Asian Research Journal of Mathematics

This paper sort to establish some risk factors of stroke and to estimate the effect of the covariates at different levels of disease states. A review of the literature on stroke risk factors did not reveal any article that estimate possible covariate effect of transition. To fill this gap, we incorporate the covariates in a Continuous Time Markov Model in multi-state models to observe the transition rates of the patients at two-monthly intervals for two years. Patient variables are age, sex, location of the patient, local treatment, smoking, alcohol intake, and hemiparesis. Secondary data from stroke patients under rehabilitation at the Tamale Teaching Hospital from 2014 to 2019 was used. It is observed that males recover earlier in all states compared to females. Old and Older patient groups have some probability of transiting to less severe states; and they have similar probabilities of transiting from mild to a more severe states. The youth are better off in the severe state than the two older groups. In severe state, a patient without local treatment lives less than two months before death, whiles patients who seek local treatment may remain with severe stroke for at least two months before transiting to a less severe state. Thus left hemiparesis patients are about twice less likely (0.06738) to transit to the severe state than right hemiparesis patients (0.1207).