Strategic room type allocation for nursing wards through Markov chain modeling.

Providing patients with the best possible care is the most essential function of any hospital. In an increasing number of countries, hospitals are governed by the number of patients they are able to attract and the corresponding services they provide for patients. One such service, which is often of significant importance for patients, is the option to choose their room type. Hospital decision makers would benefit from a strategic method for optimizing the configuration of room types among nursing wards by distinguishing between patients who prefer private rooms and those who have no preference concerning whether they are assigned to a private or shared room. Such a decision support method is currently non-existent, therefore the goal of this study is to provide a methodology for hospital management. Specifically, a mixed modeling approach is proposed which evaluates the patient flow behavior by applying a Continuous-Time Markov Chain within a heuristic search procedure. This procedure recursively improves a configuration of rooms among the wards by sampling from a gradually improved interpolation of an objective function. Based on patient data obtained from both a Danish and Belgian hospital, the performance and robustness of the proposed approach is validated through various numerical experiments, demonstrating that solutions within a relative gap of 1% from the optimum are attained in most cases.