NIHSS is not enough for cognitive screening in acute stroke: A cross-sectional, retrospective study.

The aim of this study was to investigate whether the cognitive subscale of the National Institute of Health Stroke Scale (NIHSS), the Cog-4, can detect cognitive deficits in acute stroke. This was a cross-sectional, retrospective study. The study sample consisted of people with stroke enrolled in an acute stroke unit. The index test Cog-4 was calculated based on admission NIHSS score. The reference standard instrument, the Montreal Cognitive Assessment (MoCA), was performed within 36-48 h of admission. Non-parametric statistics were used for data analyses. The study included 531 participants with a mean age of 69 years. The Cog-4 failed to identify cognitive deficits in 65%, 58%, and 53% of patients when the MoCA thresholds for impaired cognition were set at ≤25 p, ≤23 p, and ≤19 p, respectively, were chosen for impaired cognition. The agreement between the Cog-4 and the MoCA was poor; Cohen's kappa was from -0.210 to -0.109, depending on the MoCA cut-offs. The sensitivity of the Cog-4 was 35%, 42% and 48% for the MoCA thresholds for impaired cognition ≤25, ≤23 and ≤19 points, respectively. The Cog-4 has a limited ability to identify cognitive deficits in acute stroke. More structured and comprehensive tests should be employed as diagnostic tools.