Predicting cost in the health care environment is a challenging issue for healthcare professionals. Chronic disease is a long term disease that requires life time care. Physicians need to keep tracking patients' status over time including routine medical examinations. In this paper, we investigate the healthcare service scenario for chronic disease care. Then, we propose a predicted cost model for the investigated healthcare service. The cost model is based on the prediction of utilization and attendant costs through the development of a stochastic model, specifically a first-order Markov chain, can be adapted to specific diseases and/or events. The proposed model is at the initial stage and may require testing using a large administrative database of patients and hospital operational costs.