A new era for A&E targets: what will be the impact of the new basket of measures?

Long read: It was announced at the end of June that the four-hour A&E target will be replaced by a new basket of measures. But what might be the benefits and disadvantages of those new measures? Jessica Morris and Chris Sherlaw-Johnson run through the proposed A&E standards to see what they might mean for patients.

Blog post

Published: 21/08/2020

It is the end of the road for the four-hour A&E target. At the end of June, the National Medical Director confirmed that the target for 95% of patients attending A&E to be admitted, transferred or discharged within four hours will be replaced by a new basket of measures.

Citing the Covid-19 pandemic and the initial findings of the clinically led review of NHS access standards, NHS England and NHS Improvement reinforced the need for a wider set of A&E measures that will improve information, reduce crowding and drive quality improvements.

The new A&E measures will form just part of that review, which also focuses on elective care, cancer care and mental health care. The final recommendations will soon be published in a report – albeit without threshold targets for each metric, which the recent board report states will require further analysis and modelling.

With the transformation of A&E targets fast approaching, how do the new proposed measures compare to the current four-hour target? What may be the benefits and disadvantages of the new measures? Using Hospital Episode Statistics, here we unpick some of the technicalities to see what this all might mean for patients.

What are the proposed A&E standards?

Time to initial clinical assessment

A recent poll by Healthwatch found that assessing patients quickly on arrival at A&E and prioritising those with the highest level of need are both paramount to patients. This was mirrored in the interim report that also mentioned the need to identify life-threatening conditions faster, and for the measure to be understandable for patients.

Some may remember that the time to initial assessment was previously abandoned as a standard in 1997. While the original rationale for the metric has stayed the same, there remains a concern that the assessment must be clinically meaningful, and not so brief that it does not benefit the patient. It will therefore need to be tightly specified, by requiring for example a brief history, pain and early warning score as part of the assessment.

The time to initial clinical assessment is already being reported in the A&E quality indicators, but only for patients who are brought in by ambulance. It is not actively performance managed, and there are some data quality issues, which are likely to improve if the metric receives more focus as part of the basket of measures. To improve public understanding, it should be communicated that the clock starts when a patient is registered at reception, not when they walk through the door.

Time to emergency treatment for critically ill patients

The four-hour target groups all patients together, whether they went to resus, major or minor areas. The interim report suggested that for patients with life-threatening conditions – including stroke, heart attack, major trauma and others – they should complete a package of treatment in the first hour after arrival.

The time to treatment for all patients is currently being submitted by emergency departments, but the data quality is variable, and it is not being monitored closely at present. Specifying the highest priority patients and ensuring that they begin treatment in a timely manner is certainly of benefit, although it will introduce a set of challenges.

Hospitals may have to adapt their systems to enable time-to-treatments to be collected effectively for different conditions. There is prior experience of this with the National Clinical Audit Programme, which also highlighted the need to avoid gaming of targets. Ensuring that the treatment is high quality and clinically appropriate will be of prime importance, and the exclusion of certain conditions should be carefully considered.

Mean time in A&E

One of the key measures that has been proposed is the mean waiting time for all patients. The below chart shows that four-hour target performance and the mean time patients wait in A&E are closely correlated. Trusts that perform better against the four-hour target have shorter mean waits, and trusts that perform worse have longer mean waits.

There is large variation between trusts. In 2018/19, two A&E departments had a mean wait of over five hours, while one trust had a mean wait of under two hours.

Mean time in A&E compared to percentage waiting less than four hours 20/08/2020

Chart

Source:  

Nuffield Trust analysis of Hospital Episode Statistics.

Read more

The fact that the measures are closely linked is reassuring, as the introduction of the four-hour target in 2004 did dramatically improve A&E waiting times, despite performance falling over time. Knowing there is a similar measure in its place should plug the gap, but it is problematic that there will be no target threshold at first. Without a mean waiting time target to work towards, A&E waiting times are likely to increase, assuming other contextual factors remain the same.

What is less clear is how a mean wait target will change the distribution of waits at each trust. By removing the four-hour cliff edge, there could be an incentive to treat or admit the longest waiters, as this would mathematically lower the mean to a greater extent. This raises questions about which patients will benefit or lose out from the new system.

Further A&E measures

Patient perception of A&E waits

While using hospital data to monitor A&E waiting times is useful, we must not lose sight of why it is important: to improve patient experience and outcomes. But how closely does the data on mean waits correlate with patient experience of how long they feel they spent in A&E?

The urgent and emergency care survey asks respondents: “Overall, how long did your visit to A&E last?” The results show that there is a relationship between the hospital data and survey data, although there are some outlier trusts that do not follow the trend.

Patient perception of A&E waits compared to mean time in A&E 20/08/2020

Chart

Source:  

Nuffield Trust analysis of Hospital Episode Statistics.

Read more

The fact that these measures are connected is important, but it is worth remembering that other factors influence patient experience too such as effective triage, clear communication, access to pain relief and quality of care. These factors are not so well captured when looking at waiting times data alone.

Although the mean wait may be a valuable measure for clinicians and managers, how useful will this metric be for an individual patient? The four-hour target gave patients a maximum wait to expect, but the mean wait is far less intuitive. As mentioned above, mean waits can be strongly influenced by the management of those waiting a very long time, so the average wait may not reflect the experience of most patients. This could potentially lead to uncertainty and confusion for the public, who might rather value an estimated waiting time on arrival.

12-hour A&E waits

The clinical review of standards progress report mentions the proportion of patients spending 12 hours or more in A&E from the point of arrival as a measure that would ‘shine a light’ on those who wait a long time. It would replace the current 12-hour trolley wait measure, where the clock starts only when a decision is made to admit someone and ends when they are admitted to hospital.

The next chart shows that there is large variation in 12-hour waits between trusts, signalling that introducing this measure may benefit patients by encouraging improvement.

For 55 trusts, the proportion waiting more than 12 hours is low at less than 1%, but 12 trusts have a proportion of over 5%, representing a lot of people waiting for an extended period. Although the total time spent in A&E may be more important by comparison, trolley waits remain significant, and there is a risk that they could creep up if monitoring ceases.

Percentage of 12 hour or more A&E waits in different hospital trusts 20/08/2020

Chart

Source:  

Nuffield Trust analysis of Hospital Episode Statistics.

Read more

Measuring admitted and non-admitted patients separately

NHS England and NHS Improvement have indicated that they are exploring the value of monitoring admitted and non-admitted patients separately. Patients who require admission to hospital have to wait for a bed to become available, and are often seriously ill, which means that they may require more treatments and tests while in the emergency department. It does seem reasonable therefore to measure waiting times for these patients separately.

The following chart illustrates that admitted patients wait longer on average than non-admitted patients in all trusts. In 2018/19, the extra length of time that they waited varied from 18 minutes to over five hours.

Mean time in A&E compared to percentage waiting less than four hours for admitted and non-admitted attendances 20/08/2020

Chart

Source:  

Nuffield Trust analysis of Hospital Episode Statistics.

Read more

This extra time waited by admitted patients is affected by a multitude of factors – hospital bed occupancy, staff capacity, use of same-day emergency care, and many more.

One might expect that trusts with a greater proportion of patients admitted to hospital from A&E (the “conversion rate”) would also have a higher mean waiting time. But in fact there is no correlation between conversion rates and the mean time spent in A&E. This indicates that there are other factors besides admission to hospital that influence waiting times.

Percentage of patients admitted compared to mean waiting time for non-admitted patients 20/08/2020

Chart

Source:  

Nuffield Trust analysis of Hospital Episode Statistics.

Read more

The whole basket

A&E waiting time standards are being reformed, but ultimately it is the impact on patients that matters most. It is therefore imperative that the new measures incentivise the right behaviours and that any trade-offs do not jeopardise patient safety. The NHS aims to give patients the right care in the right place at the right time, and this core value must be upheld.

It is notable that in the sites trialling the new A&E measures, the proportion admitted to hospital fell while the mean waiting time for non-admitted patients went up. This is a positive finding, as it indicates that more patients are getting the treatment they need and are able to go home on the same day. Despite spending slightly longer in A&E, it may be in the patient’s interest to be spared an overnight stay in hospital. It does, however, bring home the tricky issues involved in developing measures that are meaningful to the public, not distorted by political considerations, and encourage better care.

The fact that four-hour waits correlate with mean waiting time is helpful as it shows that the current target is being replaced by a similar performance measure. However, from a patient perspective, the usefulness of an average waiting time is questionable. Performance against the four-hour target should continue to be published as we transition to the new metrics.

This would provide a useful comparison over time, enabling us to better understand the impact of the changes.

The new basket of measures will not be immune from other developments, such as the introduction of a “call first” approach that may affect waiting times in unforeseen ways. A&E departments have become a part of the NHS that is always open. Diverting people to 111 as a first point of call may offer an alternative outcome, with an increase in telephone advice and supported self-management.

One key unanswered question is the promise that politicians can make to the public. Will they be able to guarantee that patients will be assessed in A&E more quickly? Or that the mean waiting time will reduce? Without threshold targets for each metric, there is no clarity for the public about how to judge A&E services, or what to expect as a patient.

The absence of thresholds will also hamper performance improvement. Most of the measures in the basket have been trialled before, and many are already published as part of the A&E quality indicators. Without thresholds, and a clear message from national leaders that improving performance matters, it is unlikely that the measures will drive improvement. If the four-hour target is dropped as a requirement and the other measures do not have a standard against them, we can expect to see increasing variation in waiting times across the country and an overall deterioration in performance. 

The basket has the benefit that different measures will lead to incentives for different groups of patients to be prioritised, reducing the risks of clinical priorities being distorted to meet a target. However, whether this benefit will be realised will depend on whether there is a “first among equals” measure in the basket. If the mean time in A&E becomes the de facto main measure, then we may find there is very little change from the current target regime.

Given the current Covid-19 crisis, prioritising other factors including infection control and creating hot and cold areas may be more pertinent. But time is of the essence and both measures and expected performance standards are needed if we are to maintain the quality of A&E care.

*Hospital Episode Statistics data (year range 2018/04–2019/03) Copyright © (2020), NHS Digtital. Re-used with the permission of NHS Digital. All rights reserved.

Suggested citation

Morris J and Sherlaw-Johnson C (2020) A new era for A&E targets: what will be the impact of the new basket of measures?”, Nuffield Trust comment.

Comments