DESIGNING AND TESTING A PREDICTIVE MODEL FOR POST STROKE RECOVERY - AN EXPLORATORY STUDY

Background: Stroke is the first leading cause of disability, second leading cause of death worldwide.Its recovery is important for entire rehabilitation team.

Purpose: Stroke patients who receive physiotherapy may improve in function variably. So there should be identification of different factors to predict functional outcome.

Objective: To design and test predictive model for post stroke recovery.

Setting: Physiotherapy centres of Surat.

Method: For Exploratory study, Selection of stroke patients was done as per selection criteria. After ethical clearance, & CTRI registration, subjects were assessed for 21 factors as independent variables and Fugl Meyer Motor Assessment Score as dependent variable to design  and test model.

Participants: 222 stroke patients.

Outcome measure: Fugl Meyer Motor Assessment Score & Modified Barthel Index

Results: Linear regression at confidence interval 90% was applied and to design predictive model by data of 111 stroke patients. This model was checked on new cohort of 111 stroke patients for its positive and negative predictive values.

Conclusion: These findings suggest that length of hospitalization is strongest factor for prediction of functional outcome and it was found that model has satisfactory predictive and negative predictive value.

Keywords: Stroke, Predictive factors, Fugl Meyer Motor Assessment Score