Establishment of Precise Criteria for Evaluating Gait Independence to Prevent Falls in Patients with Stroke

This study aimed to establish precise criteria for evaluating gait independence to prevent falls in patients with stroke. The study verified the discriminative accuracy of a combination of three evaluation items validated in our previous research: OLST with a cane for the paralyzed side, OLST with a cane for the non-paralyzed side, and cognition-related behavioral assessment (CBA), which evaluates higher brain dysfunction.

Overall, 155 patients with stroke who were discharged from a rehabilitation hospital in Japan between March 2018 and August 2023 were enrolled in this study. Data collection was performed on the 14th day after the patients began gait in the ward, which was used as the reference date. To assess walking independence, the following evaluations were performed: OLST with a cane for the paralyzed side, OLST with a cane for the non-paralyzed side, and CBA. The discriminative accuracy of the evaluation criteria for gait independence was analyzed by calculating sensitivity, specificity, area under the curve (AUC), and generalized R-squared. A neural network model was used with 80% of the data for training and 20% for testing. The output layer consisted of two categories of gait ability in the ward at discharge, supervised and independent.

The gait independence evaluation criteria demonstrated high discriminative accuracy, with a sensitivity of 100%, specificity of 85.7%, and AUC of 0.982. This indicated a robust model with a generalized R-squared value of 0.845.

The results of this study revealed that three evaluation items—OLST with a cane for the paralyzed side, OLST with a cane for the non-paralyzed side, and CBA—are highly accurate for assessing gait independence. This method for determining gait independence is relatively simple and quantitative and has significant value for clinical applications.

 The use of these criteria by physical therapists on the 14th day after patients with stroke begin gait in the ward will lead to a more accurate determination of whether patients require supervision or can walk independently in the ward. Additionally, it will help prevent falls in individuals deemed to have achieved gait independence.