A multi-objective approach for reducing Patient’s inconvenience in a generalized home healthcare delivery setup

In this work, we present a mixed-integer programming model for a multi-objective home healthcare delivery problem. The proposed problem is modeled with minimum assumptions about the procedure attributes and can handle most of the commonly imposed restrictions in the field of home healthcare delivery. Under the imposed restrictions, the model is designed to provide selection (for caregivers and patients), assignment, scheduling, and routing decisions. In addition to some minor modifications in ‘workload balance’ constraints, a major focus of the work is to improve the quality of the schedule for the selected patients. To achieve this, we define and calculate the inconvenience caused by the unnecessarily scattered visits and their overlap with the patient-specific inconvenient time window. The model minimizes the total inconvenience cost for patients against the competitive goals of other stakeholders. Higher net profit, minimum loss of employed labor, balanced workload among staff, and maximization of fully served patients have been included as the other objectives. To solve the instances of the proposed home healthcare delivery problem, an efficient implementation for a reference point-based non-dominated sorting genetic algorithm (NSGA-III) is developed. After extensive parameter tuning using the Taguchi method of experimental design, the algorithm is used for generating a diverse set of non-dominated solutions for the decision-maker. In addition to the comparing the performance of NGSA-III with Multi-objective particle swarm optimization and Multi-objective grey wolf optimizer, experiments are also carried out to establish the relationship between patient convenience and net profit.