Abstract
A decision support model was build in order to indicate the needs for access to physiotherapy services for individuals affected by cerebrovascular accident (stroke). Methods: The research was characterized by a longitudinal study cutout of individuals affected by stroke living in the cities of João Pessoa and Cabedelo, in Paraíba State, who were admitted to a reference public hospital in the region. Results: A decision support model based on Artificial Neural Networks, feedforward type and multilayer perceptron. It was trained using a supervised backpropagation algorithm. It was composed by 15 neurons in the input layer, two hidden layers with two neurons each and an output layer with two neurons. Its accuracy was measured as 97.4%. Conclusion: The decision model developed was able to indicate the access needs for individuals affected by stroke to physiotherapy services. So, this model can help health professionals in this decision making.