ORAHS2025
Abstract Submission

150. The impact of forecasting: Dynamic appointment scheduling with elective and semi-urgent patients

Invited abstract in session ME-3: Appointment scheduling, stream Sessions.

Monday, 15:30-17:00
Room: NTNU, Realfagbygget R9

Authors (first author is the speaker)

1. Jedidja Lok - Visser
CHOIR, University of Twente

Abstract

In appointment scheduling, it is a common practice to reserve a number of slots for (semi-)urgent demand arrivals, that require service on short notice. The other slots are then given to clients that request an appointment upfront. To determine the number of reserved slots, the (semi-)urgent demand arrivals are often modelled as a distribution with static or seasonal distribution parameters. However, in many appointment scheduling processes, more information becomes available about the urgent demand arrivals over time. For example, in a radiology department, the number of patients present in the ED could forecast the required number of emergency scans.
In this study, we propose near-optimal scheduling policies that reserve slots for (semi-)urgent clients, given updated information on the arrival distribution of (semi-)urgent clients in the near future. We formulate the sequential decision-making problem as a Markov decision process, and we develop two approximate dynamic programming approaches to solve our problem, which we compare with a heuristic in a Monte Carlo simulation. We test our approaches on a real-life case study in a Dutch neurology department, where we can use remote patient monitoring information of the stroke patients to forecast the number of semi-urgent requests for outpatient appointments. We discuss first results on this case study and additional theoretical instances, and present managerial implications of near-optimal policies that we derived.

Keywords

Status: accepted

Back to the list of papers