32. An automated doctor rostering decision support tool for Knysna Hospital
Invited abstract in session MC-3: Staff scheduling, stream Sessions.
Monday, 11:00-12:30Room: NTNU, Realfagbygget R9
Authors (first author is the speaker)
| 1. | Klara Engelbrecht
|
| Department of Logistics, Stellenbosch University | |
| 2. | Linke Potgieter
|
| Logistics, Stellenbosch University |
Abstract
Hospitals often face a shortage of doctors, placing significant strain on the available staff to meet patient care demands. In public hospitals in South Africa, physician and nurse rosters are typically set up manually by the clinical manager. This is a time-consuming task that is difficult to accomplish with the number of regulations, preferences, and fairness that must be considered. In this study, a mixed-integer programming model (MIP) was formulated that can be used to automate the rostering task for physician scheduling at Knysna Hospital in the Western Cape of South Africa. Hard constraints ensure that all shift demands are met and hospital regulations are followed, while soft constraints promote fairness, minimises consultant usage, and allow doctors sufficient weekends off. Additionally, the model was integrated into a decision support tool with a user-friendly interface designed to automate the doctor scheduling process for the clinical manager at Knysna Hospital. During the development of the model and user interface, frequent discussions with the clinical manager ensured the user-friendliness of the interface and that the model correctly reflects context specific constraints and variables. In this talk, the model and decision support tool will be presented, as well as results from model analysis and comparison with the actual Knysna Hospital schedule. It was found that neither Knysna Hospital nor the model achieves perfect fairness between doctors, but the mode
Keywords
- Workforce planning and scheduling
- Decision support
Status: accepted
Back to the list of papers