Microsimulation Models for Simulating Pathways of Patients with Mental Diseases

Predicting demand for care is necessary to provide sufficient capacities in hospitals and enable better planning to cope with challenges like changes in the structure of the population. Emphasis in this work is put on the analysis of the data and on the parametrization of a simulation model for the paths of patients with mental diseases through the health care system. Survival analysis and model selection methods are used for this purpose. Data on patients and their treatments is analyzed with methods of survival analysis. Different methods for modelling the survival and hazard function are presented and compared. Hereby, the focus is on the cox model. It is used to model the hazard function and can be extended to analyze multiple events. With the use of model selection methods the significant parameters are determined. Only these are included in the simulation model. These methods shall help to raise the quality of the parametrization and therefore the whole simulation model. In the microsimulation model, every patient has a particular set of parameters and can be in one of several predefined, exclusive states. The events are implemented as state changes. The probabilities for the events are calculated using the hazard functions. These are estimated with several extensions of the cox model. The considered events in the simulation are readmissions to hospital and contacts to ambulant psychiatrists. The simulation runs for a predefined time span and the sequence of events of each patient is tracked. After the simulation, the individual paths of the patients as well as aggregated quantities such as the overall numbers of certain events are analyzed. Simulations for different populations are performed. The results for various scenarios are presented. Scenarios with and without contacts to a psychiatrist are considered as well as different maximum numbers of admissions. Also, the subpopulations are compared. For example, differences in the results for diagnosis groups are encountered. These simulations shall lead to an improvement of the prediction of the pathways of the patients and therefore help to evaluate interventions like treatment changes the health care system or the utilization of the capacities in hospitals