Saving costs and improving quality of care for hospitals by reducing variation in hospital bed occupation through the master surgical schedule
This item has file attachments that are restricted and can only be viewed from the TU Delft campus network.
Due to rising costs and an increasing demand for health care services Dutch hospitals are looking for ways to reduce costs while maintaining or increasing the level of patient satisfaction. Reducing variation in bed occupation in common nursing wards could save costs by reducing the amount of required beds and nurse staffing, and increase quality of care by stabilizing the nurse workload. Adjusting current OR schedules offers potential to reduce variation because they have often been developed without explicit consideration for the wards due to the uncertain relationship between them. The main research question is therefor ’How can the management of Dutch hospitals implement adjustments to the operating room schedule in order to reduce the variability of bed occupancy in the patient wards?’.
The variability in bed occupation can be seen over years, weeks and days. The weekly variation can be influenced by making adjustments to the master surgical schedule (MSS), a repetitive two weekly schedule that assigns all avail- able OR blocks to a surgical specialism. An OR block is a morning or afternoon session in a certain OR on a certain day. Analysis has shown that for the Sophia hospital, which is used as a case study in this research, only 16% of total admis- sions should be considered because they can be directly influenced by adjusting MSS. Research has shown this group exhibits the most variation.
Swapping OR blocks in the MSS is difficult because analysis revealed the MSS was not as repetitive as thought and differed slightly each cycle due to for example specialisms trading or cancelling blocks. To solve this the MSS was reconstructed using historical data over an entire year to find for each block which surgical specialism has used the block the most cycles for elective surgeries. The resulting reconstructed MSS represents the structural influence the MSS has had over that year. This reconstructed MSS can be adjusted by swapping MSS blocks in order to reduce the structural bi-weekly variation caused by the MSS. This variation can be seen in wards, specialism clusters, wings and the entire hospital. This research will focus on making adjustments to the structural master surgical schedule that reduces bi-weekly variation in common wards per surgical specialism cluster.
Current literature has a low rate of implementation because the influence of stakeholders is often disregarded, and rarely sees the MSS as a tool to level ward occupancies. Optimization approaches are limited because an optimal MSS is an unobtainable objective given the variety of objectives of stakeholders. This thesis presents a visualization allowing non-technical users to identify beneficial MSS block swaps between blocks from within a single surgical specialism cluster. A decision support tool (DSS) allows stakeholders to quickly evaluate the effect of those swaps on variability in their bed occupation in nursing wards. The DSS does this by adapting an implementation of the mathematical model from van Berkel that is able to relate the master surgical ward to the occupation in the nursing wards.
The DSS is able to predict the effect of the structural MSS accurately enough to evaluate changes in structural variation patterns caused by MSS block swaps. Experimentation has shown the method is able to identify a potential reduction in peak occupation by several beds for both the neurosurgery- and common surgery cluster in Isala. Another observation was that an unequal division of MSS blocks over the week caused peaks in occupation that could not be compensated with MSS block swaps. This led to the suggestion of a multi- stage approach to bed leveling through the MSS, where the division over the week of MSS blocks from a specialism cluster or hospital wing is equalized as much as possible. MSS block swaps can then be done between clusters or wings (discussed in appendix D.1), after which swaps within clusters or wings can be applied. The developed decision support tool helps stakeholders with this last step and enables specialisms to find beneficial swaps themselves. The research is supplemented with a process design that addresses how to use and embed research efforts in a hospital environment with a strong influence of stakeholders with conflicting interests. The decision support tool can benefit a hospital looking to realise a certain reduction in peak bed occupancy in order to reduce the amount of beds. The DSS can also support a sustainable process solution with incremental adjustments to the MSS, in order to allow hospital to consider ward occupation when adjusting the MSS.
This thesis has not answered the entire scope of the main research ques- tion but has identified a range of possible measures, intervention levels and a multi-stage approach to bed leveling. The method provides practical means to reduce bi-weekly variation in Dutch common nursing wards through the master surgical schedule while accounting for the influence of stakeholders. The results have yielded valuable insights for Isala, where the research will be applied and developed further. Bed leveling is not only an important issue in Isala but also becoming increasingly important in other Dutch hospitals looking to better uti- lize their expensive resources. This research has contributed to the academic field with a unique non-optimization approach to adjusting the MSS in order to reduce variation in bed occupation. Contrary to other research the method takes into account the influence of stakeholders by enabling stakeholder inter- action to find adjustments to the MSS. In conclusion, the developed method might benefit the rapidly evolving Dutch health care sector that is pressured to improve its cost-effectiveness.