ESTRO 2021 Abstract Book

S629

ESTRO 2021

Purpose or Objective Worldwide, increasing number of cancer patients is resulting in a high demand for radiotherapy (RT). Considering the current level of resources, the demand can be difficult to meet and patients may be facing long waiting times. Hospital staff assesses the urgency and decides patient priority accordingly. Treatments are scheduled with the oncologic information system (OIS) in a semi-automatic process which may be time consuming and inefficient. There is a risk of uneven distribution of cancer diagnosis with some patients falling behind others. Research so far has rarely considered the influence of factors associated with patient characteristics. Our aim is to propose a strategy, which assures consistency in waiting times between cancer types when scheduling RT. This is achieved by a using a simulation technique to illustrate effects of dynamically sorting incoming referrals of patients based on priority queuing. Materials and Methods We used real data for patient referral patterns and preferred start dates for prioritization and batching of patients. Extracted data from the OIS included patients in 12 months from Feb 2020 to Feb 2021 at one large RT department in Sweden. A multi-level prioritization approach based on current clinical practice was used to create a discrete-event simulation (DES) model (Figure 1). The model was run for 4 iterations to reach stability in tracking patients on a monthly basis. Statistical comparisons were made for mean waiting time and queue lengths between the proposed approach and current clinical practice. Simulations were performed in Arena Simulation Software (v16, Rockwell Automation, United States) assuming an inflow of 120 patients/week.

Results For retrieved data of 3559 patients, preferred treatment start was 35±15 days from the referral date. A simulation run with prioritization and batching reduced the overall mean waiting time for the majority of cancer categories by 9-15 days compared to the reference scenario. The mean waiting time for acute and breast treatments was increased by 2 and 15 days respectively, however still within the preferred treatment start range. On average, the mean waiting time for pre-treatment resources was reduced 7 days and for LINACs 20 days. Without priority and batching, 2-5% patients were looped in the system waiting for resources without exiting. In simulation 2, all patients exited within their treatment start range.