[Validity of the Wisconsin Gait Scale to predict the risk of falls in stroke patients].

Introduction: Approximately, half of the stroke patients suffered a fall during the first year after the injury, affecting its functionality. Gait evaluation scales could evaluate to predict the risk of falls.
Objective: To study the ability of the Wisconsin Gait Scale (WGS) to predict the risk of falls in stroke subjects.
Patients and Methods: Retrospective observational study that included stroke patients between May 2010 and October 2016. Data from the WGS and the Berg Balance Scale (BBS) were collected after hospitalization (acute phase), at 3 months (subacute phase) and at 6 and 12 months (chronic phases). BBS was used as the outcome variable to determine the ability of the WGS to predict the risk of falls. The calculation of the predictive validity was carried out using the ROC (Receiver Operating Characteristics) curves and the analysis of the area under the curve (AUC, area under the curve).
Results: Data were collected from 61 patients who met the inclusion criteria (after admission: 61; 3 months: 61; 6 months: 58; 12 months: 42). In all measurements, the AUC was greater than 0.9, so the WGS distinguishes stroke patients with moderate risk of falling from those with low risk.
Conclusions: The WGS scale is a valid instrument to predict the risk of falls in patients with a criterion in relation to the cut-off points established in the BBS.