Districting Decisions in Home Health Care Services: Modeling and Case Study

Abstract

Home health care (HHC) services are a growing segment in the global health care industry in which patients receive coordinated medical care at their homes. When designing the service, HHC providers face a set of logistics decisions that include the districting configuration of the coverage area. In HHC, the districting problem seeks to group small geographic basic units-BUs (i.e., city quarters) into districts with balanced workloads. In this work, we present a modeling approach for the problem that includes a mixed integer linear programming (MILP) formulation and a greedy randomized adaptive search procedure (GRASP). The MILP formulation solves instances up to 44 BUs, while the GRASP allows to solve instances up to 484 BUs in less than 2.52 min. Computational experiments performed with a set of real instances from a Colombian HHC provider, show that the GRASP can reduce workload imbalances in a 57%.