Efficient and interactive scheduling of cancer treatments during a pandemic

A current side-effect of the COVID-19 pandemic is a reduction in the number of patients with cancer seeking treatment or being referred for treatments. Some of these treatments may become critical if delayed for too long and many excess cancer deaths may occur. Efficient scheduling of treatments and clinical staff is compulsory when preventing a large backlog of outstanding treatments. Furthermore, both patients and staff may be unable to abide by a previously arranged schedule, e.g. they are required to self-isolate, therefore recovery approaches are needed to maintain an efficient clearing of treatments. Finally, patients and staff need to interact with any system supporting re-scheduling in a way suited to their cognitive needs. We will adapt and extend an existing award-winning tool that allows querying schedules and returns natural language explanations for why changes to a schedule may be infeasible, inefficient or violating known fixed decisions. The original tool was deployed for nurse rostering but its underlying framework offers an effective, general-purpose way for hospital administration to interact with schedules and react to required changes, e.g. if staff are unavailable. We will apply this framework for scheduling cancer patient treatments.