Objectives CT of the brain (CTB) for paediatric head injury is used less frequently at tertiary paediatric emergency departments (EDs) in Australia and New Zealand than in North America. In preparation for release of a national head injury guideline and given the high variation in CTB use found in North America, we aimed to assess variation in CTB use for paediatric head injury across hospitals types.
Methods Multicentre retrospective review of presentations to tertiary, urban/suburban and regional/rural EDs in Australia and New Zealand in 2016. Children aged <16 years, with a primary ED diagnosis of head injury were included and data extracted from 100 eligible cases per site. Primary outcome was CTB use adjusted for severity (Glasgow Coma Scale) with 95% CIs; secondary outcomes included hospital length of stay and admission rate.
Results There were 3072 head injury presentations at 31 EDs: 9 tertiary (n=900), 11 urban/suburban (n=1072) and 11 regional/rural EDs (n=1100). The proportion of children with Glasgow Coma Score ≤13 was 1.3% in each type of hospital. Among all presentations, CTB was performed for 8.2% (95% CI 6.4 to 10.0) in tertiary hospitals, 6.6% (95% CI 5.1 to 8.1) in urban/suburban hospitals and 6.1% (95% CI 4.7 to 7.5) in regional/rural. Intragroup variation of CTB use ranged from 0% to 14%. The regional/rural hospitals admitted fewer patients (14.6%, 95% CI 12.6% to 16.9%, p<0.001) than tertiary and urban/suburban hospitals (28.1%, 95% CI 25.2% to 31.2%; 27.3%, 95% CI 24.7% to 30.1%).
Conclusions In Australia and New Zealand, there was no difference in CTB use for paediatric patients with head injuries across tertiary, urban/suburban and regional/rural EDs with similar intragroup variation. This information can inform a binational head injury guideline.
- emergency departments
- imaging, CT/MRI
- paediatrics, paediatric emergency medicine
- trauma, head
Statistics from Altmetric.com
- emergency departments
- imaging, CT/MRI
- paediatrics, paediatric emergency medicine
- trauma, head
What is already known on this subject
In tertiary hospital emergency departments (EDs) in Australia and New Zealand, the proportion of CT of the Brain (CTB) used for paediatric head injuries is 10%, but the CTB proportions outside tertiary hospital EDs is unknown.
Studies conducted in North America have reported comparatively high proportions of CTB used for paediatric head injuries and large variation outside tertiary EDs.
What this study adds
In a retrospective study of 3072 paediatric head injuries at 31 EDs in Australia and New Zealand, there was no difference in percentages of CTB used between tertiary, urban/suburban and regional/rural hospital EDs.
There was a similar degree of variation in CTB percentages within each hospital type and patients with low Glasgow Coma Scale were seen at all types of hospitals.
This information can help inform implementation materials for a national guideline to support consistency in head injury management. The approach provides a possible template for use in other settings prior to the introduction of a guideline.
CT of the Brain (CTB) is performed, on average, for 10% of paediatric patients presenting with head injuries to tertiary hospital emergency departments (EDs) in Australia and New Zealand,1 with low-level variation across tertiary centres.2 In North America, a higher frequency of CTB has been reported, together with considerable variation across different types of hospitals.3–6 In Australasia, most paediatric patients are seen in non-tertiary EDs. At these sites, the frequency of CTB and possible variation in care are unknown, with scanning decisions generally informed by local guidance and individual clinician practice.
To identify variation in care and optimise implementation strategies in preparation for a binational head injury guideline, we assessed CTB use for paediatric head injuries in Australia and New Zealand, across different hospital types. Hospital length of stay and inpatient admissions were also assessed to determine the relationship between the proportion of CTBs performed and these important metrics.
We conducted a multicentre retrospective study of paediatric head injury presentations during 2016 to EDs in the Paediatric Research in Emergency Departments International Collaborative research network, across Australia and New Zealand.7 The study underwent central ethics review at the Royal Children’s Hospital (HREC/17/RCHM/91) and institutional review at participating sites. The development of the project was undertaken without patient or family involvement.
The strata used to categorise the hospitals was pragmatically based on the Australasian College for Emergency Medicine accredited ED role delineations.8 9 For the purposes of this study, hospital types are categorised as follows:
Tertiary=teaching hospitals in urban areas. Most have a major trauma service (MTS) and are a stand-alone paediatric hospital, or have separate paediatric specialty areas adjacent to the adult ED with dedicated paediatric emergency medicine (PEM) staff.
Urban/suburban=hospitals in urban or suburban locations without MTSs. Some have a paediatric area within the ED and dedicated PEM staff, others do not. Most are teaching hospitals with mixed EDs (adults and children) and see moderate to high volume of patients.
Regional/rural hospitals=hospitals in regional or rural areas. Some have a paediatric area within the ED and or dedicated PEM staff, others do not. Most are teaching hospitals with mixed EDs and see moderate to low volume of patients.
We included children aged <16 years, with a primary ED diagnosis of head injury and excluded return visits and those who had neuroimaging at a referring hospital. At each hospital, data were extracted for the first 100 sequential eligible cases per site (or maximum cases in the year) using a standardised report form. Primary outcome was whether or not CTB was performed. Logistic regressions were used to estimate proportions, both unadjusted and adjusted for severity using the initial Glasgow Coma Score (GCS) (GCS≤13/GCS>13). We also noted age, relevant underlying complex diagnosis (bleeding disorders, ventriculoperitoneal shunts, neurodevelopmental disability), hospital length of stay and inpatient admission (the latter two compared between sites using Mann-Whitney U and χ2 tests, respectively).
At 31 sites, 3572 visits for head injury were available. Of these, 3072 records were eligible from 9 tertiary, 11 urban/suburban and 11 regional/rural EDs (one site only had 72 eligible cases). Patient characteristics are shown in table 1. A greater proportion of children <2 years attended tertiary EDs (41.3%) in comparison with the urban/suburban (% difference (Δ)=11.3%, 95% CI 7.1% to 15.5%) and regional/rural EDs (Δ=14.8%, 95% CI 10.6% to 18.9%). The proportion of children with a GCS≤13 was approximately 1.3% in each group. Presentations with underlying complex diagnoses were higher in tertiary (4.6%) EDs compared with urban/suburban (Δ=2.5%, 95% CI 1.3.0% to 3.1%) and regional/rural hospitals (Δ=2.6%, 95% CI 0.9% to 4.1%).
The proportion of CTBs ordered among patients with head injury, and percentage that were abnormal (crude and adjusted rates) were quite similar among tertiary, urban/suburban and regional/rural hospitals (table 2).
The frequency of CTB both crude and adjusted, for individual hospitals within each group is shown in figure 1. There was similar variation in the frequency of CTB within each hospital type with tertiary, ranging from 3.0% to 14%; urban/suburban ranging from 2.8% to 13%; and regional/rural ranging from 0.0% to 13.0%. Median length of stay was: tertiary 2.6 hours, urban/suburban 2.6 hours, regional/rural 2.0 hours (p<0.001). The regional/rural hospitals admitted fewer patients than tertiary (Δ=13.5%, 95% CI 9.9% to 17.1%) and urban/suburban hospitals (Δ=12.7%, 95% CI 9.3% to 16.1%, p<0.001).
This study was undertaken to determine variation in the proportion of CTBs done after head injuries in children in Australia and New Zealand, particularly for urban/suburban and regional/rural hospitals which up until now had not been explored. Our results indicate that the frequency of CTB was low overall and that use of CTB was similar across hospital types, even when severity adjusted. Within types of hospitals, there is some individual variation; however, the range is similar across hospital types. Hence, there is likely scope for improved standardisation of CTB use in all types of hospitals. The overall unadjusted CTB rate of 6.9% for the cohort (n=3072) was lower than previously reported in the USA and Canada and we did not find high variability or higher proportions at non-tertiary centres. Comparative UK data are not available, but results may be relevant for other developed countries with CTB use lower than in North America, and where the majority of children are seen outside tertiary centres.
Our data indicate that urban/suburban and regional/rural hospitals also encounter patients with head injuries with GCS<13 and children with significant comorbidities which may increase the complexity of their assessment and management. These findings need to be considered for the development of future paediatric head injury guidelines and implementation strategies.
We had initially postulated that lower frequency of CTB may be associated with increases in admissions or observation of patients resulting in increasing ED length of stay. This has been reported to occur in some USA settings.10 In the Australasian setting, regional/rural hospitals had similar CTB use to the other hospital types but a lower length of stay in the ED, and they discharged a larger proportion home from the ED.
The main limitation of this study is that data were collected retrospectively. However, it is reassuring that the use of CTB in tertiary EDs was similar to that previously reported in prospective studies in this setting.1 In our sampling, we chose to use the first 100 consecutive records that met our selection criteria which may have introduced seasonal bias.
CTB use after paediatric head injuries in Australia and New Zealand is lower than in North America and similar across different hospital types. All types of hospitals encounter patients with low GCS and complex underlying diagnoses. The data support a generic, rather than a hospital type specific, approach to implementation of a binational head injury guideline.
The authors wish to thank research teams at the following sites: Thomas Georgeson, Shakira Spiller, Kam Sinn, Jamie Lew (Canberra Hospital); Gina Watkins, Elizabeth Walter (Sutherland Hospital); Stephen Teo (Mt Druitt Hospital); Aime Beattie, Blair Burke, Adrian Cheung, Kathryn Charlier, Emma Simmons (Tamworth Hospital); Mary McCaskill, Deepali Thosar (The Children’s Hospital at Westmead); Arjun Rao, Inas Hanna, Sophie Watkins (Sydney Children’s Hospital); Lorna McLeod, Michelle Fenton (Coffs Harbour Base Hospital); Christine Brabyn, Kirsty Greaves (Waikato Hospital); Jo Cole, Karyne Coker (Tauranga Hospital); Stuart Dalziel, Megan Bonisch (Starship Children’s Health); Adam Michael, Nicholas Edwards, Matthew Vanderberg (Bundaberg Hospital); Natalie Phillips, Sally Mcguire, Kelly Foster (Lady Cilento Children’s Hospital); Shane George, Richele Tucker (Gold Coast University Hospital and Robina Hospital); Alex King, Helena King (Toowoomba Hospital); Corey Cassidy, Amy Richter, Bo Bi, Justin Jin (Ipswich Hospital); Ben Lawton, Brooke Charters (Logan Hospital); Frances Kinnear, Ashlee Percival, Louise Spooner-Jackson (Prince Charles Hospital); Amit Kochar, Gaby Nieva (Women’s and Children’s Hospital); Lalith Gamage, Joshua Anderson (Port Augusta Hospital); Peter Archer, Lisa Vermeulen (Box Hill Hospital, Angliss Hospital and Maroondah Hospital); Simon Craig, Emma Ramage (Monash Medical Centre); Franz Babl, Ali Crichton, Cate Wilson (Royal Children’s Hospital); Mark Putland, Daniel Bourne (Bendigo Hospital); Ashes Mukherjee, Jonathon Burcham, Samantha Berkelaar (Armadale Kelmscott District Memorial Hospital); Stephen Priestley, Jessica Riordan (Sunshine Coast University Hospital and Nambour Hospital); Meredith Borland, Sharon O’Brien, Weikuei Ho, Madhuri Dama, Deirdre Speldewinde (Princess Margaret Hospital for Children); Russell Young, Tom Fox, Natalie Rudling (Albany Regional Hospital); Hugh Mitenko, Marie Draper (Bunbury Regional Hospital).
Handling editor Katie Walker
Twitter @KelFoster3, @PREDICT_network
Correction notice This paper has been updated since first published to amend author names '
Natalie T Phillips, Stephen JC Hearps, Sharon L O’Brien, Meredith L Borland, Stuart R Dalziel'.
Collaborators Paediatric Research in Emergency Department's International Collaborative: Jamie Lew, Stephen Teo, Amie Beattie, Mary McCaskill, Arjun Rao, Jo Cole, Adam Michael, Shane George, Alex King, Corey Cassidy, Ben Lawton, Frances Kinnear, Amit Kochar, Lalith Gamage, Peter Archer, Ashes Mukherjee, Russell Young, Hugh Mitenko.
Contributors FEB: conceived the study, obtained grant funding, designed the study, provided overall supervision, interpreted the data, gave final approval to be published and agreed to be accountable for all aspects of the work. CLW: designed the study, provided supervision, interpreted the data, wrote the first draft, gave final approval to be published and agreed to be accountable for all aspects of the work. EJT, EO and SD: designed the study, obtained the data, provided supervision, interpreted the data, drafted or revised it critically, gave final approval to be published and agreed to be accountable for all aspects of the work. NP, KF, SO, MB, GOOW, LM, MP, SP, CB and SC: obtained the data, provided supervision, interpreted the data, revised it critically, gave final approval to be published and agreed to be accountable for all aspects of the work. DWB: contributed to the interpretation of the data, revised the paper critically, gave final approval to be published and agreed to be accountable for all aspects of the work. SC: performed the analysis of the data, contributed to the interpretation of the data, revised the paper critically, gave final approval to be published and agreed to be accountable for all aspects of the work. Collaborators: were principal investigators at the sites and contributed their site’s data.
Funding This study was funded by Angior Family Foundation; Emergency Medicine Foundation (EMPJ-375R27-2017-PHILLIPS) Queensland; the National Health and Medical Research Council Centre of Research Excellence grant for Paediatric Emergency Medicine (GNT1058560), Australia and by the Victorian Government's Operational Infrastructure Support Program.
Competing interests None declared.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement No data are publicly available. Ethics committee approval was not provided for sharing of site data.
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.