Elena Zapata-Arriaza, MD
The National Institutes of Health Stroke Scale (NIHSS) is the most widely employed deficit rating scale in modern neurology and has became the gold standard for stroke severity rating in recent published clinical trials. However, it is essential to know that the scale was not designed to serve as a bedside rating tool for widespread use outside of research trials. The goal of NIHSS is to assure reproducibility, not accuracy.
The authors performed a retrospective analysis of all acute ischemic stroke patients admitted to the stroke unit and intensive care unit of the Lausanne University Hospital from 2003 to 2013. The main goal of this paper was to highlight the characteristics of ischemic strokes with a NIHSS = 0, as well as to determine the long-term evolution of these patients. The originality of the paper is the aim itself, which makes us wonder whether the NIHSS is a sufficiently accurate scale to determine the severity of all ischemic strokes. The authors included all patients with a suspected stroke according to the classical World Health Organization definition. The sample was divided between NIHSS=0 and NIHSS > or equal to 1, and brain and vascular state was assessed mostly by computed tomography. During follow-up, mortality, stroke or transient ischemic attack recurrence and functional outcome (3 and 12 months) were collected.
Among 2997 included patients, 108 had a NIHSS=0. After multivariate analysis, NIHSS=0 subgroup patients showed statistically significant differences. The main findings are displayed in the Table below. The most outstanding clinical and radiological findings related to NIHSS=0 strokes were a lower prestroke disability, a delay in hospital arrival, lacunar and infratentorial presentation, higher ASPECTS at admission and less arterial stenosis or occlusion.
Two NIHSS=0 patients died within 3 months from early stroke recurrence, and 5 others after 12 months (2 cancers, 1 pneumonia, 2 undetermined causes). Favorable outcome was clearly more frequent in NIHSS=0 patients. Recurrence rates were similar for strokes (6.6% in NIHSS=0 versus 8.8% in NIHSS ≥1) and for strokes and TIAs combined (11.4% versus 11.0%). About 28.5% NIHSS=0 patients had unfavorable 3 months outcome (mRS score of ≥2), and 46.7% had some remaining symptoms or handicap (mRS score of ≥1), mostly related in the multivariate analysis with infratentorial stroke localization and acute cerebellar symptoms.
Interesting assessment can be obtained from this paper. First, the NIHSS scale may be insufficient to detect certain neurological deficits with prognostic and functional implications. There may be patients with an NIHSS = 0 and disability in the post-stroke follow-up. This powerful message reveals the limitations in the clinical accuracy of a scale created to be reproducible and whose main application field are clinical trials. Likewise, the persistence of disability associated with cerebellar symptoms and stroke in the posterior territory highlights the underestimation of the NIHSS scale in this stroke localization. To determine a stroke patient’s severity, the complete and properly described neurological examination is closer to the real clinical patient state, and therefore will better guide us to the future evolution of the patient.
Secondly, the recurrence of stroke is similar in both subgroups, although patients with NIHSS = 0 have a CT without alterations. This should alert us to the following: Clinical examination is always the starting point that should lead the management and clinical follow-up. Patients with a non-quantifiable deficit are at risk of recurrence and deserve secondary prevention and adequate follow-up.
Finally, it is important to note that in this article, the image analysis is performed mainly by CT. Such assessment could misdiagnose mimics as strokes; MRI could identify situations where symptoms are not due to vascular causes. In any case, it is important to emphasize that the NIHSS is not infallible, and it must be taken into account the aim of its origin, which was reproducibility, not clinical accuracy.