An intelligent optimization strategy for nurse-patient scheduling in the Internet of Medical Things applications

Abstract

In the last years, the world has witnessed a potential increasing in the patient number resulted from the increasing number of aged persons along with the emergence of new virus and diseases. This imposes a high pressure on hospitals that suffer from a shortage of medical staff, personal equipment and adequate interventions to overcome such a challenge. Particularly, nurse scheduling is becoming a crucial operation to hospitals in order to efficiently handing the patents and increasing the performance of health system. In this paper, we present an efficient Nurse-Patient Scheduling (NPS) mechanism that is based on the patient classification according to the severity levels of their vital signs. The main goal of NPS is to balance the workload of nurses and it consists of three phases: patient monitoring, patient classification, and nurse scheduling. The first phase aims to periodically monitor the patients using a configurable window time and collect their vital signals through a set of biomedical sensors. The second phase allows for the extraction of prospective features among the collected data then to classify them according to a set of predefined criteria such as patient criticality level, patient age, and the allocated treatment time of each patient. In the last phase, we propose a novel scheduling algorithm that combines both genetic and particle swarm optimization methods in order to find the best scheduling assignment of nurses over patients. We performed simulations based on real health data and we demonstrated the performance of NPS mechanism in terms of obtaining optimal of nurses to patients according to the predefined criteria.