Abstract W P324: Acute NIHSS Profiles Predict Differential Outcomes After Mild Ischemic Stroke

Abstract

Listen

Translate article

Background: A NIHSS total score ≤ 5 is a common criteria for mild stroke. Persons with scores in this range are often considered too mild for tPa treatment and not in need of rehabilitation. There may be distinct patterns of stroke symptoms, possibly not all deficits or combination of deficits produce similar levels of disability. Aim: To investigate if patient profiles created by analysis of initial NIHSS item scores predict 90 day and 1 year outcome on the Modified Rankin Scale (mRS) and Stroke Impact Scale (SIS). Methods: Patients with ischemic stroke and initial NIHSS score ≤ 5 enrolled in the PROTECT DC trial were included. Latent class analysis of NIHSS items was used to create discrete symptom profiles: patients with similar item responses were grouped within the same profile group. Logistic regression using profile groups as independent variables predicted 90 day and 1 year mRS scores. Kruskal-Wallace tests examined differences on Stroke Impact Scale (SIS) Participation scores across the groups. Results: 153 participants were included in the latent class analysis, yielding a 3-profile model. Profile A (82 patients, 54%) represented mild strokes with low probability of abnormality on all NIHSS items. Profile B (16 patients, 11%) patients had right-side motor impairment and sensory deficits. Profile C (55 patients, 36%) patients were more likely to have facial palsy and dysarthria. The median NIHSS score decreased from the most severe Profile B (mdn=3.00) to the most mild Profile A (mdn=1.00), with Profile C (mdn=2.00) between the two extremes. Profile B had significantly higher rates (p<.05) of unfavorable outcomes on mRS (mRS ≥ 2) at 90 days (64%) and 1-year (62%). After adjusting for age, sex, premorbid mRS, and comorbidities, Profile B had 6.56 (CI: 1.45-29.72) times and Profile C had 1.77(CI: 0.73-4.27) times higher odds of being disabled then Profile A at 90 days. One year mRS findings were similar. Profile B also presented with lower levels of participation than Profiles A and C for 1-year SIS participation scores. Conclusion: Patients with initial NIHSS scores ≤ 5 with right-side motor weakness and sensory deficits had more unfavorable outcomes. We suggest that mild strokes are not a homogenous group and should not be treated equally in clinical settings.