Five ways linking ambulance and A&E data could help to improve care

My paramedic colleagues and I regularly provide life-saving care, but as soon as we hand over to the emergency department (ED), our knowledge about that patient stops. It’s like watching the first half of a play for the curtains to never open on the second act.

Ambulance services in England have a very limited understanding of what happens to the patients they treat. In most circumstances, even basic information such as admission and mortality rates are not known. It can be frustrating as an individual, not knowing whether the care you provided resulted in a good outcome, but there are also bigger opportunities being missed at a service, research and commissioning level. This is because there is currently no routine link between ambulance data and data from all other NHS organisations.

However, ambulance data is increasingly stored electronically, providing an opportunity to link this otherwise isolated dataset across organisational boundaries. Over the last few years, I’ve been running a project to do just this. The Pre-Hospital Emergency Department Data Linkage Project (PHED Data), funded by the Health Foundation as part of their Insight 2014 award, explored the feasibility of linking these datasets and any potential benefits.

We requested demographic, diagnostic and outcome data from 13 EDs for all patients taken there by ambulance. All EDs agreed to participate, signing information sharing agreements and transferring non-identifiable data.

Despite a rather lengthy process (it took on average 1 year and 13 weeks per trust) we showed that it is possible to link ED and ambulance data. The result is a dataset of just over 750,000 linked records over four years, tracking a patient record from the 999 call to the outcome for the patient in the ED and everything in between.

So what does this mean? Here are a few of the ways we think this linked data could help to improve care and outcomes for patients:

1. It could help to improve the accuracy of triage during 999 calls

On the phone, patients are triaged into categories using a traffic light system. In recent years the number of red (the most urgent) calls has increased beyond forecasted predictions and we think that our current system of triaging patients could be improved. At the moment, the system is based heavily on the patient’s symptoms (such as chest pain or difficulty in breathing), however should we also be considering other things, such as the age of the patient? This data could provide a way of safely reducing the number of red calls without increasing the risk, whilst also finding trends in green (the least urgent) calls that could benefit from being more highly prioritised. Our results support the findings of the Ambulance Response Programme, a national initiative looking at telephone triage.

2. It could help us plan services better, showing how health care professionals (such as GPs) use the emergency pathway

Health care professionals (HCP) frequently refer patients to the ambulance service and from our data we know that those referrals have a high admission rate. HCP calls follow a pattern that peaks at midday on weekdays, when they account for up to a quarter of all patients taken to hospital. Understanding this pattern helps the ambulance service and the ED to work together, planning for this predictable influx of patients.

3. It could help to ensure acutely unwell patients are recognised and treated quickly

We looked at the records of every patient who died in the ED to assess the care given by the ambulance service. Patients tended to be older and were more likely to reside in a care home indicating they may be receiving palliative care. In general, the ambulance service is already recognising and escalating care for most of these patients, however there were a minority who were not recognised. Analysis of our data suggested that respiratory distress, over other diagnoses, could be an early predictor of seriously unwell patients presenting to the ambulance service.

4. It could help us consider alternative places of care (with caution!)

The data can help us look at patient groups who on the face of it did not benefit from an ambulance trip to ED, for instance patients who were discharged with no further action required. It’s possible this could lead to community initiatives to treat these patients without an ED visit. While it’s dangerous to jump to simplistic conclusions without considering the whole picture (being discharged from the ED doesn’t necessarily mean that a patient should have avoided a visit altogether and this data will need further detailed analysis), we found there may be scope to provide training that could empower paramedics in some circumstances to safely assess patients in the community.

5. It could improve how we commission services

Despite some limitations, such as data quality and misinterpretation, linked data was considered a suitable evidence base for informing a variety of future commissioning initiatives. Commissioners have noted that the work fits with current work on the Ambulance Response Programme, the Emergency Care Dataset, and could inform decisions made by NHS England and NHS Digital about national data sets. At a local level the data could be used to understand demand for services and track the impact of a community pathway on ED attendance rates. Implications for commissioning across the wider health care system, particularly public health and preventive health interventions (such as the impact of deprivation on ambulance/ED use), were also identified along with risk algorithms.  Resource modelling and the use of simulations using this data could also help patient flow with the aim to reduce ambulance waiting times at the ED.

This is just the start. We’re only starting to scratch the surface of what can be achieved with this rich and varied dataset.

Our work looks at linked data in the context of ambulance services, however it is an example of the many ways that data can be used to find new methods to assess quality, efficiency and to contribute to service planning. Establishing relationships between organisations and using this data to implement and track changes is the next step to improving health care delivery.

Sophie Clark is a registered paramedic, and Principal Investigator of PHED Data at the London Ambulance Service NHS Trust. For more details of this work, or to receive a copy of the report when published, please contact sophie.clark@lond-amb.nhs.uk.