Modèle de données pour l’analyse des flux de patients à l’hôpital

Abstract

Summary Improving hospital operation in terms of quality of care, patient safety and efficiency requires the capacity to routinely analyse patient flow. To that end, a generic data model to represent patient trajectory across various hospital «compartments» (e.g. care units, clinical services, care levels) was developed. To build the model, time intervals during which a patient remains in the same set of compartments (e.g. the same care unit, the same clinical service, the same care level) were first identified for each stay using data extracted from the hospital information system. Each of these time intervals was then split into constant timesegments (i.e. days, hours) and the number of minutes the patient was present in each of them was counted. In addition, an entry flag (or exit flag) was attached to a timesegment if during this time-segment a patient entered (or left) a given compartment. The data model which was implemented on a MySQL-Infobright database using a starjoin schema allows, via basic SQL queries, a large array of analyses and detailed graphic visualisations of patient flow using numerous different time steps (month, weeks, days, weekdays, hours, etc.). By this means it dramatically increases the hospital’s capacity to address patient flow issues and prepare the way for patient flow simulation.