Comparing Emergency Medicine and Neurology Residents in Assessing Stroke Severity Using the NIHSS

The National Institutes of Health Stroke Scale (NIHSS) is used to assess acute stroke severity and plays a critical role in guiding treatment. There is no requirement for emergency medicine (EM) residents to be certified in NIHSS determination to assess acute stroke severity, even though they may be the primary stroke providers in future practice. We implemented NIHSS training and certification into the residency's core content in neurological emergencies.

In April 2022, all EM residents and attending physicians completed a faculty-moderated, interactive NIHSS training module. In the 6-month pilot, we prospectively assessed EM and neurology residents in their NIHSS assignment, indication for thrombolytic therapy, and large vessel occlusion (LVO) diagnosis using a Qualtrics survey completed for each acute stroke activation. Mean overall NIHSS scores from EM and neurology residents were compared using Spearman's correlation. Inter-rater agreement for each clinical category and treatment decision was calculated using Cohen's κ coefficient.

Twenty-nine matched EM and neurology surveys were analyzed. Mean overall NIHSS scores were similar between EM and neurology residents, 6.6 (IQR = 2, 10) and 6.7 (IQR = 1, 10), (p < 0.001), respectively, with substantial agreement between groups (84.4%, κ = 0.63). Individual NIHSS scores showed moderate to substantial agreement, except for horizontal extraocular movement (75.9%, κ = 0.30). There was fair agreement for indication for thrombolytic therapy (75.9%, κ = 0.39) and moderate agreement for LVO diagnosis and indication for embolization (82.8%, κ = 0.51).

Dedicated NIHSS training was effective in teaching EM residents to assess stroke severity, with moderate to substantial agreement in individual and overall NIHSS scores, except for horizontal eye movement assessment. EM residents may benefit from focused NIHSS training to support their rapid assessment of suspected stroke patients.