Patient profiling for mental healthcare management optimization in developing countries using Datamining

CPU Ibn Rochd is a reference centre for mental healthcare in Casablanca, Morocco. It has a heavy workload, overwhelming the resources allocated. However, a prior knowledge of this centre's population characteristics makes it possible to increase the healthcare efficiency by, for instance, redistributing the healthcare services according to the centre's patient profiles. Datamining techniques allow such a procedure. Based on patient data recorded in the CPU consultation information system in 2015, we applied clustering, a Datamining technique, for the discovery of different patient profiles. Then we displayed the results in the form of a health mapping to explore the clusters' geographic dimension and facilitate their interpretation by psychiatrists. Spread over two territories of Casablanca, three patient profiles have been discovered. The identifying variables for each profile are diagnosis, sex of the patient, toxic habits and family status. The results were discussed in a psychiatric staff, and recommendations regarding resource allocation and mental health management optimization have been issued. Hence, Datamining can serve as useful tool for upper management as it can enable healthcare executives to make efficient knowledgedriven and contextualized decisions.