Validade concorrente e confiabilidade para contagem de passos em indivíduos pós-acidente vascular cerebral da plataforma de monitoramento de atividades em reabilitação (MARe)

• Ano de defesa: 2019
• Autor(a) principal: Oliveira, Simone Garcia de
• Orientador(a): Russo, Thiago Luiz de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Fisioterapia - PPGFt
• Palavras-chave em Português: Sedentarismo; Reabilitação; Doenças cerebrovasculares; Tecnologias vestíveis; Acelerometria
• Palavras-chave em Inglês: Sedentarism; Rehabilitation; Cerebrovascular diseases; Wearable technologies; Accelerometry
• Área do conhecimento CNPq: CIENCIAS DA SAUDE::FISIOTERAPIA E TERAPIA OCUPACIONAL
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/11573

Introduction: Sedentary behavior after stroke and physical inactivity increases the risk of recurrent stroke, with a major impact on the world economy, especially in low- and middle-income countries. Advancing in low-cost technologies to monitor physical activity can contribute to preventive actions and creation of more effective rehabilitation programs, reducing public expenditures with hospitalizations and treatment in this population. Objective: To evaluate whether the platform entitled Activity Monitoring in Rehabilitation (MARe) has adequate concurrent validity and reliability with the StepWatch (SAM) monitor and the video camera to quantify the number of steps in people post-stroke. It was also evaluated the percentage accuracy of the MARe platform for counting steps at different walking speeds compared to SAM. Methods: Twenty-four subjects post-chronic stroke used the MARe platform and SAM simultaneously while a video camera recorded the same activities during 10MWT and TUG. The number of steps quantified by the MARe platform during the functional tests were compared with those reported by the SAM and the video. Validity was analyzed by Friedman's Anova and Spearman's correlation coefficient and the reliability by the intraclass correlation coefficient and analyzes of the Bland-Altman graphs. The accuracy percentage was calculated for each device and plotted as a function of the walking speeds during the 10MWT. Results: The MARe platform showed high correlation values for step counts when compared to SAM (rs ≥ 0.884, p = 0.000) and video (rs ≥ 0.862, p = 0.000). There were no significant differences between the devices during the repetitions at the 10MWT in the comfortable speed (p = 0.670, 0.245 and 0.787) and fast (p = 0.438, 0.052 and 0.070) and TUG (p = 0.459, 0.385 and 0.121). The MARe platform also showed excellent agreement when compared to SAM in the 10MWT at fast and comfortable speed and in TUG (ICC3,k 0.948, 0.972 and 0.973, respectively) and when compared to video (ICC3,k 0.994, 0.999 and 0.991, respectively) with values within the limit of agreement in the analyzes of the Bland-Altman graph. A high percentage of accuracy for step counting of the MARe platform was observed at all speeds during the 10MWT when compared to the SAM. Conclusion: The MARe platform showed up valid and reliable for step counting, with a high percentage of accuracy at different walking speeds, in the post-stroke population.