Using Discrete Event Simulation to explore "what-if" waiting list scenarios in NHS Scotland

Conclusion

Discrete event simulation using Simul8 proved to be a relatively quick way to build a representation of a planned care patient journey, and test the impact of proposed productivity scenarios. This modelling method depends heavily on having good quality data available in order to build robust distributions. Any implementation will be somewhat constrained by this due to disruption that arose from the COVID-19 pandemic which reduces the timeseries of available data. Nevertheless, this could be easily translated onto other specialties of interest as long as the necessary data to build the distributions is available.

The cataract surgery case study was useful for highlighting the importance of understanding the flows in the system, and the impact that unidentified flows can have on model projections. With direct TTG additions excluded, any model projections for productivity scenarios would have overestimated the positive impact of increased activity on the TTG list size. The case study also demonstrated the limitations that arise from data quality, with growing uncertainty in the projections due to diverging simulation dynamics. This method is therefore more suitable for short and medium term impact assessment if it were to be used to make operational decisions – from a theoretical perspective it is perhaps useful that the simulation does give some more quantitative visual on the wide range of possible outcomes following an intervention, and highlights the need to understand uncertainty in projections when using them for decision making.

From the case study itself, we can conclude that the increased cataract throughput productivity scenario does not have a significant impact on TTG list size in the mode where patients are re-added to the TTG list for the second eye – it could however have an impact on waiting times that this model is unable to measure. However, in a scenario where there were fewer direct TTG additions i.e. there were more immediate sequential bilateral operations, the productivity scenario would have a stronger positive impact. This also suggests that depending on local processes, different health boards could appear to be performing better or worse following such productivity improvements mostly due to how they govern flow onto the TTG list.