Emergency Medical Systems or ambulance services vary internationally. The findings here are relevant to England. Policy-makers, service providers and researchers in other countries will have to consider similarities
and differences between their system and the system researched here when considering the transferability of findings.
Our team comprised researchers with many years of experience in emergency and urgent care research, patients, paramedics and members of a body for ambulance service leaders. We have tried to take a respectful approach to services working under challenging circumstances while also identifying areas with the potential for improvement.
Patient and public involvement
As well as having PPI members as part of the project team and Project Advisory Group, patients and members of the public working with three ambulance services or research teams engaged with emergency and urgent care came together to form a VAN project PPI group. Although members associated with one of the ambulance services had to drop out over time, the two established PPI groups worked together successfully for the duration of this project. The PPI members helped to interpret the findings of the study and develop a strategy for disseminating the findings to a non-academic audience. The group met on five occasions and were also present at the final project management meeting at which the findings from each WP were presented. The group met for a final meeting after the submission of this report to finalise the plans for PPI-led dissemination.
During the PPI meetings, members of the research team presented emerging findings from each WP. PPI
members’ discussions affected the research in numerous ways, including the following:
1. PPI members were interested in variation in total non-conveyance rates as well as the different types of non-conveyance. We included graphs of total non-conveyance in our presentations to show that variation existed at this level as well as at the level of different types of non-conveyance.
2. The appropriateness of non-conveyance was very important to the PPI group. This was a key component of our study.
3. Our study focused on NHS staff views and analysis of routine data. PPI members felt that patients’
perceptions were missing from our study. They understood that we had not been funded to include
patients’ perceptions but wanted us to place our findings alongside other research focusing on patients’
views. We included a review of literature of patients’ views in the introduction to the study. We
highlight the need to focus on patients’ views of non-conveyance in our recommendations for
future research.
4. PPI members were keen to promote patient education about how best to use the ambulance service.
Conclusions
Variation in non-conveyance rates between ambulance services in England could be reduced by addressing variation in the availability of advanced paramedics and how they are used within services and in the perceptions of the risks associated with non-conveyance within ambulance service management.
Implications
Non-conveyance occurs in a context in which ambulance services are judged largely by their ability to meet response-time targets as demand for their service increases. Non-conveyance is a relatively new activity in the history of ambulance provision, which requires ambulance services to shift their focus from transporting emergency cases to hospital to managing emergency and urgent care cases in a way that is much more aligned to providing care appropriate to clinical need. Many processes vary between ambulance services, which may contribute to unwarranted variation in non-conveyance rates. These processes include the
standardisation of CAD systems, triage software, skill-mix labels, how different grades of paramedics are used in practice, and how ambulance services are commissioned in practice. We have identified that workforce configuration, how advanced paramedics are used and management motivation to undertake non-conveyance were key issues that ambulance services can address to reduce variation in non-conveyance rates between ambulance services. There are a number of implications in light of this.
Formalising definitions of non-conveyance
When policy-makers and researchers assess non-conveyance rates for ambulance services they may wrongly interpret these rates, especially for calls ending in telephone advice. It is easy to interpret published AQIs as rates of patients offered advice over the telephone by clinicians when in practice this includes an array of reasons for the call ending without an ambulance being dispatched. The evidence suggests that policy-makers might benefit from considering what they wish to see measured and putting in place plans to produce AQIs that can measure this. There remains concern about the validity of the data used to calculate AQIs, especially for calls ending in telephone advice. Real variation between services, and especially variation in calls receiving clinical advice that end in telephone advice, could easily be larger or smaller than rates reported in AQIs. A first step to better understand the important aspect of clinical telephone advice by ambulance services is to investigate the development of new indicators that separate
out what might be called‘true hear and treat’ from calls dealt with by call-handlers or calls that should
simply never be classified as ending in telephone advice, such as hoax calls. Improving the accuracy of the AQIs, and monitoring adherence to the guidance for calculating AQIs, would help ambulance services to consider real differences between themselves and other services.
Reducing variation so that all ambulance services offer the optimum rate of non-conveyance
It is possible that the ambulance service with the highest rate of non-conveyance at the time of the research did not have the optimal rate of non-conveyance. Establishing what the optimal rate is requires modelling of the cost-effectiveness, and safety and appropriateness, of different rates of non-conveyance. Some approaches to non-conveyance may not be cost-effective, may result in increased mortality or increased severity of illness, or may simply delay conveyance. This study did not set out to establish the ‘right’ rate of non-conveyance and cannot shed light on this issue. A further research study is required to address this, which repeats the linked data analysis undertaken here for one service only for all ambulance services in England and considers the relationship between mortality and subsequent service use for different rates of non-conveyance. A future study would also be necessary to identify the cost-effectiveness of different approaches to non-conveyance.
How to reduce variation
The variation in overall rates of non-conveyance mainly concerned the small number of ambulance services with extreme rates, namely one service with a relatively high rate of discharge at scene and some with relatively low rates of discharge at scene. There was no indication that the high rate was problematic
(although safety was not measured for this service and a limited measure of appropriateness was considered). There was evidence of opportunities to increase the discharge-at-scene rate for ambulance services with lower rates. Some of this unwarranted variation lay within the control of the ambulance service and was determined by their workforce policies and motivation to undertake discharge at scene. National policies cannot be the cause of variation because some ambulance services achieved high non-conveyance rates with these policies
in operation. However, a change in policies– in particular response-time targets – may help services with low
rates to prioritise investment in non-conveyance. A key issue is the considerable amount of variation in ambulance processes in terms of CAD systems, triage software for dispatch, triage software by telephone or face-to-face clinical input, paramedic nomenclature, and how staff with different skills are used. Much of this variation is not necessary and standardisation of these processes could contribute to reductions in the variation in outcomes such as non-conveyance.