Article Text
Abstract
Introduction Children who attend an emergency department and then leave without being seen (LWBS) are a major concern as they are a potentially vulnerable group who may come to harm through failure to provide a timely and accessible service.
Objectives We wished to establish the size of this population and importantly what subsequently happened to them over the following 7 days from their initial attendance, and the relevance of the College of Emergency Medicine LWBS Quality Indicator (QI 4) to this group.
Methods A retrospective case note review over 6 months of all paediatric attendances who LWBS.
Setting Southampton Paediatric Emergency Department.
Results During a 6-month study period, 10 795 attended, of which 544 (5%) LWBS. 12.6% (69/544) reattended over the next 7 days, of which 14 were admitted, 7 for <12 h.
Conclusions The majority who LWBS do so during peak times. Very few paediatric patients who LWBS then reattended required admission for >12 h (7/544, 1.3%). The rate of reattendance of those who LWBS and review of their case notes is potentially more valuable than the LWBS rate alone.
- paediatrics
- emergency department
Statistics from Altmetric.com
Key messages
What is already known on this subject?
Those children who leave the emergency department without being seen have been identified as a cause for concern leading to the introduction of the left without being seen (LWBS) Quality Indicator.
Studies from outside the UK have demonstrated that those who LWBS often do so at times of high demand, tend to be of low acuity, with low reattendance rates.
What this study adds?
This is the first study looking at the UK LWBS population.
The authors found an LWBS rate of 5% and a reattendance rate within 7 days of 12.6%.
Those who LWBS tended to present at times of high demand, were of low acuity and were unlikely to come to significant harm.
Introduction
Patients who leave without being seen (LWBS) are a concern for emergency departments (EDs), especially the potentially more vulnerable paediatric population. Do these patients subsequently need admission, do they come to harm through treatment delays? Existing research has suggested that this patient group attend at busier times when departments were overcrowded with longer waiting times, have complaints of lesser acuity and leave because the problem that triggered the presentation had either resolved or improved to a degree that the caregiver no longer felt it necessary to attend.1–9
The Department of Health in conjunction with the College of Emergency Medicine have published Quality Indicators (QI). Indicator number four stipulates LWBS rates of <5% should be achieved to reduce the risk of adverse events.10 The UK is not alone in using the LWBS rate as a means of attempting to measure quality and safety in emergency medicine. The International Federation for Emergency Medicine, which has >70 member countries (including the USA, Canada, New Zealand and Australia), highlights the LWBS rate as a process measure that can be used as part of the toolkit to assess quality and safety.11
Most of the relevant literature describes this patient group in terms of presenting complaints and demographics with comments on what subsequently happened to them, but none were UK-based studies.1–9 The authors agree this should be the primary outcome of concern. This study sought to contribute to the literature from the perspective of a UK University Teaching Hospital ED with major trauma facilities, and specifically attempted to identify whether these patients came to significant harm, and therefore to consider the usefulness of QI 4 as a marker of patient care.10 If patients who LWBS do not come to significant harm, this QI might better reflect demand rather than need. This led the authors to question whether it is more prudent to concern ourselves with the reattendance rate among those who LWBS, and argue that this might be a more sensitive indicator of quality of care delivered.
There are other factors that may mean that the results of other studies cannot be generalised to a UK population and therefore make this an important study. These include differences in healthcare system design (ie, state vs privately funded), differences in the total numbers attending the ED and ease of access to other healthcare resources (eg, primary care, walk in centres). Also, there may be important geographical and socioeconomic differences.
Methods
A retrospective case note review was undertaken for all patients <18 years of age attending the Southampton Paediatric ED, who were coded on discharge as LWBS from our electronic record system during a 6-month period (February–July 2013). This ED, supported by both paediatric high-dependency and intensive care unit facilities, sees an average of 100 000 patients annually, of which 22 000 attend the Paediatric ED.
The authors scrutinised both ED notes and all hospital electronic patient notes to identify those reattending within 7 days from their initial presentation to the ED, and recorded the need for subsequent care, admission to hospital and level of care required during admission. Patients were excluded from this review if they presented with an out of area postcode as it was felt they would be less likely to re-present to this department and would not be detected on this hospital’s database.
The primary outcome measures were the total numbers who LWBS and reattended within 7 days of initial presentation. Secondary outcomes were, for those who reattended, admission to hospital rates, additional care required and level of care required if admitted. Ethics approval was deemed unnecessary given this was an audit against a Department of Health standard.
Results
From a total of 10 795 paediatric attendances during the study period, 589 were identified as having been coded as LWBS. A total of 45 were excluded as outlined in figure 1. Of those excluded were a group of children who due to their presentation and observations had been referred directly to paediatrics (‘fast tracked’) but incorrectly identified as LWBS.
Age at time of presentation ranged from 19 days to 17 years with 5.7% attending by ambulance. Most common presenting complaints (at registration or triage) included musculoskeletal (35%), head injuries (13%), ear, nose and throat (ENT)/maxillofacial (10%) and gastrointestinal complaints (9%). A total of 24 patients (4.4%) were registered with no formal triage assessment completed. All but one patient had child protection checks performed. Two patients were known to social services and one had a child protection alert.
The majority of those who LWBS presented from Saturday through to Monday, with 58% of LWBS patients presenting between the hours of 18:00 and 23:59 (figures 2 and 3).
Primary outcome measures
Out of 10 795 attendances during the study period, 5% (n=544) LWBS. There were 69 unplanned reattendances within 1 week of initial ED presentation for this group, 12.6% (69/544).
Secondary outcome measures
In total, 39 of the 69 who reattended were managed within ED (57%), with 13 (33%) requiring additional medications such as steroids, X-rays or minor procedures (wound care, pulled elbow manipulation) (table 1). In total, 14 of the 69 were admitted as paediatric inpatients via ED or general practice. Of these 14 children admitted, 7 were discharged after a period of observation with advice or open access to the paediatric admissions unit. The remaining seven were admitted for >12 h as outlined in table 2.
Seven (10%) children who reattended ED were treated as possible or missed fractures and were referred onwards to the paediatric fracture clinic, one of which had been identified by an investigation performed at initial presentation.
Discussion
Our motivation for this study was to ascertain whether this population come to harm as a result of leaving without being seen and whether this is reflective of increased service pressures at critical times. We are also mindful that the LWBS group could incorporate those children at greater risk of child abuse. Our results indicated that those who LWBS are unlikely to experience serious adverse outcomes. Interestingly and perhaps unsurprisingly, a large proportion of these LWBS occurred during times of high demand, namely over the weekend and out of hours. This finding of lack of adverse outcomes in those who LWBS has been previously highlighted in other studies.1–2 ,4–6 ,12
Anecdotally and during the review of case notes a long wait with an associated lack of information about this and a caregiver’s perception that a child's condition had improved contributed to caregivers leaving prior to being seen. These findings are in keeping with the existing research and have prompted a review of internal processes within our department to look at how we manage demand and caregiver/patient expectations, particularly during peak times.
The authors were surprised by the number of unplanned reattendances that were 12.6% in our population. These data, however, are in keeping with pre-existing research from the USA, Canada and Australia, which is summarised in table 3.
Most recently, Gravel et al report higher levels of unfavourable outcomes, none of which included death or suicide attempt. Six children required surgical procedures and one required fracture reduction. Twenty-one children were hospitalised in their cohort; however, the authors did not qualify the level of care they required during those admissions or the length of stay. Potentially this higher level of unfavourable outcomes reported in comparison to the other studies cited, including this work, reflects the author's accepted broader definitions of what constituted an adverse outcome.
The authors recognise certain limitations with this study. By restricting our audit to those presenting with a Southampton post code in the belief that they would only reattend this hospital, it is possible to have biased our findings. Lack of access to the notes from general practice (GP) surgeries, other EDs, minor injuries units or walk in centres may mean that we did not have an accurate picture of the true reattendance rates. However, it is likely that significant further treatments, for those with a Southampton post code, would have involved either the hospital inpatient or outpatient services that our searches included.
As only a 6-month period was considered, there may also be seasonal variations in LWBS rates that are not reflected by these data. Finally, our results may not be representative of other UK centres and therefore be difficult to generalise. However, our results do seem to reflect the wider international picture as illustrated in the existing literature.
Conclusions
While concerns exist regarding children who LWBS and efforts to minimise these remain important, our results offer some reassurance that these children are generally of lower acuity and unlikely to come to significant harm. It is important, however, that further prospective, multicentre studies are conducted to ensure these findings are consistent across the UK.
Our results and those from international sources demonstrate that patients who LWBS tend to occur during times of peak activity, that is, out of hours and weekends. We suggest that for the paediatric population the LWBS QI may be a less sensitive measure of quality of care and may in fact be a more relevant marker for demand, and that reattendance rates for those who LWBS are potentially a more useful indicator of the quality of care delivered.
References
Footnotes
Contributors NS conducted audit, writing and submission of paper, and revision of paper TP: conducted audit, writing for paper, revision and re-submission of paper. MC proposed audit, supervisor and edited manuscript.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.