With a travel bill of approximately £1 million per year, sub-optimal use of its estate, and an agenda to move care closer to home, Royal Cornwall Hospitals NHS Trust (RCHT) has long been keen to better organise its provision of outpatient services across Cornwall. The service is comprised of around 30 clinics and around 250 consultants visiting these clinics. Therefore, mathematical optimisation approaches are necessary in order to try to find optimal configurations.
A three-phase project aims to answer the broad question “What’s the best way to distribute outpatient clinics and clinicians across the existing sites in Cornwall, to minimise travel for patients and clinicians, and maximise usage of the sites where clinics are based?”.
The first phase sought to determine which clinics should be held and where to minimise travel distances for patients. A second phase looked at how this configuration would differ if minimising travel distances for consultants. A third phase will use decisions on “optimal” configurations identified through the first two phases to build a scheduling model that looks at how these clinics should be scheduled. If there is significant disparity between the optimal configurations identified in phases one and two, multi-objective optimisation techniques will be used to identify a range of optimal configurations, allowing the collaborators to make trade-off decisions.
Outputs and impact
The collaborators have been wanting to optimise their outpatient service configuration for some time, and plan to use the results of the larger project in order to redesign their service provision. The results from phase one of the project have been directly used to inform the development of phases two and three.
There is likely significant potential for a research paper to be generated from the larger project outlining how geographic modelling and optimisation methods can be used to optimise resource allocation in the context of outpatient clinics.
Aims and activity
The main aims of the first phase of the project were to answer the following questions:
- Based on current activity, which outpatient clinics should be held at which sites in Cornwall in order to minimise travel times for patients?
- Is this configuration likely to change based on forecasts of activity in five years’ time?
We built an Excel-based geographic model to determine the optimal configuration of where to run each outpatient clinic, trying to minimise the travel time of patients who would access those clinics.
For the forecasting, we used a Seasonalised Linear Regression Trendline to forecast trends in activity, taking into account seasonal fluctuations.
Phase One findings
- The optimum centralisation scenario predicted by the model is for all outpatient clinics to be run at RCH Treliske, West Cornwall Hospital and Bodmin Hospital.
- The model predicts that for 40% of patients this would lead to a reduction in travel time, and for a further 49% there would be no change in travel time. Only 1% of patients would see an increase of more than 30 minutes, and no patient would see an increase beyond 35 minutes.
- If this scenario was adopted, it is predicted that significantly reduced capacity for outpatient clinics would be needed at RCH Treliske, as there would be 123,176 fewer clinic visits per year. However, significant additional capacity would be required at West Cornwall Hospital (86,566 additional clinic visits per year) and Bodmin Hospital (119,335 additional clinic visits per year).
- The 1% of patients who would have to travel the furthest if all clinics were centralised to RCH Treliske, West Cornwall Hospital and Bodmin Hospital have a near identical proportion of older people in the population as compared to the remaining 99%, and a near identical level of deprivation (in fact, they are marginally less deprived). The analysis suggests that the patient transport provision is much lower in these areas – however, this is most likely due to ‘border effects’, as these patients are all located close to the border with Devon, and therefore many are likely to be accessing services out of county. This can be seen in the lower levels of attendance at Cornwall outpatient clinics for such postcodes in Phase 1 of the project.
The findings were presented to RCHT Outpatient Improvement Group in Summer 2016; this will inform decisions regarding a potential re-configuration of outpatient clinics in Cornwall.
Aims and activity
Phase two of the project built on the findings from phase one, and aimed to answer the following questions:
- What would be the impact on patient travel time of centralising outpatient services at Treliske and West Cornwall Hospital, assuming:
- All outpatient clinics could be run at both sites?
- Haematology, oncology, breast surgery clinics could only be run at Treliske, and all other clinics could be run at both sites?
- What would be the impact on patient travel time of centralising outpatient services at Treliske, West Cornwall Hospital and Bodmin Hospital, assuming:
- All outpatient clinics could be run at all three sites?
- Haematology, oncology and breast surgery clinics could only be run at Treliske, and all other clinics could be run at all three sites?
- How many more patients would need to be seen per year at each clinic at each site for each of the above four scenarios?
- How many more patients would need to be seen per year at each clinic at each site if services were located closest to patients (such that patients could always access the clinic they needed at the nearest site to their home)?
- What is the age and deprivation profile of those travelling furthest to the three candidate centralisation sites, and how does this compare to the rest of the population in Cornwall accessing outpatient services?
- How does the non-emergency transport provision in Cornwall map to those travelling furthest to the three candidate centralisation sites?
View more PenCHORD projects