Danger of glazing over ambulance performance as the NHS crisis worsens
It is now widely recognised that the NHS is facing its worst ever winter. The recent deaths of patients waiting in hospital corridors prompted the British Red Cross to speak of a “humanitarian crisis” facing hospital and ambulance services. Questions are now being asked about why this situation wasn’t addressed earlier. An investigation by the NHS Support Federation has found that over the last year important data about the performance of the ambulance service has been removed from public view. Underfunding and efficiency drives are combining to disable any comprehensive system of performance monitoring.
Removed statistical measures
Ambulance crews often wait with sick patients - in some cases for many hours - until they can be handed over to hard-pressed hospital staff. After figures published showed the highest ever number of ambulances queuing outside A&E units in 2015, NHS England decided to remove this and other key statistics from its Winter Situation Reports. In response, Dr Cliff Mann, President of the Royal College of Emergency Medicine said it seems as if “someone has decided to to glaze over the windscreen." Since then, more evidence has emerged suggesting the problem, is getting worse.
As one South East Coast Paramedic told me, awareness about the numbers of ambulances queuing is of vital importance, because these vehicles and their crews are not available to attend other emergencies. I asked NHS England the rationale for removing this important statistic and was told this was to reduce the “data burden on NHS Trusts.”
Deciding to stop the collection of important data to reduce paper work is certainly controversial. However, we discovered that ambulance trusts have in-fact continued to collect the figures. The only change is that NHS England have decided to stop publishing them. I asked NHS England how they were reducing bureaucracy by not publishing the data and, how this could be squared with the 2013 Francis Report’s recommendation that the NHS should ensure that all relevant data was made public. They declined to answer.
Mark Docherty, Director of Clinical Commissioning and Service Development at West Midlands Ambulance Service, told me that he believes the statistic should be reintroduced as a valuable measure of system pressures. He added that it is “a disgrace”, that frail and elderly people wait without privacy for hours on trolleys in hospital corridors to be treated.
Recent Freedom of Information data obtained by the Labour Party has revealed what removal of the queuing measure had obscured: Nationally, the handover delays greater than an hour rose from 28,000 in 2013-14 to 76,000 in 2015-16.
These figures back up the results of my own Freedom of Information requests in 2016. The ambulance trusts responsible for the North West, North East and in London all told me that, in March 2016, some patients waited between six and ten hours in the care of ambulance crews before being handed over to hospitals.
South East Coast ambulance service (SECAmb) told me that the number of ambulances waiting over thirty minutes increased from 2953 in April 2015 to 4167 in April 2016. According to a SECAmb report the trust “lost over 47,000 hours to hospital handover and turnaround delays in 2015/16.” SECAmb admitted that these delays presented an “Increased risk to patient safety, quality of care and dignity “ as well as “Increased risk to the wider patient community arising from the reduction and a reduced capacity to respond to new 999 emergency incidents. “
Focussed on managing demand: less monitoring and targets changed
The funding of the NHS has been tightened since 2010 to increases of under 1% a year in real terms. For the ambulance service this has meant insufficient investment in staff and vehicles. Ambulance trusts have been forced to rely upon on efficiency drives to cope with growing demand.
Under a scheme called the “Ambulance Response Programme,” changes have been made to how the speed of response to the most serious ambulance calls is recorded. This too coincides with a sharp decline in performance and a rise in missed targets.
Since 2012, the most urgent (Category-A) calls have been divided into Red-1 (most urgent) and Red-2 (serious but less time critical). Other less urgent calls fall into several Green categories. According to a national standard, ambulances should reach 75% of all Category-A emergencies within eight minutes. In recent months, this target has been missed.
According to NHS England - in November 2016, emergency response to life threatening Red 1 calls arrived within 8 minutes on 68% of occasions and 63% of the time for Red 2 calls.
In February 2015, NHS England rolled-out an Ambulance Response initiative aimed at increasing efficiency by more accurately determining what level of response that callers needed. The new scheme introduced a new set of pre-triage questions and a system called Dispatch On Disposition (DOD).
Under DOD, life threatening Red-2 calls that can include conditions such as stroke and fits, are now delayed for an additional two minutes for extra triaging.
As of October 2016, all trusts of mainland England had adopted the maximum clock-start time of DOD. This did not affect the Red-1 category but effectively severed any remaining link between past and present Red-2 performance data. This is because the changed clock-start means that data about response-time performance cannot now be compared with data recorded before the introduction of the scheme. An NHS England statistical note recognises that:
The differing clock start times mean that [Red-2] data…. are not comparable with each other. Red 2 calls comprise the vast majority of Category A calls, so 19 minute Category-A data are also not comparable….
The inability to make a like-for-like comparison of data over time hampers proper safety and performance monitoring. Proponents of the Ambulance Response Programme argue that with a lack of additional money for staffing and vehicles these changes were unavoidable.
The Clinical Coding Review: scrapping existing call categories and response-targets
A second Ambulance Response initiative called the Clinical Coding Review (CCR), has now entirely removed the established call categories for three trusts. The early version of CCR introduced a single Red category which retained an eight-minute response standard but which contained a much smaller number of calls than the previous Category-A as a whole. Red-2 calls had made up 95% of all Category-A, life-threatening emergencies. Despite this, the Red-2 category was now scrapped and these calls included within a much larger Amber category with no response-time requirement. These calls consequently went unrecorded in NHS England statistics. This is because only calls with a response- time standard are monitored in national figures.
After it was accepted by NHS England that the Amber category was too large and poorly distinguished the urgency of patients’ needs, another set of categories was introduced. Mark Docherty told me that the three ambulance trusts in the CCR pilot, now use a larger number of categories. Each of these is then further divided into a “response” and “transport” category. [NT1] The aim of the system is to more efficiently match ambulance service response to patients’ needs.
No information is currently published by NHS England on performance for these new categories. In a letter to Phillip Dunne MP in October 2016, Professor Jonathan Benger, National Clinical Director for Urgent Care NHS England explained the new CCR clinical coding set. In none of the categories listed was there an associated specific response-time standard mentioned. Mark Docherty at West Midland Ambulance Service said that this may be because CCR is still being piloted.
He told me that his trust does still observe an eight-minute response standard for the very most urgent calls. However, the number of calls with a response-standard is smaller than it would have been for Red-2 and figures for performance against this standard are not published nationally. Even if they were, because these new categories use different criteria, the data would not be comparable with the trusts’ own old Red-1 figures or the Red-1 data of the other seven trusts. This lack of comparable data has important consequences for performance monitoring.
National removal of the Red-1&2 eight-minute response
There are proposals to roll-out CCR to other trusts when the pilot has completed which would completely remove the current Red-1&2 categories and their associated response-time.
It is remains to be seen what kind of response standard is applied to any of the new categories but on the evidence to date, it is likely to apply to a smaller number of calls than for Red-2 presently.
When I spoke to Mark Docherty, he recognised the continuing importance of existing national response-time standards, stating that:
“you have got to have a benchmark. The risk is that if you stop measuring something, it slips.” He added that “National benchmarks are really helpful in flagging up when trusts need additional support.”
Safety concerns with telephone triage
Central to the Ambulance Response initiatives are telephone triaging systems. Similar systems are being touted as the solution to meeting demand throughout the NHS. Whilst NHS England emphasises that the systems used in DOD and CCR are safe, James Pavey, Clinical Operations Manager at South East Coast Ambulance service has told me that even under existing triaging arrangements, triaging-down more than ten per-cent of calls was always considered unsafe. Mark Docherty told me that “telephone triage unfortunately in a time critical environment is inherently ‘difficult.” As telephone triaging lacks the benefit of a physical examination, unsurprisingly, the numbers of people re-contacting ambulance services after discharge by telephone are consistently higher than those following face to face discharge.
In August 2016, ambulance trusts using the new Dispatch On Disposition triage system recorded the three highest re-contact rates: West Midlands ambulance Service 13.5%; North East Ambulance Service 12.6% and South Western ambulance Service 10.4%
Mr Docherty felt that response times were “still being calibrated” under CCR. Concerningly however, he told me that:
“Some of the calls triaged down are children under Five that should have automatically been classified as urgent calls because of their age.”
He told me that this group will usually go into the highest category of response precisely because “triaging over the phone for this age group is difficult.” “Unfortunately,” he said “the age of the patient was not a triage question” under the new arrangements. He said that:
“It can be two or three minutes into a call before they realise the mistake, by which time it is too late to make an eight-minute response.”
Me Docherty said that this was unavoidable because there are more clinically important questions required by the triage system before asking the age of the patient.
Trusts accused of fiddling performance figures
The failure of these efficiency schemes to make up for lack of resources is in part confirmed by the fact that some ambulance trusts have been accused of adjusting data to meet response targets. In January this year, a report in The Times stated that:
Ambulance trusts have been accused of routinely manipulating 999 response times by using a loophole to claim that they reached life-threatening emergencies in less than ten seconds. Ten of the country’s 11 trusts have taken advantage of NHS rules allowing them to log ambulance response times as near-zero if there was a defibrillator within 200 metres of a patient, and someone nearby was trained to use it. The rule applied even if the device was not used or unsuitable for the patient’s condition.
The need for comprehensive performance and safety monitoring throughout the NHS
The growing pressures on the NHS make it essential that performance is properly monitored to ensure that patient safety is not suffering. Crucially, performance of the ambulance service may be viewed as a key indicator of pressures within the wider NHS. Not only has the number of 999 calls risen by more than five per-cent for many years, but ambulance crews cope with a lack of primary and social care and wait with sick patients at overloaded A&E departments.
How the Ambulance Response programme undermines basic monitoring requirements
In 2013, the Francis report into failings at the Mid-Staffs NHS Trust recommended a greater emphasis on adequate staffing, the setting of minimum standards and the proactive publication of information “in a form that facilitates comparison with the organisation’s own previous performance and with national comparators” It also recommended data “consistency across the country.“
This post-Francis emphasis on greater transparency and enhanced performance data would now appear to be under threat. In the i-News on January 20th, Paul Gallagher wrote that in a week that half of NHS trusts were at breaking point and declaring operational alerts:
NHS England has now changed the way data is recorded so the number of major alerts – previously called red and black alerts – cannot be compared with the previous week.
What has not been widely understood is how the initiatives of the Ambulance Response Programme, contradict the basic requirements set out by Francis. The different clock-start times introduced by DOD prevent comparison with a trusts previous performance. The application of CCR to three trusts, disables both long term comparison and consistency across the country. If rolled out to other trusts it will make all past and present call categories incomparable. It is likely to remove response-targets for a large proportion of emergency calls and as a result leave these unmonitored in NHS statistics.
In 2015, the record-low ambulance performance on several measures was reflected in steeply falling graphs. The monitoring system that enabled such a graphic representation of an emergency medicine system struggling to cope has now effectively been reset and disorganised. It is impossible to accurately plot these past trends against present and future performance. Even on present measures, the ambulance service is in crisis but it will take many years of data collection to enable similar comprehensive long-term assessment.
In December, James Lazou, Unite's Research Officer for Health told me that:
"it is extremely worrying that the NHS has decided to reduce monitoring at a time when ambulance services are under so much pressure. Effective monitoring is a crucial part in ensuring that ambulance services are operating safely and perform to the highest standards. Unite is calling for the recruitment of more paramedics so that the service can be staffed with the necessary skills to deliver the most efficient and safe service."