Predicting 3-month independency using postural assessment scale for stroke in acute stroke unit: total score and item-level analysis

Purpose: This study aimed to establish the optimal cutoff scores for the Postural Assessment Scale for Stroke (PASS) and to identify the specific PASS items at admission and discharge from an acute stroke care unit to predict independent functional status in activity of daily living (ADL) three months later. The identified cutoff scores can assist in prognosis estimation and rehabilitation goal setting, while the specific PASS items can guide task-oriented interventions in acute stroke rehabilitation. Method: Participants were 109 individuals with acute ischemic stroke admitted to the acute care unit, all with dependence in ADL upon admission (Barthel index, BI, ˂ 60). Data on PASS at admission and discharge were collected, with a total PASS score range of 0 to 36. Participants were categorized into independent (BI ≥ 60) or dependent (BI ˂ 60) groups at three months post-stroke. Multivariate logistic regression analysis, with age, National Institutes of Health Stroke Scale score (NIHSS), length of acute care unit stay (LOS), admission BI as the covariates, and total PASS score as the independent variable, was performed to identify predictors of 3-month independence. In item-level analysis, each item underwent univariate logistic regression analyses, and those with significant associations were included in multivariate logistic regression analysis to determine items significantly predicting the ADL status at 3-month while controlling for confounders. Results: On average, participants were discharged from the acute ward on day 11 after stroke onset. PASS scores at admission (OR = 1.09, 95% CI = 1.01-1.19) and discharge (OR = 1.17, 95% CI = 1.07-1.27) individually predicted ADL status at three months after stroke, with the same covariates controlled. The cutoff scores at admission and discharge were 14 (area under curve = 0.97, 95% confidence interval = 0.94 - 0.99) and 21 (area under curve = 0.99, 95% confidence interval = 0.96 - 1.00), respectively. Among all items, item 1 at admission (OR = 2.77, 95% CI = 1.42-5.39) and item 10 at discharge (OR = 4.00, 95% CI = 1.90-8.42) were significant predictors, with similar percentages correctly classifying ADL status three months post-stroke as using PASS total scores. Conclusions: The identified total PASS scores and PASS items can assist in setting rehabilitation goals and plans, estimating prognosis, and facilitating discharge planning.