NHS 111 was set up as a helpline for urgent medical concerns and receives a growing number of calls – over 15 million every year. As we highlighted in an earlier analysis, a frequent criticism of NHS 111 is that it directs a relatively high proportion of patients to A&E, and there have been concerns that this might have added to the pressures on A&E departments, which can be particularly high during winter.

In response to these concerns, NHS England has increased the number of calls receiving clinical assessment by a nurse or a doctor. In January 2018, they reported that 39.5% of calls had been handled in this way, up from 23% in January 2017.

What impact does clinical input into NHS 111 have on A&E departments?

It is still unclear what impact clinical input to NHS 111 calls has on A&E departments and very few research studies have examined this (a gap in the evidence base that was noted by the Primary Care Workforce Commission in 2015).

The Health Foundation’s data analytics team aims to shed light on just this kind of issue. Working in collaboration with researchers from the Royal College of Paediatrics and Child Health, we have analysed data on NHS 111 calls about children and young people in Hammersmith and Fulham, Kensington and Chelsea and Westminster. Using these data, we were able to examine whether those children and young people who were reviewed by a GP during or shortly after their call to NHS 111 were subsequently less likely to attend A&E.

The research is still work in progress. We have shared the results at national conferences, and this month published them in provisional form on a pre-print server prior to peer-review. We’re keen to share our findings at this stage to generate discussion about both their potential implications and the methods used.

The project involves linking together existing NHS data from over 10,000 patient episodes from multiple sources. We began with data from the operational systems used by NHS 111 call handlers in these three areas of London, from which we obtained records of out-of-hours calls to NHS 111 for people aged 15 years or younger between July 2013 and February 2015.

By linking these data to a database of appointments with out-of-hours GPs, we identified 7,458 calls who received an appointment with a GP, either as part of the original call to NHS 111 or very shortly afterwards. We compared these patients with a group of 2,898 callers who were not reviewed by a GP, but advised by NHS 111 that their health conditions could be managed at home.

We found that 16.2% of the GP group attended A&E within ten hours of the NHS 111 call, compared with 14.9% of the stay-at-home group. Of course, the patients who were reviewed by a GP also tended to have higher levels of health need – for example, they were more likely to have breathing difficulties, febrile illnesses and diarrhoea or vomiting than patients who were advised to manage their health needs at home.  We therefore risk adjusted for differences in the characteristics of the two groups.

After this adjustment, we found that the GP group was around 14% less likely to attend A&E departments than the group who was advised to manage their health needs at home. Thus, clinical review by a GP was associated with fewer A&E attendances. In another analysis, we looked at the impact of being reviewed by a nurse rather than GP and found the same pattern.

What do the data really tell us?

On the surface of it, these findings appear to corroborate the drive towards increasing the frequency of clinical input into NHS 111 calls. However, the story seems to be more complicated when we drill down further into the data.

When breaking the analysis down by type of A&E, we found that the patients who were reviewed by a GP were around 68% less likely to attend minor injury units than those who were advised to manage their health needs at home, but there was no sign of an impact on visits to major A&Es, which are more expensive to the NHS.

One explanation of the findings is that there may be a group of parents who are in need of reassurance about their child’s wellbeing, who might normally consider taking their child to a minor injury unit, but probably not to a major A&E department. When offered the option to speak with a GP on the telephone, many of these parents opt not to attend the minor injury unit, explaining the results seen. However, there is another group of parents who are very concerned about their child’s wellbeing (and perhaps rightly so) and will go to a major A&E department regardless of whether their child is reviewed by a GP.

Looked at in this way, the findings suggest that clinical input during (or soon after) calls to NHS 111 might provide valuable reassurance to callers, but not avert pressures on major A&E departments.

When is clinical input most effective in the urgent and emergency care pathway?

The findings raise questions around how clinical input can be most effectively targeted within the urgent and emergency care pathway. Primary care is under pressure as well as hospital care, and GP and nurse time is valuable and increasingly scarce. If clinical expertise is more valuable for some patients and at some points in the pathway than others, then it makes sense to try to design health systems that target it optimally.

Further analysis would help. For example, to examine the views of the people calling NHS 111 (these could use surveys developed previously) as well as rates of A&E attendances, since a high-quality health care service is about more than managing need. There is also a need to repeat our analysis on a larger sample, for example including other age groups and localities, where the patterns of service utilisation might be different.

With health services under pressure, it is more important than ever to understand the factors that are leading people to attend A&E, so that approaches to improving care can be designed carefully. As this piece of analysis shows, the answers will not always be straightforward, but that is only to be expected in an area like urgent and emergency care where patients move around the system in such complex ways.

Adam Steventon is Director of Data Analytics at the Health Foundation.