Shizuoka city. In the ED setting, formatted neurological examinations were supposed to be recorded by physicians and double-checked by senior emergency physicians, neurologists, or both. Between August 1, 2008, and September 31, 2013, we retrospectively collected the clinical information of patients with vertigo or dizziness in the ED using a hospital-based database and electronic medical records. To identify patients with vertigo, dizziness, or both, we extracted patients with the hospital-based original ED diagnosis codes of “vertigo or dizziness,” “CVD,” “acute coronary syndromes,” “arrhythmia,” “central nervous system disorders (except CVD),” “hypoglycemia,” and “metabolic disease (except hypoglycemia)”, including a disease list by Newman-Toker et al. as diseases that have risk of stroke with vertigo or dizziness from all patients visiting the ED (Table 1).5 We then extracted eligible patients using a text search of ED visit records of these patients for the following terms: “vertigo,” “dizziness,” or “dizzy.” We excluded patients who had faintness or presyncope without any other symptoms, who had intestinal bleeding, who were aged
