You are here

Article Text

Article menu

Download PDFPDF

Original article
How do clinicians with different training backgrounds manage walk-in patients in the ED setting?
  1. Tim Harris1,
  2. Keith McDonald2
  1. 1Emergency Department, The Royal London Hospital, London, UK
  2. 2GP Health E1 Homeless Medical Centre and Clinical Lead GP Streaming Service, Barts Health NHS Trust, London, UK
  1. Correspondence to Professor Tim Harris, Emergency Department, The Royal London Hospital, Whitechapel Road, London E1 2BB, UK; tim.harris{at}bartsandthelondon.nhs.uk

Abstract

Objective To compare the initial assessment and management of walk-in emergency department (ED) patients between different types of healthcare providers.

Setting A large teaching hospital with an annual ED census of 140 000 adult patients.

Methods A random sample of 384 patients who self-presented to the ED was obtained. A detailed analysis of each patient record was performed by two clinicians. Data were obtained on the presenting condition, and disposition of each patient, either into the ED for further assessment, or discharge.

Results GPs were significantly more likely to discharge patients home as compared to emergency nurses. ED senior nurses were more likely than GPs to stream patients into the ED for further assessment. Of the patients referred into the ED for further assessment by senior ED nurses, the majority were discharged home. There were insufficient numbers of emergency physician assessments for meaningful statistical analysis.

Conclusions The clinician groups studied here demonstrated different patterns of discharge and referral, reflecting their training and experience. When planning operational procedures, the training and background of the staff allocated to each area should be considered.

  • clinical assessment
  • emergency departments
  • emergency care systems, admission avoidance
  • emergency care systems, efficiency
  • emergency care systems, emergency departments

https://doi.org/10.1136/emermed-2013-202844

Statistics from Altmetric.com

    Request Permissions

    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.

    Key messages

    • What is already known on this subject

    • Walk-in patients are frequent emergency department (ED) attendees, and the vast majority of these patients are discharged home. There is no international consensus on which walk-in patients should be cared for in the ED and which should be referred to a primary healthcare provider. There is little objective data to compare the clinical behaviour of different healthcare providers in the assessment of this cohort of patients when attending the ED.

    What this study adds

    • General practitioners (GPs) and emergency physicians discharged a higher proportion of patients home than senior nurses or ED nurses. Over half the patients referred into the ED for assessment by senior nurses were subsequently discharged. The number of patients streamed by emergency physicians was small in this study. The differences in streaming patterns were observed least for minor injury patients. Most patients streamed out of the ED were discharged home, with low referral rates to outpatient or colocated walk-in centre facilities. Clinicians with different training backgrounds will offer different patterns of clinical behaviour when engaged in streaming activity.

    Introduction

    Emergency department (ED) overcrowding has been linked to an increased morbidity and mortality, and is unpopular with patients.1 ,2 ED attendances are increasing in many countries, with just over 16 million attendances in England in 2010–2011, of whom 22% were admitted.3–,5 This suggests the public choose ED attendance over other healthcare access points for at least some of their healthcare. There have been a variety of strategies aimed at reducing the number of unplanned admissions to hospital. A recent review examining 13 interventions using data from 274 separate studies concluded that only a few techniques targeting specific groups were effective in reducing unplanned admissions.6 Those found not to be successful included specialist clinics, community interventions, medications review, ED interventions, continuity of care and hospital at home. Since most EDs are not sufficiently staffed to cater for these increasing attendances exploring alternative staffing and healthcare delivery models is vital.

    Around 76% of all London ED attendees are ‘walk-in’ with 84% of these being discharged, 22% within 1 h and 50% within 2 h of arriving.7 The proportion of walk-in patients suitable for primary care, and who should provide care for them, is widely debated with figures varying between 4.8% and 90%.8–,11 The document ‘Healthcare for London: A Framework for Action’ sets out London's strategy to improve the capital's health services and improve the efficiency of front-line care provision.12 One recommendation is the development of primary care-led urgent care centres. These may be colocated with EDs, operating as a polyclinic or as an integral part of a wider system. This model is proposed so as to reduce attendance in EDs and improve patient care, but evidence suggests that this strategy may not be successful.13–16 The process of streaming patients is an integral part of this model.7 Streaming involves the rapid assessment of patients and immediately referring them to the service most suitable to provide for their healthcare needs. Proposed models of streaming include by clinician or by trained staff using a clinical decision tool.

    At our large urban teaching hospital (census 140 000, 28 000 paediatric) we introduced a two-tier streaming system designed to facilitate early assessment, investigation and referral of all attending patients. At the time of this study, the ED consisted of three geographical regions; one dedicated to managing patients with critical illness (resuscitation), one for medical and surgical patients requiring evaluation (majors) and a minor injuries unit. The GP and nurse-led walk-in centre (WIC) was located around 50 m away from the main ED, and operated 08:00–22:00. Patients may attend the ED via a common entry point or self-present to the WIC.

    Adult walk-in patients who self-presented to our ED initially booked in at reception and were then assessed by a clinician (as defined below in the Methods section) in a dedicated area consisting of three cubicles. This area was termed streaming 1, and operated from 10:00 to 22:00. The streaming process consisted of a focussed history and examination to assess the presenting complaint. Focussed examination and physiological measurements were made at the discretion of the streamer.

    Following the initial streaming 1 assessment, the patient could be further managed in the community, or on-site, as detailed in figure 1. Each streamer worked 4 h shifts.

    Figure 1
    Figure 1

    Patient disposal from streaming 1. WIC, on-site nurse and GP-staffed walk-in centre; OP, outpatient services at hospital.

    Streaming 2 is similar to streaming 1 but designed to receive ambulance patients and patients referred from streaming 1. It is run by emergency physicians (EP) and nurses. Patients receive a focussed history and examination, baseline physiological observations, an ECG, initial blood tests and any urgent radiological investigations required are requested. As such, streaming 2 is the initial part of the patients’ journey through the ED.

    Following the introduction of this system, we undertook to audit our practice in order to assess the differing patterns of referral and discharge from the streaming 1 area by different clinicians.

    Methods

    We randomly prospectively identified 4 h blocks of streaming 1 activity (using random number generation). During each assessed block of streaming activity, we collected the patients’ hospital numbers and later obtained the patient record for the attendance. Data was collected on 28 days during March to May 2010. We identified 4 h steaming blocks on days when the study member responsible for data acquisition was on duty and randomly identified one block each day, a total of 28 blocks. The service offered six streaming blocks each day, so the potential number of streaming blocks was 168 during the study period. A GP and an EP (KM and TH) reviewed each patient record. During the study, a policy change saw the EP's roster altered, and they were no longer allocated to streaming 1 but, instead, were allocated to focus on the patients attending by ambulance. Thus, the number of EP's notes audited were less than for other groups. We scrutinised the notes for patient demographics, path through the ED, case mix, investigations, specialist referral, note keeping and admission data. All results and radiological investigations were reviewed. Data were stored on a password protected Microsoft excel database, which was then interrogated to provide the results. This paper focuses on the results obtained for the referral patterns by the healthcare providers working in streaming 1.

    Staffing at streaming 1

    The staffing at streaming 1 consisted of a GP, senior ED nurse or EP. The EP was either a consultant or fellow; the latter is a non-training middle-grade doctor with 2–3 years of postgraduate training and at least 6 months in the ED. The nurses all had at least 5 years ED experience and had completed a dedicated training in emergency nursing practice. The majority were also Emergency Nurse Practitioners (ENP), who had taken further competency-based training in the treatment of minor injuries. They were able to practice independently, and discharged patients.

    No staff received specific training other than in the operational policy of streaming 1.

    Setting

    This study took place in a large urban ED serving one of the poorest areas of the UK but also including some of the richest economic parts of the country. We serve a high proportion of people recently arrived in the UK particularly from Eastern Europe and Asia (26% white British, 24% Asian British, 9% Black British, 9% white/other, 7% Asian, 7% other). Many are temporary residents, and a high proportion do not speak English or speak it as a second language. Our population is also young (51% male, 60% 18–34 years old) and includes a high proportion of patients who work in our catchment area but live outside it.

    Results

    Three hundred and eighty-four patient records were audited for streaming 1 attendance. The breakdown of the heathcare professional performing the assessment was: GP 250 (65%), ED senior nurse—ENP 81 (21%); ED fellows 31 (8%); ED consultant 22 (6%). This reflects the proportion of streaming shifts performed by the healthcare professional.

    Overall, 85% of patients were discharged from the hospital. Seven per cent were admitted under inpatient care, and 2% to the clinical decisions unit (CDU). Disposition data was not clear in 6%; they were admitted to hospital, but whether under an inpatient unit to the CDU consequent upon bed pressures or under ED to the CDU is not clear.

    The dispositions of patients from streaming 1 are displayed in table 1 and figure 2. For analysis, all patients referred away from the ED are included as one group referred to as ‘discharged to community’ in figures 1 and 2 (this includes patients referred to their GP or outpatient clinics, to the WIC or discharged from streaming 1). The dispositions of patients referred into the ED from streaming 1 are displayed in table 2 and figure 3. The breakdown for patients discharged to the community is displayed in figure 4. A total of 20 patients were referred to the WIC from streaming, which is 5.4% (20/373) of all attendances, and 13.2% (20/151) of those streamed away from the ED. The breakdown was only available for 11 of the ED consultant patients, 9 of whom were discharged to community as shown in figure 4. The majority of patients discharged to the community by GPs, ED consultants and ED senior nurses were discharged with a diagnosis and not referred to other healthcare providers. The ED fellows referred more patients than other groups to the WIC.

    View this table:
    Table 1

    Patient disposal (either out of ED, to ED streaming 2 for assessment, to minor injury unit or direct to specialty team) by assessor at streaming 1 (GP, EP consultant, EP staff, ED nurse). Also see figure 2.

    View this table:
    Table 2

    This shows the numbers and percentage of patients referred from streaming 1 into any area of the emergency department (ED), who were discharged from the ED following review and/or investigation. See also figure 3

    Figure 2
    Figure 2

    Patient disposal (either out of emergency department (ED), to ED streaming 2 for assessment, to minor injury unit or direct to specialty team) by assessor at streaming 1 (GP, emergency physician (EP) consultant, EP staff, ED nurse).

    Figure 3
    Figure 3

    This shows the numbers and percentage of patients referred from streaming 1 into any area of the emergency department (ED), who were discharged from the ED following review and/or investigation.

    Figure 4
    Figure 4

    Destination breakdown for patients discharged to community, by clinician group.

    Statistical analysis

    The clinicians were compared using a multinomial logistic regression. This method analyses the decisions made by each group as compared with a baseline group. GPs form the majority of healthcare providers who perform the initial assessment of patients who self-present to the ED. Therefore, the decisions made by EP consultants, EP fellows and ED nurses were all compared separately to those made by the GPs by using a χ2 test for the nurses and by a Fisher's exact test for the EP consultants and staff (due to smaller numbers). If there was a significant overall difference, the contributions of each decision (either out of ED, to ED streaming 2 for assessment, to minor injury unit or direct to specialty team), to the overall χ2 test was examined to identify which specific decisions differed.

    There was evidence that GPs and ED nurses streamed patients differently (p<0.001). In particular, ED nurses streamed significantly fewer patients out of ED compared to GPs (12% vs 47%; p<0.001). ED nurses also streamed significantly more patients for ED assessment (37% vs 18%; p=0.001). Of the patients that are streamed into ED, 58% were subsequently discharged by nurses compared with 24% by GPs (p<0.0001).

    There was no evidence that either GPs and ED staff streamed patients differently (p=0.208), or GPs and ED consultants streamed patients differently (p=0.582). However, the numbers of encounters by ED staff and ED consultants is small, and the data underpowered to answer the question. As the overall differences were not significant, individual comparisons were not investigated. The patients streamed into the ED by nurses were significantly more likely to be discharged from the ED than those streamed in by GPs (p<0.0001). The remaining data involve too few patients for meaningful analysis.

    Discussion

    This study focuses on the clinical behaviour of different healthcare providers in the initial assessment of patients who self-present to the ED. Overall, 85% of all walk-in patients were discharged from the ED, which is similar to the figures for London overall (84%).7 We found that ED nurses discharged significantly fewer patients from the ED as compared with GPs and, therefore, sent a higher proportion of patients into ED for assessment. A higher proportion of the patients streamed into the department by nurses were subsequently discharged as compared to GPs. The small number of ED consultants and staff in the study precluded meaningful analysis. This study is clearly underpowered to inform on this. The small numbers reflect changes to our department structure during the study, with ED staff removed from the streaming 1 process to focus on patients arriving by ambulance.

    A 1996 Irish study found GPs admitted and referred fewer patients classified as semiurgent (defined as triage category 3 or 4 with normal physiology, and excluding patients referred by community-based GPs) from the ED to hospital, as compared with ED doctors. The grade of ED staff is not clear from this paper, but as the ED was staffed by 1 consultant, 2 registrars and 10 junior doctors, it is likely that juniors were involved in the majority of consultations. The staffing pattern in our streaming area was by consultants and staff with at least 12 months ED training. The 1996 paper does not compare EPs with GPs, but junior doctors working in the ED with GPs, and the findings, therefore, are not surprisingly in favour of more experienced doctors being more efficient decision makers.17

    The senior ED nurses were less likely to discharge patients home and requested ED review on a higher proportion of patients than ED consultants or GPs, yet in the end, a similar number of patients screened by these groups were discharged home. This reflects that nursing staff operate upon a competency-based framework, and are not trained or empowered to discharge patients that fall outside this. The nursing staff employed at ED streaming was not specifically trained for this work. It may be that had the ED nurses received appropriate training, their performance would have been similar to the GPs and EPs. Indeed, all clinicians made similar numbers of referrals to minor injuries suggesting a similar competency in this area.

    The GPs and ED nurses referred more patients directly for inpatient specialty team assessment than ED trainees and ED consultants. However, these results need to consider in the context that the primary care trust (PCT) encouraged direct referral to inpatient teams, and this may have biased a positive referral. These data suggest that the EPs were more confident in the ED, managing a wider range of conditions than GPs and ED senior nurses.

    Based on this small study, staffing streaming with GPs or ED consultants would result in a higher proportion of patients being discharged home with less impact on the ED and inpatient services. This is likely to represent a more efficient mode of healthcare delivery.

    Alternatively, these data suggest that the staff at streaming 1 need to be provided with guidelines and training as to what can be managed in the ED and what requires referral to operate more efficiently. Many of the GPs may have trained in emergency medicine several years in the past, and many of the ED nurses were ENPs and, so, may have not worked in majors for some years. It is conceivable that the newer pathways for rapid assessment and discharge were not well known to these clinicians (eg, algorithms for possible pulmonary embolism, possible cardiac chest pain, acute headache, subarachnoid haemorrhage) and these patients groups were consequently referred.

    We did not undertake a formal economic analysis. The streaming service was funded by the PCT, with an agreement to reclaim the lowest ED tariff (£75) for each patient streamed out of the department. This refund was applied equally to patients referred directly to a specialist team where the ED had not been involved in management. In the year of the study, the total number of patients streamed away from the department was 27 475 generating a credit to the PCT of £2 060 625. The cost of running the service was just under £1 m, although this does not include costs for estates, nor all administrative charges (most of which were absorbed by the GP out of hours (OOH) service). The WIC was funded on a block contract, so no extra costs were incurred from this pathway. Even allowing for some administration costs, there appears to be a significant net saving to the PCT. However, as this study was not designed to consider cost effectiveness, nor productivity of the service, a further study to investigate this would be desirable.

    Limitations of study

    This study is based on a small proportion of the attendees at our institutions. Thus, this work can only be viewed as pilot data, and a much larger study should be undertaken over a much larger number of hospitals and associated institutions. The number of ED staff and consultants involved was small, due to changes in staffing at streaming 1, thus making the study underpowered to assess their performance. We obtained data from our senior nurses as a single data point, and were not able to identify whether the practitioner was an ENP (the majority) or a senior nurse. Also, this study did not address the potential role for specifically trained streaming nurses.

    This study is based on data for a highly diverse population, in terms of wealth and racial background. In working hours, a significant proportion of ED attendees are from outside the hospital's catchment area, as they work near the hospital but live outside of London. As such, it may not represent many other areas of the UK accurately.

    Patients were not followed-up, and the diagnosis not substantiated, thus, discharged patients may have had incorrect care or an inaccurate assessment. This study provides no data on the very important aspect of quality of care delivery.

    The study would have had greater validity had we included senior ED nurses and ED trainees in the reviewing group. The study funding allowed only two reviewers to be allocated the time to perform the study. This should be addressed in subsequent work.

    The ED staff were not trainees in emergency medicine and the ED trainees were not rostered to attend streaming 1 in the study period. The staff were less experienced than trainees, and their performance cannot be translated into that expected of trainees in emergency medicine.

    Summary

    We found that GPs discharged a significantly higher proportion of patients attending streaming as compared with ED senior nurses. The EPs referred fewer patients to specialties reflecting a better familiarity with the resources provided by the ED. This is not surprising, as the patients attending ED are the group in which they have undertaken specialist training. We would, therefore, suggest that consultant EPs or GPs are the most effective practitioners to run the streaming process, minimising the time patients spend in the ED. However, few UK EDs have sufficient resources for an ED consultant-delivered service to assess walk-in patients, and this may not be the most appropriate use of their skill sets.

    Acknowledgments

    The authors would like to acknowledge the support of Jane Baylis for commissioning the study, Liz Price for analysing the results and Susie Hannah for organising patient notes and performing the structured telephone follow-up.

    References

    1. ↵
      1. Richardson D
      . Increase in patient mortality at 10 days associated with emergency department overcrowding. Med J Aust 2006;184:213–16.
      OpenUrlPubMedWeb of Science
    2. ↵
      1. Sprivulis P,
      2. Da Silva J,
      3. Jacobs I,
      4. et al
      . The association between hospital overcrowding and mortality among patients admitted via Western Australian emergency departments. Med J Aust 2006;184:208–12.
      OpenUrlPubMedWeb of Science
    3. ↵
      Opening up Government. London NHS A&E Performance Report. http://data.gov.uk/dataset/london-nhs-accident-emergency-dept-performance-report (accessed 3 Mar 2012).
    4. Australian Bureau of statistics. Population by age and sex, Australian states and Territories, June 2010. 2010. http://www.abs.gov.au/Ausstats/abs@.nsf/mf/3201.1 (accessed 4 Mar 2012).
    5. ↵
      Health and Social Care Information Centre. HESonline hospital episode statistics. Headline figures 2010–2011. http://www.hesonline.nhs.uk (accessed 4 Mar 2012).
    6. ↵
      1. Moberly T
      . Most Schemes to cut unplanned admissions do not work. GP. 16 July 2012. http://www.gponline.com/news/article/1141156 (accessed 12 Mar 2013).
    7. ↵
      A service delivery model for urgent care centres: Commissioning advice for PCTs. Pub Healthcare for London. http://www.healthcareforlondon.nhs.uk. Document ID:hfl.00012010.
    8. ↵
      1. Durand A,
      2. Gentile S,
      3. Devictor B,
      4. et al
      . ED patients: how non-urgent are they? Systematic review of emergency medicine literature. Am J Emerg Med 2011;29:333–945.
      OpenUrlCrossRefPubMed
    9. King's College Accident and Emergency Primary Care Project. Providing for primary care: progress in. London: Kings College School of Medicine and Dentisrty, 1991.
      1. Foroughi D,
      2. Chadwick L
      . Accident and Emergency abusers. Practitioner 1989;233:657–8.
      OpenUrlPubMedWeb of Science
    11. ↵
      1. Bader J-M
      . Revamp of emergency units. Lancet 1993;342:857.
      OpenUrl
    12. ↵
      Healthcare for London: A framework for action document. http://www.londonhp.nhs.uk/publications/a-framework-for-action/.
    13. ↵
      1. Heyworth J
      . Darzi centres. Flawed assumptions have hampered emergency care. BMJ 2010;341:c6852.
      OpenUrlFREE Full Text
      1. Dent A,
      2. Phillips G,
      3. Chenhall A,
      4. et al
      . The heaviest repeat users of an inner city emergency department are not general practice patients. Emerg Med 2003;15:322–9.
      OpenUrlCrossRef
      1. Nagree Y,
      2. Ercleve T,
      3. Sprivulis P
      . After-hours general practice clinics are unlikely to reduce low acuity patient attendances to Metropolitan Perth emergency departments. Aust Health Rev 2004;28:285–91.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Davis P
      . Darzi centres: an expensive luxury the UK can no longer afford? BMJ 2010;341:c6287.
      OpenUrlFREE Full Text
    17. ↵
      1. Murphy A,
      2. Gerard B,
      3. Plunkett P,
      4. et al
      . General practitioner versus usual medical care in an urban accident and emergency department: process, outcome and comparative costs. BMJ 1996;312:1135.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Contributors The study was conceived and performed by both authors. TH wrote the paper and KM reviewed the manuscript. Statistical advice was obtained professionally.

    • Competing interests None.

    • Ethics approval The study was classified as audit by the chair of ethics committee two, East London, and registered along local guidelines.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement The paper is paired with a second which was published online on August 28th 2013. Complete data is presented in the two papers.

    Linked Articles