Objectives: To determine the true impact of emergency care practitioners (ECPs) on admissions relative to emergency department (ED) attendance.
Methods: Two groups of patients ringing 999 were studied: those classified as having breathing difficulties and elderly patients (>65 years) with a fall. Routinely collected data by Yorkshire Ambulance Service were compared with a historical comparison group from the local ED. Initial contact comparison was undertaken along with statistical analysis of secondary care attendance 28 days after initial contact with the ECP service.
Results: The ECPs showed decreased rates of admission to hospital in both groups at initial contact and at 28 days (p<0.001 for those with breathing difficulties, p<0.05 for elderly patients with a fall).
Conclusions: ECPs help to prevent attendances and admissions by delivery of clinical care and assessment at point of access to health care beyond that traditionally provided by UK ambulance services. This study was limited in scope owing to the difficulties in ensuring an accurate comparison group.
Statistics from Altmetric.com
There has been reference in national documents concerning the role of emergency care practitioners (ECPs) within ambulance services and their potential impact on admission avoidance.1–4 There remains no national definition of an avoided admission and how this can be applied in practice. In Yorkshire the Primary Care Trusts (PCTs) collect data from all commissioned services of the number of admissions avoided. We currently give a figure which is based upon the ECP’s own expectation of the likelihood of the patient seen being admitted if they went to the emergency department (ED). Unfortunately this is difficult to validate on a case-by-case basis.
This paper looks at two specific clinical presentations to the ECP service in an urban population which were felt to benefit from ECP management. The aim was to establish the true impact in terms of avoided admissions using a control group from the local ED to test our hypothesis.
No ethics committee approval was required for this review.
Since January 2006, data have been collected routinely by the Yorkshire Ambulance Service using an intranet-based electronic log accessible only to the ECPs themselves and administrators on a Structured Query Language (SQL) platform. This is completed by the ECP tasking the service from the ambulance service control room. This log records basic demographic data of the patient, source of referral, the Advance Medical Priority Dispatch System (AMPDS) code where appropriate, ECP diagnosis, patient outcome, drugs used and any issues arising for performance monitoring and audit review. The job number is used as a cross reference to the Computer Aided Dispatch (CAD) system allowing the log records to be linked to demographic details.
From January to April 2006 inclusive, all the patients seen by the ECP service who had rung 999 with a diagnosis of either breathing difficulties or an elderly patient (>65 years of age) with a fall were reviewed. The data were used to review the records at Sheffield Teaching Hospitals NHS Foundation Trust which is the primary receiving unit for emergency admissions across two sites (Northern General Hospital and Royal Hallamshire Hospital) and has the only adult ED in Sheffield. Each patient seen by an ECP was searched for in the hospital computer system for attendances or admissions in the 72 h and 28 days following attendance by an ECP. Owing to the difficulty in assessing the contribution of the initial complaint to apparently unrelated admissions, all admissions were considered related for the study. In the cohort of elderly patients with a fall, all ED attendances were considered potentially relevant owing to the multiple causes of falls in this group. This is likely potentially to underestimate the true benefit of ECP intervention.
Comparison data were taken from January to April 2005 inclusive for attendances to the same ED for patients with the same criteria as above seen by non-ECP ambulance service personnel. These dates were chosen because, during this period, the ECP service was not tasked to patients with breathing difficulties and Yorkshire Ambulance Service had only 12 operational ECPs during this comparison period compared with 24 whole-time equivalent operational ECPs during the study period. The outcome of patients was then reviewed to establish the proportion of patients who were admitted or discharged.
No other changes in the pathways for these patient groups were commenced during the period from January 2005 to April 2006. These dates were chosen because of expected high numbers of patients with breathing difficulties presenting to medical services during the winter period.
Comparison for statistical significance was made using a χ2 two-sample test with the level of significance set at p<0.05.
Patients with breathing difficulties
From January to April 2006 (inclusive) the ECP service saw 186 patients who rang 999 with a problem classified by AMPDS as difficulty breathing (fig 1). Of these, 119 were treated at home and referred on for primary care review where appropriate, a 64% initial avoided attendance rate. From January to April 2005 (inclusive) 1130 patients attended the local ED via 999 ambulance with a diagnosis ascribed as difficulty breathing. Of these, 855 were admitted with 266 discharged directly or to other services, a 24% avoided admission rate (table 1). If we assume the same discharge rate as from the ED in 2005 then, of the 186 patients seen by an ECP, only 45 would have been discharged (vs 119).
The patients who were seen by an ECP were then checked against ED data at the Sheffield Teaching Hospitals for attendance or admission within 72 h or 28 days from initial service contact. Of the 119 patients who were managed at home, 18 attended the ED of which 15 were admitted within 72 h of the ECP attending. Two of the attendances were unrelated to respiratory problems. At 28 days a further 19 patients had attended the ED of which 13 had been admitted. Three of the attendances were unrelated to the presenting problem. Two patients lived outside the study area and were lost to follow-up (table 1).
When those admitted or attending with related problems within 28 days are taken into account, the 28 day avoided admission rate is 46.7%. This suggests that ECPs reduce admissions by 30% at 28 days relative to the ED “initial contact” figure (from 76% to 53.3%, n = 1307, df = 1, p<0.001).
ECP diagnoses are recorded using a 100 diagnosis pick list formulated locally. The diagnosis proportions for this group are shown in table 2.
Elderly patients with a fall
From January to April 2006 (inclusive) the ECP service saw 233 patients who rang 999 having fallen and who were over 65 years of age (fig 2). Of these, 171 were treated at home and referred on for primary care review where appropriate, a 73% avoided attendance rate. From January 2005 to April 2005 (inclusive) 772 patients aged >65 years attended the local ED via 999 ambulance with a diagnosis of a fall. Of these, 396 were admitted with 369 discharged directly or to other services, a 48% avoided admission rate. If we assume the same discharge rate as the ED then, of the 233 seen by an ECP, 112 would have been discharged (vs 171). Of the 171 patients who were kept at home at initial contact, 21 attended the ED or were admitted within 72 h of the ECP attending. At 28 days a further 19 patients had attended the ED or been admitted.
When those admitted or attending within 28 days are taken into account, the 28 day avoided admission rate is 56%. This suggests that ECPs reduce admissions by 17% at 28 days relative to the ED “initial contact” figure (from 52% to 44%, n = 1005, df = 1, p<0.05).
The diagnosis proportions for this group are shown in table 3.
Avoiding inappropriate hospital admissions is a high priority for PCTs in the UK, but difficulty exists in defining when an admission has been avoided. It is extremely difficult to predict from a single clinical encounter whether the patient seen would have been admitted if he/she had presented to secondary care. Most admission avoidance schemes are required to present a figure to support their continued working practices; however, these figures are highly subjective and unlikely to reflect the true picture without accurate follow-up over a 28 day period.
This study aims to give a more accurate reflection of the impact of ECPs on the management of two specific conditions. A historical comparison group from Sheffield Teaching Hospitals was used to try to decrease bias. Assuming that all subsequent attendances or admissions by patients in the group of elderly patients with falls may have been related to the initial presentation is likely to lead to an underestimation of the true benefit of the ECP intervention. The dates chosen should minimise bias caused by the ECPs themselves (as discussed in the Methods section). Analysis was facilitated by the service having only a single acute trust in its operational area.
The ECP service in Yorkshire accepts referrals from ambulance crews, direct from care homes and other sources, as well as responding directly to 999 calls. This risks selection bias. To minimise this, cases other than direct 999 calls were discounted to endeavour to get as close a match as possible for comparison with the ED. The ECPs in the Yorkshire Ambulance Service are all paramedics or nurses with more than 3 years experience who underwent a 1-year full time course at a local higher educational institution to become ECPs. Although at the time of training there was no nationally agreed curriculum framework, their skill set fits with that of the recent competence and curriculum framework from Skills for Health.5 The ECPs do not have any specific criteria for admission or discharge other than those available to all medical staff such as the National Institute for Clinical Excellence (NICE) guidelines for chronic obstructive pulmonary disease (COPD),6 and are expected to use their clinical judgement in decision making. Some additional training in COPD was added to the initial training as a short module delivered jointly by the ambulance service and the local higher educational institution.
Most PCT-based initiatives centre around proactive management of patients with chronic disease supporting patients to prevent multiple readmissions, although recent research has questioned the impact of this approach.7 Calls to ambulance services are increasing annually and admissions from EDs are also increasing. A combination of both reactive and proactive management of patients is necessary to prevent admissions, and ECPs can provide part of the reactive element.
There are multiple access points for the general public to health care: their GP or out-of-hours provider; the ED or Minor Injuries Unit; contacting NHS Direct; or ringing 999. Currently the 999 pathway is the only one in which, in the majority of cases, neither a doctor nor advanced skills nurse practitioner assesses the most suitable care delivery pathway for the patient at point of contact. This occurs when the patient arrives at the ED. One of the problems with assessing 999 calls is that most of the research which has labelled calls as inappropriate has been retrospective and decisions made on clinical appropriateness after medical assessment has taken place. It is unreasonable to expect a layperson to make an informed clinical judgement about their health needs when studies suggest most callers feel their 999 call to be appropriate.8
The study of alternative emergency care pathways for patients is relatively new. In 1998 the Audit Commission questioned the need for a fully crewed ambulance to attend all 999 calls,9 but a study in 2000 looked at all UK ambulance services and found that only three were carrying out specific work on evaluating initiatives which mainly concerned alternative care pathways for category C (AMPDS classified non-urgent) patients.10 A case-control study in 2003 looked at using computer assisted assessment and advice for “non-serious” 999 calls. Category C calls, deemed as non-serious, were identified and a computer aided decision making tool was applied. Of 635 cases in the study, 330 did not require an emergency ambulance and 119 of these did not attend the ED, 155 required care within 24 h, 85 had a routine appointment with the GP and 89 for whom self-care would suffice. Worryingly, 161 of the 635 cases prioritised by the computer system as non-serious (25%) required immediate care.11
The same study commented on the conflict between developing alternative pathways, which could delay ambulance responses by increased on-scene times, and the need to meet response targets.12 One of the major difficulties with the type of patients discussed in this paper is that most will present as either category A or B which necessitates a performance-related response time linked to ambulance service ratings. This has been criticised by consumer groups as stifling innovation.13
ECPs are now becoming established across the UK working in a variety of unscheduled care settings. The role of the ECP is designed to occupy a space between GP, nurse and paramedic, and training includes social care and other NHS services to provide a more holistic approach to patient care.14 A recent report from the Department of Health has highlighted the positive aspects of this role.3
Studies reviewing “treat and release” protocols by paramedics have shown an increased risk of inappropriate decision making.15 It is likely that the extended role and training received by the ECPs will decrease the risk. In cases such as difficulty in breathing it would be unreasonable to expect paramedics without extended training to make a clinical decision about the suitability of patients to stay at home. One of the difficulties for any medical practitioner is that patients can deteriorate—as shown in the 72 h admission figures in this study. Much of the challenge lies in balancing patient safety against a desire of commissioners to avoid hospital attendances and admissions. In the group with breathing difficulties, 13 of the 15 admissions at 72 h were due to patient deterioration and the other 2 were due to unrelated problems. This would suggest that the misdiagnosis rate was low and that, with appropriate patient advice on seeking follow-up if required, safety is relatively high. All of these patients reattended the ED via an emergency ambulance. Owing to the multiple pathologies which can lead to falls, it is very difficult to say whether or not the 72 h admissions in this group were due to misdiagnosis, but the majority were due to inability to cope despite increased support with some further falls or confusion.
The figures presented in this paper suggest that savings may be made by admission avoidance using ECPs at this level and by tackling these categories of higher risk patients. Payment by results sets the figures by which such schemes should be judged. An ED attendance will cost between £54 and £99, an elderly patient with a breathing problem admitted as an unscheduled case will cost up to £3176.16
Patients presenting with breathing difficulties have an average tariff of around £2500 per case, with an average trim point (proxy for length of stay) of 30 days. Preventing these admissions by appropriate referral to a community pathway could give substantial savings. Elderly patients at risk of falls are represented on a wide range of Payment by Results Health Resource Group (HRG) codes, and a simple average of those codes with “elderly” in the description gives a cost of around £4000 per case and an average trim point of 59 days.
Most category C calls attract only an ED attendance tariff. Our figures demonstrate the potential admission savings from higher risk cases with only one respiratory admission prevented saving the same as 25 high tariff A&E attendances. The challenge remains for both commissioners and providers of such services to accept that targeting high-risk patients in this way will cause some decrease in the number of avoided ED attendances as these patients are at greater risk of admission, but that the cost savings are potentially higher.
A further barrier to development exists in the current NHS position of financial difficulty. There needs to be a significant initial investment to train the ECP workforce and equip them appropriately for their role. This has been estimated at around £25 000 per ambulance service-based ECP with the position of positive cash flow not reached for up to 4 years.2 To this figure also needs to be added the backfill costs for operational staff to maintain normal service provision. A scheme will consequently be heavy in cost in its initial stages and no return will be seen during the time required to train, which is currently around 12 months. For many organisations who have other priorities—both financial and operational (such as call connect)—such development is likely to remain a challenge requiring a 5-year plan to realise the full potential of ECPs.
Emergency care practitioners can help avoid attendances and admissions to EDs for patients requesting a 999 response who are classified as either having difficulty in breathing or a fall in the elderly. Potential cost savings to the NHS are significant when reviewed against the tariff set by the Department of Health for payment by results.
The advanced practitioner who can assess and treat the patient at point of access to health care is increasingly important, with the ECP at the front end of the ambulance service’s response to its changing role. Further development is required in the way calls are assessed and performance indicators designed to allow this work to function most effectively and with the highest levels of patient care.
The authors thank Sue Cross, Sheffield Teaching Hospitals, for data collection; Hugo Minney, South Yorkshire Academy, for data review; and Yorkshire Ambulance Service Emergency Care Practitioners.
Competing interests: None declared.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.