Use of Markov models to estimate transition probabilities in a follow up study
Abstract
In a follow up study to assess treatment of BV. 207 female adults presenting with BV were
treated. Follow up information was obtained at one week. one month. two months and three
months. One outcome measurement at each follow up was BV status. Since BV status could
be influenced by various habits and practices. the outcome is analyzed by transition first order
Markov model. We relate transition probabilities to covariates and previous outcomes through
a logistic regression model with Markov structure. This approach allows us to assess not only
the baseline characteristics but also the effects of past outcomes in analyzing longitudinal
binary data.
Citation
Masters of Science in BiometrySponsorhip
University of NairobiPublisher
Department of Mathematics, University of Nairobi