Neglect scoring modifications in the National Institutes of Health Stroke Scale improve right hemisphere stroke lesion volume prediction

Background: The National Institutes of Health Stroke Scale (NIHSS) does not equitably assess stroke severity in the two cerebral hemispheres. By attributing a maximum of two points for neglect and seven for language, it undervalues right hemisphere deficits. We aimed to investigate if NIHSS equally predicts right hemisphere lesion volumes in patients with and without neglect, and if a modification of the neglect scoring rules could increase its predictive capacity.

Methods: We analyzed a prospective cohort of acute right middle cerebral artery ischemic stroke patients. First, we calculated the correlation between NIHSS scores and lesion volume and analyzed the partial correlation of neglect. Then, we applied different modifications in the neglect scoring rules and investigated how they interfered with lesion volume predictive capacity.

Results: A total of 162 ischemic stroke patients were included, 108 with neglect and 54 without. The correlation between lesion volume and NIHSS was lower in patients with neglect (r = 0.540 vs. r = 0.219, p = 0.004) and neglect was a statistically significant covariate in the partial correlation analysis between NIHSS and lesion volume (p = 0.017). With the neglect score tripled and with the duplication or triplication of all neglect modalities, the correlation was significantly higher than with the standard NIHSS (p = 0.043, p = 0.005, p = 0.001, respectively). With these modifications, neglect was no longer a significant covariable in the partial correlation between lesion volume and NIHSS.

Conclusion: A modification of NIHSS neglect scoring might improve the scale's capacity to predict lesion volume.