Reducing left-right bias in the NIH Stroke Scale: Prospective two-centre validation of a modified item 9 for spatial neglect screening during language assessment.

BACKGROUND

The National Institutes of Health Stroke Scale (NIHSS) is central to acute stroke assessment but underrepresents right-hemisphere deficits such as spatial neglect (SN). The standard NIHSS item 11 primarily captures extinction and lacks sensitivity for the broader behavioural manifestations of SN, contributing to underestimation of right-hemisphere stroke severity. We evaluated the diagnostic accuracy and feasibility of a refined NIHSS item 9 (n-NIHSS-9), designed to preserve language and aphasia assessment while incorporating a structured screening for SN using the Cookie Theft picture description task.

PATIENTS AND

METHODS

In this prospective two-centre observational study, consecutive adults with hemispheric stroke were assessed ≤7 days from onset. The n-NIHSS-9 was compared with the Star Cancellation Test (SCT) and the Line Bisection Test (LBT). The simplified Catherine Bergego Scale (sCBS) served as the reference standard. Diagnostic accuracy metrics and non-completion rates were recorded.

RESULTS

Of 632 patients (51.7% male; mean age 70.4 ± 12.9), SN prevalence was 27.5% (left-sided: 19.7%; right-sided: 7.8%). The n-NIHSS-9 showed 84.75% sensitivity, 96.92% specificity and 93.57% accuracy relative to the sCBS. Administration time was <30 seconds. The non-completion rate of the n-NIHSS among patients with SN was 8.7%. The SCT also showed high sensitivity (76.77%) but more instances of non-completion (11.71%), while the LBT showed lower sensitivity (37.50%) and the highest non-completion (14.71%).

CONCLUSION

The n-NIHSS-9 provides a brief, feasible method for incorporating structured SN screening into the standard NIHSS examination. Its diagnostic performance and minimal time burden may support earlier and more balanced detection of hemispheric stroke deficits in routine practice.