Stochastic scheduling of chemotherapy appointments considering patient acuity levels

Appointment scheduling for medical diagnostic centers considering time-sensitive pharmaceuticals: A dynamic robust optimization approach

“…Other studies have targeted the business-oriented aspects of a healthcare provider ( Gupta & Wang, 2008;Stanciu, 2009 ). The analytical approaches including, queuing theory ( De Vuyst, Bruneel, & Fiems, 2014;Kortbeek et al, 2014 ), integer programming ( Bastos, Marchesi, Hamacher, & Fleck, 2019 ), dynamic programming ( Green et al, 2006;Wang, Chen, & Xu, 2018 ), stochastic programming ( Alvarado & Ntaimo, 2018;He, Sim, & Zhang, 2019;Karakaya, Gul, & Celik, 2022;Legrain, Fortin, Lahrichi, & Rousseau, 2015 ), and non-analytical tools such as discrete-event simulation ( Pérez et al, 2013;Rau et al, 2013 ) and online tools ( Jain, Mohan, Zacharia, & Sanders, 2022 ) have been the leading frameworks to study different perspectives of patient appointment policies.…”

Clustering and Stochastic Simulation Optimization for Outpatient Chemotherapy Appointment Planning and Scheduling

“…A two-stage stochastic integer programming formulation considering the infusion stage only have been developed in [ 35 ]. In similar works, the authors of [ 36 , 37 , 38 ] used stochastic integer programming models and heuristics.…”

“…Therefore, our work falls under the category of stochastic OCA planning and scheduling. However, different from the single-stage models in [ 35 , 36 , 37 , 38 , 64 ], we considered the multi-stage process, including the patient registration, triage, blood sample extraction, blood test, health condition review, discharge, as well as drug order activation, validation, preparation, delivery, and administration. Moreover, we took into account the different required human and physical resources of each stage, not only the nurses and chairs.…”

“…Nevertheless, we used the acuity levels of patients to restrict the simultaneous nurse tasks. The models in [ 35 , 36 , 37 ] used a positive integer variable to decide the appointment time without time slots, while the models in [ 38 , 64 ] assigned patients to time slots using a binary variable. We combined both techniques by using discrete variables and restricting their values to a set of feasible values associated with the time slots in the planning days.…”

“…The models in [ 37 , 64 ] were solved to optimality using optimization solvers. In contrast, heuristics were used in [ 35 , 36 , 38 ] to find approximate solutions within a reasonable computational time. We searched for solutions by utilizing an artificial intelligence-enabled General-Purpose Optimizer (GPO) developed by OptQuest.…”