Designing and testing a predictive model for post stroke recovery an exploratory study

Abstract

quotBackground: Stroke is the second common cause of death and the fourth common cause of disability worldwide. Its recovery is important for the entire rehabilitation team. newlinePurpose: Stroke participants who receive physiotherapy may improve variably. So there should be the identification of different factors to predict post-stroke recovery. newlineObjective: To design and test a predictive model for post-stroke recovery. newlineSetting: Physiotherapy centers of Surat. newlineStudy design: Exploratory study newlineMethod: After ethical clearance, the CTRI number was taken. The selection of stroke participants was done as per selection criteria. In phase 1, a total of 111 stroke participants were assessed for 21 predicting factors as independent variables and Fugl Meyer Motor Assessment Score as a dependent variable to design a predictive model. In phase 2, this model was tested on a new cohort of 111 stroke participants by the investigator. In phase 3, the model was also tested by 16 external validators on a new cohort of 35 stroke participants. newlineOutcome measure: Fugl Meyer Motor Assessment Score newlineResults: Linear regression at a confidence interval of 90% was applied and found that three factors namely length of hospitalization (p=0.000), type of stroke (p=0.043), and post-stroke duration (p=0.073) as most significant factors. These factors were used to design the predictive model. The model when tested by investigator and external validators had overall accuracy 0.69 and 0.72 respectively. newlineConclusion: Designing and testing a predictive model for post-stroke recovery was exercised that was tested by investigator and external validators. Length of hospitalization is the most significant factor followed by the type of stroke and post-stroke duration to predict post-stroke recovery. newlineKeywords: Stroke, Predictive factors, Fugl Meyer Motor Assessment Score, Post-stroke Recoveryquot newline