Children account for nearly one-fourth of all care provided in emergency departments (EDs) in the UK and the USA.1^–6 Two issues highlight the differences between children and adults who seek care in ED: first, there are physiological differences between adults and children; and second, the variety of disease and injury patterns seen in children are different from those in adults.7 8 In the USA, it is estimated that children receive their emergency medical care from paediatric emergency medicine physicians (PEMPs) only 23% of the time4 8 and in separate paediatric EDs (PEDs) only 7% of the time.2 8 With children receiving the majority of their care in general emergency departments (GEDs), and the minority in PEDs, it is possible that the variability of physician training and types of EDs may impact quality of care.

Recognition of the need for specialised emergency care for children formally began in the US in 1985 with the development of the Emergency Medical Services for Children (EMSC) programme.6 Since that time, educational tools, medication and equipment standards, protocols, practice guidelines and facility preparedness standards have been developed.7 Paediatric emergency medicine (PEM) training programmes have been established, and the last decade has seen a 64% increase in PEM fellows in training.9 Currently there are 46 accredited PEM fellowships in the US10 and 1450 PEM providers nationwide.11

In the UK, the Specialist Training Authority officially recognised PEM as a subspecialty in 2003, and in 2007 the first RCPCH National Grid trainee completed formal training in PEM; however, employment in senior medical staff positions does not currently mandate accredited PEM training.12 13 The 2007 “Services for Children in Emergency Departments” document recommends that a consultant with subspecialty training in paediatric EM be appointed for each emergency department with greater than 18 000 annual paediatric ED visits,5 13 a goal not currently being met.13 14 The College of Emergency Medicine recently increased the requirement for paediatric training for EM doctors, mandating an increase to 6 months of paediatric-focused training.5 Most recently, in 2008 The Confidential Enquiry into Maternal and Child Health (CEMACH) report proposed a national audit on the standards of delivery of care to children in emergency departments and advocates for the role of paediatric consultants in the Emergency Department.5 The authors noted the inability of junior doctors to recognise serious paediatric illness and poor provision of senior ED doctors to provide a safety net for children as liabilities within the present system of caring for children in their emergency rooms. The benefit of consultant staff was also supported in a 2007 study by Geelhoed and Geelhoed who retrospectively demonstrated that the provision of additional consultant medical staff in an Australian PED coincided with decreases in the percentage of admissions, patient complaints, average waiting times and operating costs.15 These data highlight both the benefit and deficit of PEMPs, but what is the clinical impact on children cared for in systems without access to PEM providers?

METHODS AND ANALYSIS

We performed a review of the literature using Medline and Google Scholar with key words: “practice pattern variation; paediatric versus adult emergency medicine; emergency variation paediatric; trauma paediatric versus adult; paediatric and emergency medicine.” To better incorporate publications outside the US, all above searches were repeated using “paediatric” and “paediatric emergency competencies.” References of each chosen article were also scanned for inclusion of related cited studies. In total, 18 studies that evaluated differences in care provided to children in emergency service settings were identified. Of these, three assessed care provided to febrile children, one of which was specific to febrile seizures,16^–18 and seven compared the delivery and outcomes of trauma care.19^–25 A thorough summary by Stylianos and Nathens in 2007 compiled 15 years’ worth of relevant comparative studies of trauma care at children’s versus adult hospitals, but this was not included in the analysis, as it was not a primary study.26 Additionally, a 2007 comparison of practice pattern variation of febrile infants by Seow et al from Taiwan was identified yet eliminated, since the comparison group was general paediatricians and not PEMPs.27 A 1998 study by Schweich et al was also identified but not included, since it was a survey more than a decade old, not outcome-based, and 50% of its responders were not fellowship-trained.28 In this review, we will explore the differences in the evaluation and management of paediatric patients with traumatic injuries and fevers, including febrile seizures, but we will not discuss single-studied topics such as DKA, croup, analgesia, asthma and bronchiolitis,29^–33 although all primary studies are summarised in table 1.

TRAUMA CARE: PROCESS OF CARE AND OUTCOMES

Injury remains the single greatest cause of morbidity and mortality in the UK and US among children beyond 1 year of life and adolescents. Two studies, collectively including 73 526 children, evaluating paediatric injury care demonstrate lower mortality outcomes at paediatric trauma centres (PTCs) compared with adult trauma centres (ATCs) with additional qualifications (ATCAQ) in paediatric trauma care26 or general ATCs.20 Similarly, Densmore et al determined that in more than 79 000 patients, mortality and other outcome data such as length of stay and hospital charges were significantly higher when paediatric patients are treated in adult hospitals compared with children’s hospitals when injury severity score (ISS) was controlled for; however, he did not evaluate outcomes by trauma centre designation.20 In contrast, Osler concluded that the lower mortality at PTCs disappeared following regression analysis for ISS.23 Splenectomy rates, a measure of process of care, more than doubled when paediatric trauma victims were treated by adult surgeons, with nearly equivalent injury severity scores.21 24 Furthermore, functional feeding, another outcome measure, at time of hospital discharge has been shown to be improved at PTCs compared with ATCAQs and ATCs.26

FEVER EVALUATION: PROCESS OF CARE AND OUTCOMES

Caring for young children with fever has remained a controversial topic, despite American Academy of Paediatrics (AAP) and American College of Emergency Physicians (ACEP) sponsored and widely disseminated guidelines in both the paediatric and adult emergency medicine literature.18 34 35 Similarly, published practice parameters supported by the AAP outline diagnostic evaluation recommendations of healthy infants and children who present for assessment following their first simple febrile seizure, a common emergency room scenario for children aged 6 to 60 months.36 Isaacman demonstrated multiple variations in physician practice patterns between PEMPs and GEMPs in the evaluation and management of febrile infants, as did Hampers et al with regard to the management of febrile seizures.16 18 In the Isaacman review, PEMPs were more likely to obtain CBCs, blood cultures and urine cultures, while GEMPs were more likely to obtain chest x rays and CSF samples and to treat with empiric antibiotics. These results are similar to those of Hampers et al, who found that GEMPs were more likely to perform LPs and administer parenteral antibiotics when evaluating infants with febrile seizures. Belfer et al demonstrated overall poor adherence to published guidelines for both PEM fellowship directors and EM residency directors with significant practice variance in the hypothetical management of febrile infants and children.17 Unfortunately, none of these studies included any true outcome measures, such as cost of care, adverse clinical events, parental satisfaction or unscheduled revisit rates.

DISCUSSION

We have tried to answer the question: when children are acutely ill and seek emergency services, who should care for them? Unfortunately, the data are limited, although there is a suggestion that children who require trauma care have better health outcomes, including lower mortality and splenectomy rates, if they are cared for in a PTC or ATCAQ. This finding, while certainly relevant to the care of acutely ill children, is not directly related to our original question given that the physician composition of PTCs or ATCAQs is not uniquely dependent on the presence or absence of PEMPs, and the decision to move to splenectomy is based on the experience and training of the surgeon, not the emergency room physician. The information in the one other area in which there is more than one article, the care of febrile children, is less conclusive, since only adherence with guidelines and the use of ancillary testing were measured, rather than examining true health outcomes.

What is evident from our review is that the majority of children receive their acute care from GEMPS, who continue to be the nearest capable physician in most settings. In recognition of this, and to address the unique needs of children in emergency departments, the AAP and ACEP jointly published guidelines in 2001. In 2007, The Institute of Medicine of the National Academies published a book entitled “Emergency Care for Children: Growing Pains” which was aimed at arming the emergency care workforce with improvements in paediatric knowledge and skills.37 This was followed in 2008 by an AAP policy statement on the management of paediatric trauma.5 38 The minimal requirements as outlined in these documents of prehospital care services and in-hospital needs, such as staffing, medication, equipment and supply requisites, must be met by any system that cares at any time for children in order to effectively evaluate, stabilise and, when necessary, transfer acutely ill children. Continued research through the EMSC programme and the Pediatric Emergency Care Applied Research Network (PECARN) is needed to develop evidence-based protocols for key management issues of the injured or acutely ill children in an effort to provide higher and more consistent quality of emergent care for children, a substantial need most specifically highlighted in 2002 by Moody-Williams et al,39 and a need that persists today, since as the IOM report best summarises, the care of children in emergency departments today is “uneven.” 37

Despite our search strategy, we may have not identified every relevant study; however, in addition to using multiple terms, we also examined the reference lists of all identified papers. Second, our conclusions must be tempered by the overall lack of data. Our results, nonetheless, are consistent with those in other areas of medicine in which subspecialty care of children has been compared with generalist care and found improved national guideline adherence with subspecialty care, but without measurement of true outcomes. This is clearest for paediatric asthma care.40 Conclusions are similarly limited in the adult arena, studied most recently by Smetana et al who reviewed 49 articles in an analysis from 2007 which compared generalist versus specialist care and concluded that despite the common belief that specialty care is superior, the influence of specialty training on outcomes could not be determined secondary to methodological shortcomings of available studies.41 Third, the inability to reliably distinguish between levels of training and accurately categorise “PEMPs” was another obstacle of this review given that a multitude of paths to subspecialty training exist, both in the US and in the UK, and that studies do not typically declare levels or paths of training. Ideally, it might be more useful to compare competence rather than species, but unfortunately neither is possible from the available data.

Every society is committed to providing outstanding healthcare for its children. The general perception is that children do better if they are cared for by paediatric subspecialty physicians. Our review hints that this may be true in the field of emergency medicine, particularly with respect to trauma; however our results are limited by the amount of data, the inability to reliably interpret the data given that formal training differentiation is not always possible, and the focus on process, rather than health outcomes. In many respects, the delivery of emergency services to children is like other specialties where there is limited information that specialty-trained providers improve health outcomes. Future studies should focus on the prospective evaluation of true healthcare outcomes when care is provided by paediatric emergency medicine subspecialists compared with GEMPS. And perhaps even more importantly, future studies should address “how” children should be cared for—rather than “who” is providing their care—and if improved adherence to current and future AAP and ACEP guidelines would improve the care that they receive in all varieties of emergency settings.