By 2050, over 35 million new cancer cases are predicted, leading to a corresponding increase in the need for chemotherapy treatments. To manage this surge, health facilities are adopting the Outpatient Cancer Center paradigm, infrastructures where patients with various types of tumors are treated. Optimizing the planning and scheduling of chemotherapy treatments in these facilities is crucial due to the limitations of several resources, including medical and nursing staff, consultation and exam rooms, and chairs and beds for drug infusion. Unlike many previous contributions that focus solely on either the Tactical or the Operational level, this work addresses both levels simultaneously, developing an approach suitable for practical application. At the tactical decision level, available resources are assigned to macro groups of cancer pathologies. At the operational level, scheduling patients' appointments involves decisions made over a shorter planning horizon. Specifically, we tackle the organization of activities in an Outpatient Cancer Center over a one-month planning horizon. This involves creating weekly assignments of visit rooms to specific cancer pathologies, known as the Master Chemotherapy Problem, and managing the clinicians’ rostering. Subsequently, we have to assign the days and time slots for visits and infusions to patients and determine the chair or bed where they will receive the infusion. We may consider two criteria: maximizing the number of scheduled patients during regular working hours and minimizing the overtime needed to meet the entire demand. The problem turn out to be computationally challenging, with various ILP/MILP models proving inadequate due to their size for real-world instances. Therefore, we developed a multi-step heuristic based on problem decomposition and an exact approach involving resources aggregation. The proposed approaches were tested on real data from an Italian hospital, demonstrating outstanding performance. The combination of these methods provides a powerful tool for solving real life cases.
Entro il 2050, si prevede che saranno diagnosticati oltre 35 milioni di nuovi casi di cancro, con un incremento della domanda di cure chemioterapiche. Per fronteggiare questa crescita, le strutture sanitarie stanno adottando il paradigma degli Outpatient Cancer Center, strutture centralizzate capaci di gestire pazienti affetti da diverse tipologie di tumori. Ottimizzare la pianificazione e la programmazione dei trattamenti chemioterapici all'interno di queste strutture è cruciale a causa delle risorse limitate coinvolte, tra cui il personale medico e infermieristico, gli ambulatori per visite ed esami, e le sedie e i letti utilizzati per l'infusione dei farmaci chemioterapici. A differenza di molti contributi presenti in letteratura, dove il livello Tattico e quello Operativo vengono trattati separatamente, con le conseguenti limitazioni, questo lavoro si prefigge l'obiettivo di trattare congiuntamente questi due livelli. A livello tattico, le risorse disponibili vengono assegnate a macrogruppi di patologie oncologiche. A livello operativo, vengono pianificati gli appuntamenti dei pazienti. In particolare, ci occuperemo dell'organizzazione delle attività in un Outpatient Cancer Center su un orizzonte di pianificazione di un mese. Questo implica la gestione dell'assegnamento settimanale degli ambulatori a specifiche patologie oncologiche, noto anche come il Master Chemotherapy Problem, e la gestione della turnazione dei medici. Successivamente, viene gestito l'assegnamento dei giorni e delle fasce orarie per le visite e le infusioni ai pazienti, così come viene determinata la risorsa, sedia o letto in base alle condizioni mediche, dove riceveranno l'infusione. Due criteri principali possono essere considerati: massimizzare il numero di pazienti gestiti all'interno del regolare orario lavorativo, e minimizzare il tempo di straordinario necessario per soddisfare l'intera domanda. Il problema è computazionalmente difficile e, i vari modelli ILP/MILP impiegati si dimostrano inadeguati in un contesto reale a causa delle loro dimensioni. Pertanto, si introduce una euristica multi-step basata sulla decomposizione del problema, e un approccio esatto basato sull'aggregazione delle risorse. Gli approcci proposti sono stati testati su dati reali di un ospedale italiano e hanno dimostrato prestazioni eccellenti.
Optimization approaches for joint tactical and operational level in an Outpatient Cancer Center
Ceffa, Paolo
2023/2024
Abstract
By 2050, over 35 million new cancer cases are predicted, leading to a corresponding increase in the need for chemotherapy treatments. To manage this surge, health facilities are adopting the Outpatient Cancer Center paradigm, infrastructures where patients with various types of tumors are treated. Optimizing the planning and scheduling of chemotherapy treatments in these facilities is crucial due to the limitations of several resources, including medical and nursing staff, consultation and exam rooms, and chairs and beds for drug infusion. Unlike many previous contributions that focus solely on either the Tactical or the Operational level, this work addresses both levels simultaneously, developing an approach suitable for practical application. At the tactical decision level, available resources are assigned to macro groups of cancer pathologies. At the operational level, scheduling patients' appointments involves decisions made over a shorter planning horizon. Specifically, we tackle the organization of activities in an Outpatient Cancer Center over a one-month planning horizon. This involves creating weekly assignments of visit rooms to specific cancer pathologies, known as the Master Chemotherapy Problem, and managing the clinicians’ rostering. Subsequently, we have to assign the days and time slots for visits and infusions to patients and determine the chair or bed where they will receive the infusion. We may consider two criteria: maximizing the number of scheduled patients during regular working hours and minimizing the overtime needed to meet the entire demand. The problem turn out to be computationally challenging, with various ILP/MILP models proving inadequate due to their size for real-world instances. Therefore, we developed a multi-step heuristic based on problem decomposition and an exact approach involving resources aggregation. The proposed approaches were tested on real data from an Italian hospital, demonstrating outstanding performance. The combination of these methods provides a powerful tool for solving real life cases.File | Dimensione | Formato | |
---|---|---|---|
THESIS_Paolo_Ceffa.pdf
non accessibile
Descrizione: Thesis Paolo Ceffa
Dimensione
3.37 MB
Formato
Adobe PDF
|
3.37 MB | Adobe PDF | Visualizza/Apri |
Executive_Summary_Paolo_Ceffa.pdf
non accessibile
Descrizione: Executive Summary Paolo Ceffa
Dimensione
472.08 kB
Formato
Adobe PDF
|
472.08 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/223123