Assessing the risk for falls among Portuguese community-dwelling stroke survivors. Are we using the better tools? Observational study.

This study assesses the estimation of the risk for falls among community-dwelling stroke survivors referred for ambulatory physiotherapy and explores factors that affect the risk. Methods: Observational, cross-sectional with nested case-control study, of individuals, referred to physiotherapy less than 12 months after stroke and able to walk independently. Berg Balance Scale, Timed Up and Go Test, and the Motor Assessment Scale were applied. Berg Balance Scale ≤45 or Timed Up and Go Test > 14 were used to estimate the risk for falls. The discrimination ability of the estimation was assessed. Alternative models were explored by logistic regression analysis. Results: One hundred sixty-seven patients fulfilled the inclusion criteria. Patients were 21 to 87years old (median 66), 98 men (58.7%), and in 133 (79.6%) the stroke occurred in the last 6 months. Falls were reported by 78 (46.7%) of the patients but 139 (83.2% [95%CI 76.84-88.14]) were estimated as having risk for falls. The discrimination ability of the estimation of the actual occurrence of falls by Berg Balance Scale ≤45 or Timed Up and Go Test >14 was 55% (95%CI 47.5-62.4). The actual occurrence of falls was associated only with Motor Assessment Scale, as a protective factor. The discrimination ability of the estimation of the actual occurrence of falls by Motor Assessment Scale alone was area under the curve 0.69 (95%CI 0.60-0.77). Conclusions: Different tools with better performance are needed to identify the risk for falls after stroke.