ARTICLE IN PRESS

e-NIHSS: an Expanded National Institutes of Health Stroke

Scale Weighted for Anterior and Posterior Circulation Strokes

Silvia Olivato, MD, Silvia Nizzoli, MD, Milena Cavazzuti, MD,

Federica Casoni, MD, Paolo Frigio Nichelli, MD, and Andrea Zini, MD

Background: The National Institutes of Health Stroke Scale (NIHSS) is the most
widespread clinical scale used in patients presenting with acute stroke. The merits
of the NIHSS include simplicity, quickness, and agreement between clinicians. The
clinical evaluation on posterior circulation stroke remains still a limit of NIHSS.
Methods: We assessed the application of a new version of NIHSS, the e-NIHSS
(expanded NIHSS), adding specific elements in existing items to explore signs/
symptoms of a posterior circulation stroke. A total of 22 consecutive patients with
suspected vertebrobasilar stroke were compared with 25 patients with anterior
circulation stroke using NIHSS and e-NIHSS. Results: We compared the NIHSS
and e-NIHSS scores obtained by the 2 examiners, in patients with posterior cir-
culation infarct (POCI), using the Wilcoxon test. Patients with POCI evaluated
with e-NIHSS had an average of 2 points higher than patients evaluated with
classical NIHSS. The difference was statistically significant (P < .05), weighted by
the new expanded items. Conclusions: The NIHSS is a practical scale model, with
high reproducibility between trained, different examiners, focused on posterior
circulation strokes, with the same total score and number of items of the existing
NIHSS. The e-NHISS could improve the sensitivity of NIHSS in posterior circu-
lation stroke and could have an impact on clinical trials, as well as on outcomes.
Further studies are needed to investigate a larger number of patients and the cor-
relation between the e-NIHSS score and neuroimaging findings. Key Words: Stroke
management—stroke scale—stroke care—posterior circulation infarct (POCI).
© 2016 National Stroke Association. Published by Elsevier Inc. All rights reserved.

Introduction mography and clinical outcome at 3 months. 2 It is

composed of 15 items: level of consciousness, horizon-
The National Institutes of Health Stroke Scale (NIHSS)1
tal eye movements, visual field test, facial palsy, motor
is a valid instrument that provides a standardized eval-
arm, motor leg, limb ataxia, sensory, language, speech,
uation of symptoms and signs in stroke. The NIHSS has
extinction, and inattention. The score extends from 0
acquired good predictive value with 7 days’ cranial to-
(normal neurological examination) to 42 (unresponsive-
ness coma); a score of 10 or higher is more probably related
From the Nuovo Ospedale Civile S. Agostino-Estense, AUSL Modena, to a large-artery occlusion.3
Department of Neuroscience, Neurology Clinic, Modena, Italy.
The score has a good correlation with the middle ce-
Received October 23, 2015; accepted August 8, 2016.
All authors contributed equally to the manuscript.
rebral artery territory size infarct but underestimates clinical
Address correspondence to Andrea Zini, MD, Nuovo Ospedale severity in posterior circulation stroke.4,5 Although some
Civile S. Agostino-Estense, AUSL Modena, Department of items related to the vertebrobasilar system can be scored,
Neuroscience, Neurology Clinic, via Giardini 1355, Modena 41100, other elements receive no score (e.g., diplopia, dyspha-
Italy. E-mail: a.zini@ausl.mo.it.
gia, gait instability, hearing, and nystagmus). A specific
1052-3057/$ - see front matter
© 2016 National Stroke Association. Published by Elsevier Inc. All
scale for posterior circulation (Israeli Vertebrobasilar Stroke
rights reserved. Scale6), a scale for the emergency triage (Emergency Triage
http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2016.08.011 Stroke Scale7), or the modified NIHSS, derived from the

Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2016: pp ■■–■■ 1
 ARTICLE IN PRESS
2 S. OLIVATO ET AL.
8
NIHSS by deleting redundant items, has been proposed. In item 6 of the e-NIHSS, the examiner performs the
These simplifications did not solve the lack of sensitiv- test for facial palsy. The patient is asked to open the mouth
ity for posterior circulation stroke. and move the tongue from side to side. After the facial
test, the patient is asked to say “ahh,” and it should be
Materials and Methods observed if the palate rises symmetrically at the back of
the oral cavity. Lingual palsy or asymmetry in velum el-
e-NIHSS evation is scored 3.
The e-NIHSS (where “e” stands for expanded) is derived NIHSS item 11 tests limb ataxia, whereas e-NIHSS also
from NIHSS and focuses on anterior and posterior cir- tests trunk ataxia. The patient is asked to sit up in the
culation strokes, adding specific elements in existing items bed. If latero- or retropulsion is present in the sitting po-
to explore signs or symptoms of posterior circulation stroke. sition, the score is 2. If the sitting position is maintained,
The e-NIHSS consists of 15 items with scores from 0 to the patient is asked to stand up in the Romberg posi-
42, with 5 categories added on the existing items 4, 6, tion. Instability in the Romberg position is scored 1.
and 11 (Table 1). The scale was firstly approved by 2
experts in cerebrovascular disease who did not partici-
Patients
pate in the study.
NIHSS item 4 tests only horizontal eye movements. In A total of 22 consecutive patients with suspected
the e-NIHSS, horizontal and vertical eye movements are vertebrobasilar stroke hospitalized in the Stroke Unit of
tested, and patients are asked to track a pen or a finger Nuovo Ospedale Civile S. Agostino-Estense in Modena
from side to side and then up and down. Multidirec- were recruited and compared with 25 patients with an-
tional, vertical, horizontal, lateral, and dissociated terior circulation stroke.
nystagmus is scored 1. The presence of Horner syn- The e-NIHSS was used within 24 hours of symptoms
drome is scored 1. onset, 30 minutes after the NIHSS evaluation, by 2 in-

Table 1. NIHSS items compared with expanded e-NIHSS items

Item 4, NIHSS Item 4, e-NIHSS

Horizontal eye 0 = normal Horizontal and 0 = normal

movements 1 = partial gaze palsy (gaze is abnormal vertical eye 1 = partial gaze palsy (gaze is abnormal in 1
in 1 or both eyes, but forced deviation movements or both eyes, but forced deviation or total
or total gaze paresis is not present) gaze paresis is not present)
2 = forced deviation or total gaze 1 = nystagmus and/or Horner’s syndrome
(paresis not overcome by the 2 = forced deviation or total gaze (paresis not
oculocephalic maneuver) overcome by the oculocephalic maneuver)

Item 6, NIHSS Item 6, e-NIHSS

Facial palsy 0 = normal Facial, 0 = normal

1 = minor paralysis (flattened nasolabial hypoglossal and 1 = minor paralysis (flattened nasolabial fold,
fold, asymmetry on smiling) glossopharyngeal asymmetry on smiling)
2 = partial paralysis (total or near-total palsy 2 = partial paralysis (total or near-total
paralysis of the lower face) paralysis of the lower face)
3 = complete paralysis of 1 or both sides 3 = complete paralysis of 1 or both sides
(absence of facial movement in the (absence of facial movement in the upper
upper and lower parts of the face). and lower parts of the face).
3 = deficit of IX nerve (soft palate paralysis)
3 = deficit of XII nerve

Item 11, NIHSS Item 11, e-NIHSS

Limb ataxia 0 = absent or untestable Limb and trunk 0 = absent or untestable

1 = present in 1 limb ataxia 1 = present in 1 limb
2 = present in 2 limbs 1 = imbalance in Romberg position
2 = present in 2 limbs
2 = trunk ataxia or retro- or lateropulsion

Abbreviations: e-NIHSS, expanded NIHSS; NIHSS, National Institutes of Health Stroke Scale.
 ARTICLE IN PRESS
e-NIHSS: AN EXPANDED NIHSS 3
dependent examiners with NIHSS certification. We classified Table 2. Total patients with relative scores evaluated with
stroke subtypes using the Bamford classification con- NIHSS and e-NIHSS
firmed by brain tomography and/or magnetic resonance
imaging. Patient Sex Age NIHSS e-NIHSS ob 1 e-NIHSS ob 2

1 F 59 4 4 4
Statistical Analysis 2 M 78 3 3 3
3 F 81 22 24 23
We used the specific validation criteria for scales:
4 M 78 35 35 35
- Intrarater reliability (“retest”) is a method used to 5 F 71 20 20 20
compare the reproducibilities of a test, applied to 6 M 75 5 5 6
the same patient, by the same examiner at differ- 7 F 69 21 23 23
ent times. Inter-rater reliability is used to compare 8 F 82 22 25 25
the reproducibilities of a test, applied to the same 9 F 63 20 18 20
patient, by 2 blinded, independent examiners. These 10 F 73 11 11 11
criteria are measured using the Fleiss k coefficient. 11 M 78 7 7 7
12 F 49 2 5 5
Fleiss k coefficient is rated with a range from 0 to
13 M 52 2 4 4
1 (where k < 0: no agreement; 0 < k < .20: poor agree-
14 M 68 4 6 6
ment; .41 < k < .60: moderate agreement; .61 < k < .80: 15 F 83 3 7 7
good agreement; .81 < k < 1: perfect agreement). 16 F 68 8 8 8
- Internal consistency assesses the consistency of results 17 F 76 12 15 15
across items within the scale. 18 M 73 2 2 2
- The validity of each item evaluates symptoms or 19 M 65 5 6 6
disabilities in relation to already validated items in 20 F 70 16 18 18
other neurological scales. We compared any statis- 21 F 76 3 3 3
tical difference between each item and global score 22 M 69 6 8 8
valued with the 2 scales using Spearman correla- 23 M 56 8 9 9
24 M 73 18 19 18
tion. We also estimated the statistical weight on the
25 M 61 3 4 4
final score of new items in the e-NIHSS using the
26 M 74 6 6 6
Wilcoxon test. Results are considered statistically sig- 27 F 52 4 4 3
nificant with a P value less than .05. 28 M 91 4 5 5
29 F 74 3 3 3
30 M 46 7 8 8
Results
31 F 70 4 5 5
We enrolled 47 patients (median age: 69; males 23; 32 M 70 4 4 4
females 24) within 24 hours of stroke onset. A total of 33 F 55 5 7 6
22 patients presented with symptoms suggestive of pos- 34 F 78 0 3 3
terior circulation stroke and 25 presented with symptoms 35 M 48 3 7 7
36 F 79 1 3 3
of anterior circulation stroke. The e-NIHSS and NIHSS
37 M 48 4 5 5
were used. The total scores are reported in Table 2.
38 F 77 21 21 21
The intrareliability of each item of the e-NIHSS for the 39 F 79 1 2 2
single independent observers and the inter-reliability of 40 F 67 17 19 19
the total score of the e-NIHSS between the 2 indepen- 41 F 74 2 2 2
dent observers showed a statistically significant accordance 42 F 49 5 7 7
for both single items and the total score (Table 3). 43 M 85 9 13 13
The correlation between the 2 examiners obtained a 44 M 66 5 5 5
perfect accordance between the 2 scales: e-NIHSS/ 45 M 73 1 3 3
NIHSS, .933*(observer 1); and e-NIHSS/NIHSS, .930* 46 M 63 6 6 6
(observer 2). 47 M 85 9 11 11
We also classified patients according to stroke sub-
Abbreviations: e-NIHSS, expanded NIHSS; F, female; M, male;
types, using Bamford classification: 22 patients (46%) had NIHSS, National Institutes of Health Stroke Scale; ob 1, observer
posterior circulation stroke, confirmed by neuroimaging. 1; ob 2, observer 2.
We compared the NIHSS and e-NIHSS scores obtained
by the 2 examiners, in patients with posterior circulation NIHSS score of 5 and average e-NIHSS score of 7), in-
infarct (POCI), using the Wilcoxon test. Patients with POCI dependently by 2 examiners (Table 4). The difference was
evaluated with e-NIHSS had an average score of 2 points statistically significant (P < .05), weighted by the new ex-
higher than patients evaluated with classical NIHSS (average panded items. Nystagmus, Horner’s syndrome and deficit
 ARTICLE IN PRESS
4 S. OLIVATO ET AL.

Figure 1. e-NIHSS items 4 and 6 compared with item 11 in patients with POCI. Abbreviation: POCI, posterior circulation infarct.

of IX and XII cranial nerves resulted more specific items The NIHSS does not consider specific signs of poste-
for posterior circulation stroke symptoms than trunk ataxia rior circulation stroke, such as dysphagia, cranial nerve
and Romberg imbalance (Fig 1). deficit, and gait imbalance, and underestimates stroke se-
verity in vertebrobasilar strokes. The presence of nystagmus
Discussion and deficit of IX and XII cranial nerves resulted more
specific for posterior circulation stroke than trunk ataxia
We elaborated a standardized stroke scale that in- and Romberg imbalance, which are present both in an-
cludes anterior and posterior circulation strokes, called terior and posterior strokes.
e-NIHSS. Posterior circulation strokes are usually not in-
cluded in clinical trials for many reasons, in particular,
worst outcome and therapeutic benefit. Some items in Table 4. Patients with posterior circulation infarct evaluated
the NIHSS are useful for evaluating vertebrobasilar stroke with e-NIHSS and NIHSS by two observers
patients (e.g., visual field abnormalities and limb ataxia).
Patient NIHSS e-NIHSS ob 1 e-NIHSS ob 2

Table 3. Intrareliability and inter-reliability of each item of 1 11 11 11

e-NIHSS for the single independent observers 2 7 7 7
3 2 5 5
e-NIHSS Fleiss coefficient k 4 2 4 4
5 4 6 6
1. Level of consciousness 1* 6 3 7 7
2. Questions 1* 7 3 4 4
3. Commands 1* 8 12 15 15
4. Best gaze .967* 9 2 2 2
5. Visual 1* 10 5 6 6
6. Facial palsy .949* 11 7 8 8
7. Motor arm (right side) .982* 12 4 5 5
8. Motor arm (left side) .966* 13 6 8 8
9. Motor leg (right side) 1* 14 5 7 7
10. Motor leg (left side) .973* 15 1 3 3
11. Limb ataxia 1* 16 4 5 5
12. Sensory 1* 17 1 3 3
13. Best language .953* 18 4 5 5
14. Dysarthria .946* 19 1 2 2
15. Extinction and inattention 1* 20 17 19 19
e-NIHSS ob 1/e-NIHSS ob 2 .968* 21 9 13 13
22 5 7 7
Abbreviations: e-NIHSS, expanded NIHSS; NIHSS, National In-
stitutes of Health Stroke Scale; ob 1, observer 1; ob 2, observer 2. Abbreviations: e-NIHSS, expanded NIHSS; NIHSS, National In-
*k Fleiss results rated from 0 to 1. stitutes of Health Stroke Scale; ob 1, observer 1; ob 2, observer 2.
 ARTICLE IN PRESS
e-NIHSS: AN EXPANDED NIHSS 5
The e-NIHSS is a quick scale, with high reproducibil- References
ity between trained, different examiners. e-NIHSS might
1. Lyden P, Brott T, Tilley B, et al. Improved reliability of
be a new useful version of NIHSS with expanded items,
the NIH Stroke Scale using video training. NINDS TPA
focused and more specific for posterior circle symp- Stroke Study Group. Stroke 1994;25:2220-2226.
toms, with the same total score and number of items of 2. Brott T, Marler JR, Olinger CP, et al. Measurements of
the existing NIHSS. acute cerebral infarction: lesion size by computer
Unlike specific scales created for posterior circulation tomography. Stroke 1989;20:871-875.
3. Sato S, Toyoda K, Uehara T, et al. Baseline NIH Stroke
strokes, the e-NIHSS assesses both the anterior and pos-
Scale Score predicting outcome in anterior and posterior
terior circulation strokes. circulation strokes. Neurology 2008;70:2371-2377.
e-NIHSS score of POCI was higher than classical NIHSS 4. Linfante I, Linas RH, Schlaug G, et al. Diffusion-weighted
score with a statistically significant difference. The scale imaging and National Institutes of Health Stroke Scale
shows a greater sensibility for posterior stroke, com- in the acute phase of posterior-circulation stroke. Arch
Neurol 2001;58:621-628.
pared with NIHSS, particularly for modified items 4 and
5. Inoa V, Aron AW, Staff I, et al. Lower NIH stroke scale
6. scores are required to accurately predict a good prognosis
The e-NHISS could improve the sensibility of NIHSS in posterior circulation stroke. Cerebrovasc Dis 2014;37:251-
on posterior circulation stroke and has an impact on clin- 255. [Epub 2014 Mar 25].
ical and therapeutic trials, as well as on outcomes of 6. Gur A, Lampl Y, Gross B, et al. A new scale for assessing
patients with vertebrobasilar stroke—the Israeli
posterior strokes. This is a pilot study limited by a small
Vertebrobasilar Stroke Scale (IVBSS): inter-rate reliability
number of patients who lack long-term disability follow- and concurrent validity. Clin Neurol Neurosurg
up. Assessing the 3 months’ modified Rankin Scale score 2007;109:317-322.
could be another primary end point for an enlarged, mul- 7. Lyden P, Lu M, Jackson C, et al. Underlying structure of
ticenter study. the National Institutes of Health Stroke Scale: results of
a factor analysis. Stroke 1999;30:2347-2354.
8. Whelley-Wilson CM, Newman GC. A stroke scale for the
Acknowledgment: The authors thank Riccardo Giannandrea, emergency triage. J Stroke Cerebrovasc Dis 2004;13:247-
English native speaker, for linguistic revision. 253.