Iza Faria-Fortini1, Janaíne C Polese2, Christina D C M Faria3, Aline Alvim Scianni3, Lucas R Nascimento4, Luci Fuscaldi Teixeira-Salmela3. 1. Department of Occupational Therapy, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil. Electronic address: ifaria@ufmg.br. 2. Department of Physical Therapy, Faculdade Ciências Médicas de Minas Gerais, Belo Horizonte, Brazil. 3. Department of Physical Therapy, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil. 4. Center of Health Sciences, Discipline of Physiotherapy, Universidade Federal do Espírito Santo, Vitória, Brazil.
Abstract
BACKGROUND: Falls, which are common events after stroke, may lead to activity limitations and increased dependence. It is important to identify which commonly employed clinical measures could differentiate individuals, who are fallers from the non-fallers. AIM: To investigate specific cut-off values of clinical measures that could discriminate fallers and non-fallers individuals with chronic stroke. METHOD: This cross-sectional study involved 105 community-dwelling individuals with stroke. The primary outcome was report of falls over the last six months. The clinical predictors included measures of mobility (walking speed, stair ascent/descent cadences, time to perform the Timed Up and Go test, and ABILOCO) and the Fall Efficacy Scale - International (FES-I) scores. To identify which measures were able to detect between-group differences, independent Student's t-tests were employed. For measures which were able to discriminate fallers from the non-fallers, the Receiver Operating Characteristics (ROC) and the Area Under the ROC Curve (AUC) were calculated. RESULTS: Out of the 105 participants (61 men), 41% reported falls over the previous 6 months. Stair ascent cadence, ABILOCO, and FES-I scores significantly differentiated the groups, but only the FES-I demonstrated acceptable discriminatory ability (AUC = 0.71). The optimal FES-I cut-off score was 28 points (sensitivity = 0.71; specificity = 0.57; positive predictive value = 51%; and negative predictive value = 74%). CONCLUSIONS: The FES-I demonstrated good discriminatory ability to classify individuals with chronic stroke, who were fallers from the non-fallers. The use of the established cut-off value of 28 points is recommended and may help clinical reasoning and decision-making in stroke rehabilitation.
BACKGROUND: Falls, which are common events after stroke, may lead to activity limitations and increased dependence. It is important to identify which commonly employed clinical measures could differentiate individuals, who are fallers from the non-fallers. AIM: To investigate specific cut-off values of clinical measures that could discriminate fallers and non-fallers individuals with chronic stroke. METHOD: This cross-sectional study involved 105 community-dwelling individuals with stroke. The primary outcome was report of falls over the last six months. The clinical predictors included measures of mobility (walking speed, stair ascent/descent cadences, time to perform the Timed Up and Go test, and ABILOCO) and the Fall Efficacy Scale - International (FES-I) scores. To identify which measures were able to detect between-group differences, independent Student's t-tests were employed. For measures which were able to discriminate fallers from the non-fallers, the Receiver Operating Characteristics (ROC) and the Area Under the ROC Curve (AUC) were calculated. RESULTS: Out of the 105 participants (61 men), 41% reported falls over the previous 6 months. Stair ascent cadence, ABILOCO, and FES-I scores significantly differentiated the groups, but only the FES-I demonstrated acceptable discriminatory ability (AUC = 0.71). The optimal FES-I cut-off score was 28 points (sensitivity = 0.71; specificity = 0.57; positive predictive value = 51%; and negative predictive value = 74%). CONCLUSIONS: The FES-I demonstrated good discriminatory ability to classify individuals with chronic stroke, who were fallers from the non-fallers. The use of the established cut-off value of 28 points is recommended and may help clinical reasoning and decision-making in stroke rehabilitation.
Authors: Barbara Spanò; Maria G Lombardi; Massimo De Tollis; Maria A Szczepanska; Claudia Ricci; Alice Manzo; Simone Giuli; Lorenzo Polidori; Ivo A Griffini; Fulvia Adriano; Carlo Caltagirone; Roberta Annicchiarico Journal: Brain Sci Date: 2022-01-27