• Accident & Emergency
• Epidemiology
• Health Service
• Health services research
• Paediatric Practice

Admission to hospital can have a major impact on children and families. Unnecessary admission suggests a failure in the system. Gill et al have analysed hospital episode statistics for children less than 15 years in England between 1999 and 2010.1 They confirm anecdotal evidence that there has been an increase of 28% in emergency admission rate over the last decade with a persistent year on year increase from 2003 onwards. There has been a twofold increase in the number of children admitted for less than a day with short-term illness. An increase in upper respiratory tract infection admissions by 22% and lower respiratory tract infection admissions by 40% illustrates their findings and they also show admission rates for chronic conditions falling by 5.6%. This effect has been shown particularly in children aged 5 years and less.

This is a temporal association, which of course does not infer causation. However Gill's conclusions present the paediatric community with a huge challenge. They conclude that this increase in short-term admissions represents a systematic failure in primary and hospital care. Solving this problem is likely to require a restructuring of the way we deliver acute paediatric care. Although this is very complex with multiple factors, Gill's main conclusion is that the timing of the rise would implicate two main issues; the National Health Service (NHS) contractual arrangements allowing general practitioners to opt out of providing cover ‘out of hours’ and the impact that the 4-h wait target in the emergency department (ED) has had on patient management.

So what is it in the system that is failing? Do these data really suggest a failure of the way primary care services are provided? Is this due to changing attitudes and expectations towards healthcare services from parents? Is it really a combination of anxious parents, inexperienced junior doctors and pressure from the 4-h waiting target in EDs? Is it the use of care pathways that discourage proper assessment before formal admission? Does the increased use of short-stay units make it easier to admit for observation rather than making a diagnosis and offer a definitive management plan and discharge? Do rapid results technology and bedside testing like oximetry, increase the likelihood of admission? Are practitioners becoming more risk averse as the perceived risk of legal claims increases, so are more likely to admit children to hospital? Is paediatric training in general practice and secondary care in need of restructuring with further emphasis on competent management of simple conditions? Could it really be the effect of the tariff system in England influencing the desire for hospital admission? How much do public misconceptions and understanding about illnesses and health services, affect their expectations? How does the health-seeking behaviour of parents with increasing likelihood for direct presentation to the EDs, bypassing primary care, have an effect on acute presentation and admission to hospital? Gill et al have certainly laid down the challenge.

There is qualitative and anecdotal evidence that a consultant-delivered service will reduce unnecessary admissions in emergency care, improve safety and training and supervision. The Royal College of Paediatrics and Child Health (RCPCH) supports the role of the consultant in delivering care supporting the principles of quality, safety and sustainability.2 Is it too expensive to deliver this model and are consultants willing to deliver this kind of patient-centred care? Perhaps this is part of the answer. We need more quantitative data on the impact of this intervention.

Is the 4-h target fit for purpose or is it simply a means to move a problem from one department of the NHS infrastructure into another department? Gill et al may be right, that all it does is increase the pressure on junior doctors to admit and move on the decision of management and definitive treatment to elsewhere, rather than deal with it there and then. In a recent systematic review by Jones and Schimanski, the impact of the introduction of an ED time target and had not resulted in a consistent improvement in care which appeared to be markedly varying effects being reported between hospitals. They warned other countries wanting to ‘emulate the UK experience’ that they should proceed with caution.3 We do not know the real impact of the 4-h target in the children's EDs in the UK. This needs further assessment.

What is the evidence that the changes in out-of-hours arrangements have made a major impact in ED referrals for children? In a study examining hospital admission from 1997 to 2006, Saxena's conclusions were that the majority of paediatric admissions were minor illness episodes that could be better managed by primary care in the community and may be evidence of a failure of primary care services.4 This was before the changes in the contractual hours for general practitioners (GP). So perhaps this failure in primary care goes back beyond the out-of-hours contract changes and the new contract has simply magnified it. Do GP training programmes have enough paediatric aspects to them and indeed should all training programmes have a mandatory component of paediatric experience?

Does NHS Direct advise parents appropriately, about paediatric referrals to ED? Delivering healthcare telephone advice about illness severity in children is difficult without seeing the child. More prospective studies are needed to assess NHS Direct's impact on clinical assessment, especially in children under 5 years of age.

What tools do we have to help us decide how unwell a child is when they present to unscheduled care? This remains a challenge to all involved in assessing unwell children, not only inexperienced junior doctors. Innovations such as the Paediatric Observation Priority Score may help us in the emergency setting in hospitals, but this score is still undergoing evaluation. There are no well-validated tools to help the general practitioner identify children with potential serious illness in primary care. There are also few paediatric tools to help us assess the appropriateness of admission once a child is admitted.

There are other alternatives to admission to hospital such as developing a community children's nursing team and an acute Hospital in the Home service. In reality there are little UK data to help us pull this challenge apart. In a systematic review examining the interventions that may prevent hospital admission in children by Coon et al,5 their conclusion was that there was very little evidence on which to base an optimal strategy for reducing admission rates.

These data are from England only. It would be very interesting to see if similar trends have been occurring in Wales, Scotland and northern Ireland where the NHS structure varies; funding and resources, consultant to patient ratios, tariff systems, commissioning of healthcare and the organisation and policy process are all different. This geographical variation must be a rich source to discover what works well and what doesn't work at all. National databases and minimum datasets on common problems should be collated to examine in detail, disease patterns and current management of unscheduled care presentations.

General practice and the paediatric community have to rise to this challenge. We need to review the possible causes and inter-relationships, which may account for these changes and provide a more effective model for delivering healthcare for acutely unwell children. It will require very careful evaluation using multifaceted assessment tools for examining complex interventions and realistic evaluation methodology. There is an urgent need for quality, well-conducted, qualitative and quantitative research to enable informed service change.

As we work through the current workforce issues, service reconfiguration and changes in postgraduate training, we need to use this as an opportunity to get it right. There are a number of research networks recently established such as the Paediatric Emergency Medicine Research Group of UK and Ireland and as the Medicines for Children Research Network and paediatric non-medicines network undergo new changes in their structure, the paediatric community needs to develop well-funded multicentre studies addressing these issues using these networks. Key stakeholders need to be involved with this process, underpinned by the new operating framework for the NHS and the Standards for Children and Young People in Emergency Care Settings developed by the Intercollegiate Committee for Standards for Children and Young People in Emergency Care Settings. As Gill et al say, the restructuring must be done with empirical evaluation of the policy options. This must be done at a national and coordinated level.