Integrated Care Pathways of Elderly in Hospital – PSIH

The work carried out over the period revolves around two axes: macroscopic modeling of the hospital and analysis of the care pathway of multimorbid patients.

This first axis, carried out in collaboration with the Saint-Etienne University Hospital, led to the creation of a discrete-event simulation model. This tool will allow us to evaluate the care pathways of multimorbid patients and measure the weight of these patients on the system.

The second axis is related to multimorbid patients. It consisted of statistical work on the identification of the patients under study, the preparation of data and the application of analysis methods. To identify patients, we conducted a literature review on multimorbidity scores. These scores were analyzed by the project team. Some were selected to identify multimorbid patients. After the extraction, the database was cleaned up for analysis. Different statistical methods have been applied and a process mining analysis has highlighted differences in the care pathway of the multimorbid population.

The two major results of this thesis are the realization of the systematic literature review and the hospital simulation model. The first allowed us to deepen the definition of multimorbidity while the second will be the basis of all studies related to the care pathway, and in particular improvement scenarios.
Process mining studies have highlighted the fact that the majority of multi-morbid patients have a simple hospital journey, passing through a service or group of services in a sequence that appears to be under control. Among the «complex« pathways, there are sequences that are a medical necessity and those whose erratic course reveals a difficulty in care due to multimorbidity. A study to distinguish these categories will be conducted to identify the population on which we will act.

The perspectives concern the development of an integrated approach making it possible to (1) automatically build the care pathways of a multimorbid population from medico-administrative data, (2) characterize the «good« and «bad« care pathways, (3) propose organizational scenarios for better follow-up of the patients via a «digital twin« approach.

Literature review on multimorbidity (under review).
ORAHS 2019 - Oral communication.
WSC 2020 - Article in proceedings (under review).

In a context of an aging population and an increase in life expectancy related to the progress of medicine and a better prevention policy, the demand for geriatric care is likely to increase in the incoming years. This project aims to provide a decision support tool for better follow-up and efficient care of the elderly patient (EP) in the hospital from his/her first admission. The purpose of this tool is to rationalize all required resources to take care of EP by decompartmentalizing the services usually organized in silos and by favoring the collaboration of health professionals. The secondary objective is to propose a performance evaluation tool for the integrated care pathway of EP in order to quantify the impact of the proposed measures and prioritize the most efficient ones. This project is composed of 3 main phases. A first phase corresponding to the first scientific challenge relates to the modeling of the clinical pathways. A mixed approach to model the health status of EP combined with pathway modeling will be proposed. We are considering a similar approach in order to capture the complexity of the journey in the hospital and to emphasize the impact of the medical decisions on the health status of EP. In order to feed the model, a data collection will be carried out in collaboration with the University Hospital of Saint-Etienne and the Hospital of Roanne. Observations will be made in the facilities to understand clinical pathways, and hospital data will be used to build clinical pathways models and health status models. A second phase corresponds to the design of the hospital reorganization. For this, several implementation scenarios will be proposed via the optimization of the necessary resources. The model proposed in Phase 1 will be used to build a new service adapted to the population of interest while respecting the strong constraints of the hospital (constant resources, limited space). Several mixed-integer linear models based on a formalization of the problem using queueing theory will be proposed. Finally, the third phase aims at evaluating the performances of the organizational scenarios that will emerge from phase 2. A mixed simulation model with discrete / multi-agent events will be proposed with a view to a systemic approach. A joint assessment of flows and medico-economic aspects will be proposed through collaboration with a health economist. The objective is to simulate several possible scenarios to find the best management policy (discharge policy, diagnosis related group, ...) and the best sizing of services in the hospital in terms of quality of service / length of stay and cost (economic study). The first admission control of an EP is critical. It is crucial to propose a control strategy of the path of PA from its first admission to the hospital through a coordination of its journey in the hospital and after the stay, to avoid costly and disastrous rehospitalization for the health of the EP. Thus the approach proposed at the end of the project will provide healthcare professionals with methodological guides to better support the EP care, from his/her first visit, thus minimizing the risk of deterioration of health and increasing his/her life expectancy.