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

Limitations

Data quality is often an issue for discrete event simulations, as they aim to replicate real world processes in a virtual model which requires sufficient data. The data used in this case study (waiting list figures, attendances, arrivals, removals etc.) are revised at each quarterly publication as there can be issues with initial data collections. This leads to small variations and the need to re-calculate the distributions following each publication. The impact of this on short and medium term projections is mitigated by having wide distributions in the model , however this reduces the utility of the model for longer term projections.

As this was a case study and exploration of the method we considered the case study subject at a Scotland-wide level and specialty level. This meant that the model still relied on some averages and potentially misses some of the variance between health boards which could impact the flow of patients Scotland wide.

It is important to note that in the case study we did not consider the impact of diagnostics on the flow of patients, or the impact of return outpatients (and how the number of return outpatients would change over time) on consultant availability. The model also assumed that conversion rates and direct TTG rates were stable in time and independent of any changes to surgical activity. A suggested future expansion of this simulation model would be to explore in more detail the relationship between surgical activity and additions to the new outpatient list. With these limitations the model outputs are more illustrative of the potential impacts of any productivity scenarios in a highly idealised system.