Markov chain prediction of the recurrence of common diseases in a community and their corresponding proportions

Abstract

Multi-state Markov models have been used to analyze the recurrence times, the average first passage times and proportions of individuals who are likely to suffer from common diseases in the long run with in the community. Data was collected from two different places using two methods. At Busabala Nursing Home, 358 medical records were used to get the necessary data while at Kibuli Hospital, 324 patients were interviewed to obtain the necessary data. At both Busabala and Kibuli, diseases such as Malaria, Typhoid, Peptic Ulcers, Urinary infections, Pneumonia, Cough, Urinary Infections, Diarrhoea and Hypertension were common. 44%, 10%, 5%, 8%, 10%, 7%, 12% and 5% of individuals are estimated to suffer from Malaria, Peptic Ulcers, Typhoid, Cough, Urinary Infections, Diarrhoea, Pneumonia and Hypertension in the long run respectively using data of individuals inter- viewed from Kibuli hospital while 32% 8%, 8%, 21%, 8%, 5%, 4% and 14% of individuals are likely to suffer from Malaria, Malaria and Typhoid, Peptic Ulcers, Urinary Infections, Pneumonia, Diarrhoea, Typhoid and Hypertension in the long run respectively basing on the data gathered at Busabala nursing home. At both places Typhoid was least recurrent among individuals while Malaria emerged as the most recurrent disease. The mean first passage times between the common diseases at both places was also computed as well as the average time interval with in which individuals fall sick which was obtained three months.