Article Text
Abstract
The new accident and emergency (A&E) unplanned reattendance rate clinical quality indicator is intended to drive reduction of avoidable reattendances. Validation data for reattendance rates in children are awaited. The aim of this three site observational study is to establish the rate and reasons for unplanned reattendance to UK paediatric A&Es. Each centre undertook retrospective case note review of children attending at least twice within 7 days. Unplanned reattendance rates at the three centres were 5.1%, 5.2% and 4.4%. Reducing unnecessary unplanned reattendances is beneficial for patients, service capacity and efficacy. This study has identified two groups for targeting reattendance reduction: parents of children returning with the same diagnosis, severity unchanged and parents who bypass primary care resources. Clear communication and early involvement of experienced clinicians are paramount. This study has indicated that a 1%–5% unplanned reattendance rate is realistic, achievable and can drive improvement in children's services.
- Accident & Emergency
- Epidemiology
- Paediatric Practice
Statistics from Altmetric.com
What is already known on this topic
-
Reattendance consumes significant resources and reflects underlying parental anxiety.
-
Reducing unnecessary unplanned reattendance is beneficial for patients, service capacity and efficacy.
-
A significant proportion of unplanned reattendances are avoidable.
What this study adds
-
Identification of potential groups where targeted reduction of unplanned reattendance rates may be possible.
-
Communication is paramount. Acute service providers have a responsibility to positively reinforce messages given to parents by General Practitioners.
-
Maximum 5% paediatric unplanned A&E reattendance is a realistic and achievable target which can be used to drive improvement in children's services.
Introduction
Urgent and emergency care is changing. Eight new Emergency Department clinical quality indicators1 (CQI) were introduced in 2011 to provide comprehensive assessment of emergency clinical care and to drive service improvement.
One indicator, the unplanned reattendance rate, is intended to assess care and communication delivered during a patient's first presentation and thus reduce avoidable reattendances. The evidence underlying this indicator is largely from adult service data where a range of 1%–5% is considered acceptable. Validation data for rates of reattendance in children is awaited (Cooke MW, personnel communication). Reasons for reattendance are multifactorial encompassing patient characteristics, clinical care received and communication.2 ,3 High rates are undesirable but so too are low rates which may indicate excessively risk averse clinical management.
Reattendance rates are likely to differ in children compared with adults due to factors specific to the paediatric population such as parental concern. There have been a limited number of reports of reattendance rates in paediatric services but none in the UK (see online supplementary appendix 1).
The aim of this three site observational study is to establish the rate and reasons for unplanned reattendance to paediatric A&Es across three UK units and thus help to inform the national reattendance target for paediatric units. The study considered the causes of avoidable unplanned reattendance and how these may be reduced while simultaneously ensuring that children whose reattendance was appropriate are accurately identified.
Methods
Datasets were collated from three paediatric centres allowing comparison of unplanned reattendance across a wide demographic range.
Each centre undertook a retrospective observational case note review of children (<16 years) making unplanned reattendances at A&Es. Methods for each hospital are detailed in table 1.
All children attending at least twice within 7 days during the study periods were identified by computer database review. The number of unplanned reattendances was determined by case note review. The unplanned reattendance rate was calculated by dividing the number of unplanned reattendances by the total number of paediatric attendances during the study period. Children attending for planned A&E review, recalled with test results or who did not wait to be seen were excluded from the numerator. Centres 1 and 2 used fixed study periods while the Centre 3 period was defined by retrospectively identifying 70 consecutive reattendances.
Following rate calculation, detailed case analysis was undertaken. Centres 1 and 3 assessed samples of their reattenders: this was a pragmatic decision due to the longer study period and larger number of returners in the first centre and availability of case notes in the latter. Centre 2 analysed all reattenders. Demographics, mode of presentation, diagnosis and overall outcome at both presentations were recorded. Illness severity was judged to have worsened if admission or augmented treatment was required or triage category increased at second review. Reattendance diagnosis was compared with that at primary attendance and classified as follows:
-
Same diagnosis
-
severity not worse
-
worsened severity
-
-
Reclassified primary diagnosis
-
Unrelated problem.
Each unit undertook subtly different detailed analysis of the retrieved case notes prior to collaboration. Therefore, for clarity, only results which could be standardised and compared across the three units are presented. Notes were assessed by registrars or senior SHOs with consultant supervision. Clear criteria were set for reattender classification to ensure reviewer seniority had no effect on results.
Results
Detailed results are for each centre are provided in table 2 for comparison.
Discussion
The study revealed unplanned reattendance rates towards the top end of the CQI target of 1%–5%. Comparison of data from three very different centres gives confidence that these rates could be extrapolated for a significant proportion of UK Paediatric A&Es.
In this study, we suggest a significant proportion of the unplanned reattendances could have been avoided. This includes children who, at reattendance, are assessed as having the same diagnosis without worsening severity. Although the majority of these children had viral illnesses, a significant proportion had soft tissue injuries, which may reflect unrealistic advice given regarding time course for recovery. These children could potentially have been seen in primary care or stayed at home if better advice had been given initially. Therefore, the study data suggest that, with this reduction, the 5% CQI upper limit should be a realistic and achievable figure.
Deciding where to set the lower limit is more difficult. Centres should not be driven towards very low reattendance rates: this could encourage excessively risk averse clinical care which may not provide the best service. Discharge home for a trial of therapy with robust safety-net advice4 regarding when to return is often appropriate and so inevitably a small number of children will worsen and need to return. A range is required which encourages reducing reattendance while not entirely excluding it. Perhaps 1% would appear reasonable based on our findings of a mean of 1.4% of children reattending with worsening disease severity.
The high rate of self-presentation on second attendance is striking, especially given the high number of children with unchanged illness severity upon return. This may reflect that parents are choosing to by-pass primary care resources2 ,3 or a failure of the A&E team to recommend General Practitioner (GP) review as part of safety-net advice.
A significant proportion of returning children were admitted (34%, 21%, 18%), which may indicate good safety netting and appropriate reattendance. However, this may also include some who, with hindsight, should have been admitted on the first attendance.
Getting it right on first review is the key and close supervision by senior clinicians may reduce avoidable reattendance. In the document ‘Facing the Future’,5 the Royal College of Paediatrics and Child Health has clearly defined a transition to consultant delivered care, ensuring experienced ‘decision makers’ are involved early in assessment and management planning. Good communication and rapport at first consultation are also essential and will impact strongly upon reattendance. It was not our purpose to examine trainee supervision, the nature of the advice given to families or the advisability of admission on their first attendance in this study. These areas are all appropriate targets for future audit and quality improvement initiatives.
There are limitations to our data. Each centre designed and conducted their own study prior to collaboration resulting in differing methodologies and study periods. Where necessary, data were reanalysed to standardise the results presented. Direct comparison of centres is restricted by different study periods which will affect case mix and disease prevalence. However, broadly comparable results across the three centres indicate the effect of this is likely to be small. The study period at Centre 3 was retrospectively determined by the date of the reattendance for 70 consecutive cases. That is, all children who reattended during the study period were eligible for inclusion providing their index attendance was within 7 days of the reattendance, regardless of whether the index attendance fell within the study period.
In order for valid comparison of unplanned reattendance rates to be made across multiple centres, it is essential methods across all units are standardised to those used by Department of Health to performance manage the CQIs. Further clarification from DoH may be required.
In this study, we were unable to assess the impact of factors such as access to primary care, the seniority of clinician reviewing the child and social factors such as parental preference and expectations. In addition, for each individual unit the unplanned reattendance rate is just one indicator of performance. Exploring parental perspectives on reattendance, perceptions of disease severity, actions taken prior to reattending and preferred service providers will be essential when attempting to reduce reattendance rates.
Conclusions
Reducing unnecessary unplanned reattendances is beneficial for patients, service capacity and efficacy. Reattendance consumes significant resources and reflects underlying parental anxiety.
This study has identified two potential groups where a reduction of unplanned reattendance rates may be possible: (1) parents of children returning with the same diagnosis, severity unchanged and (2) parents who bypass primary care resources. To this end, acute service providers have a responsibility to positively reinforce messages given to parents by GPs and should work with GP colleagues to look at ways of reducing A&E attendances.
Improving communication is paramount in helping parents to make an informed decision as to when their child's condition requires further medical review. Providing a clear explanation of the diagnosis, natural history of the illness, a realistic time-course of recovery and unambiguous safety netting backed up by the provision of written literature could all reduce reattendance.
In conclusion, this three centre study has indicated that a maximum 5% unplanned reattendance is a realistic and achievable target which can be used to drive improvement in children's services.
Acknowledgments
Dr S Hobbins#, Dr R Raychaudhuri#, Dr J Grice* and Dr O Marzouk*. #Paediatric Department, Princess Royal University Hospital, Farnborough, Kent, UK. *Emergency Department, Royal Liverpool Children's NHS Foundation Trust, Alder Hey, UK.
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online appendix
Footnotes
-
Contributors Each author contributed equally to the planning and conduct of this work, including writing and revising the manuscript report. NJP accepts overall responsibility as the guarantor for the manuscript content and submission.
-
Funding None.
-
Competing interests None.
-
Ethics approval Anonymised, retrospective review data only.
-
Provenance and peer review Not commissioned; externally peer reviewed.