Incidence and prevalence

Feverish illness is very common in young children. Figure 1.1 shows the proportions of children from a birth cohort of all infants born in one English county (Avon) whose parents either reported a high temperature or presented to a doctor for this reason.¹ It can be seen that a high temperature is reported by nearly 40% of parents of children aged under 6 months, and in over 60% of children in the other age ranges between 6 months and 5 years. Between 20% and 40% of children in the various age ranges are taken to a doctor because of fever, with the highest proportions presenting between the ages of 6 and 18 months. It has been estimated that an average of eight infective episodes occur in otherwise healthy children during the first 18 months of life.²

Figure 1.1

Proportions of children reporting and presenting to doctors with high temperature by age range; data from Hay.

The prevalence of feverish illness in children is reflected by statistics from primary care. Fever is probably the most common reason for a child to be taken to the doctor. In a study of 1% of the national child population, the mean general practice (GP) consultation rate was 3.7 per child per year and almost double that rate for children aged under 4 years. Infections and respiratory disorders made up over 40% of the consultations.³ In the fourth national study of morbidity in general practice, which included nearly 10 000 children, the annual consultation rates for infections were 60% of the population aged less than 12 months, 36% aged 1–4 years and 20% aged 5–15 years.⁴ Not surprisingly, fever in children is also a common reason for seeking health advice out of hours. In one service, 34% of calls concerned children under 5 years of age.⁵ Fever was a concern in 52% of calls about children aged under 12 months and in 64% of calls about children aged 1–5 years.

Feverish illness is also one of the most common reasons for children to be seen in hospital emergency departments and it is a leading cause of admission to children’s wards. In a study from an emergency department in Nottingham, 32% of the 120 000 annual total attendances were for children.⁶ Febrile illness was the second most common medical reason for attendance, accounting for 20% of such cases. On children’s wards, at least 48% of admissions are associated with infection. Most of these infections present with a feverish illness with or without other symptoms such as breathing difficulty, fit, rash or cough. Feverish illness is second only to breathing difficulty as the most common presenting problem leading to acute hospital admission in childhood.⁷

Issues for healthcare professionals

Feverish illness in young children is a diagnostic challenge for healthcare professionals because it is often difficult to identify the cause. In most cases, the illness is due to a self-limiting virus infection and the child will recover quickly without intervention. However, fever may also be the presenting feature of serious bacterial illnesses such as meningitis, septicaemia, urinary tract infections and pneumonia. Estimates of the incidence of these and other serious infections are given in Table 1.1. Although there is quite a large variation in the estimated incidences according to the source of data, it appears that up to 1% of children aged 0–5 years may have one of these infections each year.

Table 1.1

Estimated incidence of serious infections in children aged 0–5 years in the UK; data from Hospital Episode Statistics (HES).

In some children with fever there will be symptoms and signs that suggest a particular infection, such as an inflamed eardrum in a child with otitis media or a non-blanching rash in a child with meningococcal septicaemia. When these features are identified, the diagnosis can be established relatively easily and the child can be treated appropriately. There will remain a significant number of children, however, who have no obvious cause of fever despite careful assessment and investigation. These children with fever without apparent source (FWS), are a particular concern to healthcare professionals because it is especially difficult to distinguish between simple viral illnesses and life-threatening bacterial infections in this group.⁸ In general, FWS tends to be a problem in young children, and the younger the child the more difficult it is to establish a diagnosis and assess the severity of illness. Because of these problems, a number of diagnostic and management strategies have been developed for feverish illness without obvious source in young children.⁹

To further complicate the problem of assessment and diagnosis, the clinical picture often changes rapidly in young children. The condition of young children with serious illness may deteriorate within hours of onset but, on the other hand, an ill-appearing child with a viral illness may make a rapid recovery. Thus, another challenge for healthcare professionals is to determine when to observe the child for a period of time, and when to investigate and begin treatment.

Most healthcare professionals are aware that infectious diseases were, and remain, an important cause of mortality and morbidity in childhood. In the past hundred years there have been impressive reductions in childhood mortality. The infant mortality rate in the UK, for example, has fallen from 20% to 0.5% since 1890. Much of this improvement has been due to public health measures, and immunisation against infectious disease has increasingly been an important factor. In recent years, the reduction in childhood mortality has changed only a little. In other countries, mortality rates have continued to fall and some European countries now have childhood mortality rates that are 30–40% lower than that of the UK. These figures suggest that more can be done to reduce childhood mortality in this country.

Figure 1.2 shows that infection is a major cause of mortality in children aged 0–5 years. There are over 100 deaths from infection in children aged 1–12 months each year in England and Wales. In the first year of life, infection is second only to congenital defects as a cause of death. In children aged 1–4 years there are around 30 deaths from infection per year of life, and infection is the most common cause of death in this age group.

Figure 1.2

Contributions of the four major causative categories to childhood mortality, England and Wales, 2004; neonatal deaths and deaths due to perinatal events have been excluded; data from the Department of Health, courtesy of R MacFaul.

It is possible that the childhood mortality rate in the UK could be reduced to a figure in line with other European countries if the proportion due to infections could be reduced. Immunisation will probably play an important part in this process. For example, the new pneumococcal conjugate vaccine, which was introduced into the UK schedule in 2006, has led to a dramatic reduction in invasive disease due to Streptococcus pneumoniae in other countries.¹⁰ However, it is likely that improved recognition, evaluation and treatment of febrile illnesses in children could also lead to a reduction in mortality from infectious disease. For instance, a recent national study investigated deaths from meningococcal disease, which is the leading cause of mortality from infectious diseases in children.¹¹ The researchers found that mortality from meningococcal disease is often associated with late identification, sub-optimal treatment and other deficiencies in health care.

There is some concern that there is considerable variation in the provision of care for children with feverish illnesses across the UK. At present there are no national guidelines on the management of such children and practice has developed in different ways in different areas. For example, in many paediatric units it is common practice to observe febrile children for several hours while assessment takes place, but in other units it is not. In some situations there is evidence that differences in practice can affect outcome. For example, this could be construed from the above-mentioned study of meningococcal disease.¹¹ It is also known that the outcome from infectious diseases can be associated with the degree of affluence or deprivation of the area in which children live. Another study of meningococcal disease has shown that the mortality rate from the disease for children in the most deprived areas is three times that of children from the most affluent areas of the UK.¹² The case mortality rates are also significantly higher in children from deprived areas. Differences in childhood mortality rates due to health inequality are well recognised, and it is an objective of a Public Service Agreement issued by the Department of Health in 2001 that the gap in infant mortality between different social groups should be reduced by 10%, by 2010. Addressing differences in the management of febrile illnesses in young children may be one way of helping to achieve this target.

Parental concern

It is clear that febrile illnesses continue to have a considerable impact on childhood mortality and morbidity. This impact is reflected in the concerns of parents and carers. Kai conducted a survey of parents’ responses to acute illness in their children and found that fever, cough and the possibility of meningitis were parents’ primary concerns when their children became acutely ill.¹³ The parents were often worried that an illness might herald potential harm. In the case of fever this included the development of meningitis or fits, or permanent impairment of some kind, such as brain damage or even death. Parents were also concerned that the presence of fever itself could damage their children. This concern, which can lead to what has been described as fever phobia,¹⁴ is quite widespread and tends to increase with the height of temperature. In scientific terms, fever is a natural response to infection and is not harmful in itself. It is the underlying infection that has the potential to cause harm. Indeed, there are some theoretical grounds to suggest that fever is beneficial in the body’s response to infection. In any event, it is clear that parents and carers could receive more useful advice about feverish illness from healthcare professionals. This could include information about detecting potential serious infections and how to manage fever.

Need for guidance

It is a requirement of the Children’s National Service Framework that all ill children should have access to high-quality, cost-effective, evidence-based care.¹⁵ From the above, it can be seen that there is a need for evidence-based guidance to inform healthcare professionals about how to judge whether a child who presents with a fever is likely to develop a serious illness. Healthcare professionals also need advice to support their decision on whether to observe the child, to perform diagnostic tests, to start treatment such as antibiotics or to refer onwards for specialist care. The guidance would also usefully include advice on the best ways to detect fever, the management of fever itself, and what to inform parents and carers who have made contact with the health services. The guidance should be applicable to primary and secondary care and should take account of the number of agencies that are involved in giving health care and advice to parents and carers. It is also important that parental preferences, as well as the child’s best interests in terms of health outcomes, should be taken into account when considering the various options for investigation and treatment.

What is covered

1. the accuracy of different measurements of body temperature, including the methods and sites, and how to interpret the height of fever
2. in a child presenting with fever, identification of signs and symptoms that would help to establish the possible diagnoses and focus for infection
3. in a child presenting with fever, identification of clinical signs and symptoms that would help to predict the severity of the child’s illness
4. identification of which clinical signs and symptoms would direct the healthcare professional to carry out further investigations, what these investigations should include and how to interpret them
5. when a child presenting with a fever should be started on treatment (for example antipyretics and/or antibiotics) to try to improve their condition or manage their illness
6. thresholds for referral:
   • what clinical signs or symptoms can be used to identify young children who should be referred
   • what additional factors should be taken into consideration when deciding whether or not to admit a young child to hospital
   • which clinical signs or symptoms should be used to identify young children who should be referred directly to intensive care
7. what advice should be given to parents and carers following the child’s initial assessment by the healthcare professional, including the use of antipyretic drugs and other cooling methods.

What is not covered

1. management after a specific diagnosis has been made
2. management beyond initial stabilisation
3. feverish illness in children already admitted to hospital
4. children with a pre-existing co-morbidity for which the presentation of fever is already covered by an established management plan by their specialist team, for example those with cystic fibrosis or immunosuppression
5. children presenting with recurring and/or persistent fever
6. management of febrile convulsions
7. children with tropical diseases.

Definitions used in the guideline

It was necessary for the GDG to define certain terms that could be used as inclusion or exclusion criteria for the guideline and literature searches.

Care pathway

The GDG designed an outline care pathway early in the development process to explore how a child with feverish illness might access, and be dealt with by, the health services. The resulting pathway is shown in Figure 1.3. The pathway starts with a child at home with fever, and the pathway and guideline come into effect when parents or carers decide to access the health services. The figure also illustrates a number of other concepts that were crucial to the guideline development process. More detailed clinical questions evolved from the pathway and the pathway was modified at the end of the development process to incorporate the recommendations derived from the updated clinical questions.

Figure 1.3. Care pathway for feverish illness in children (PDF, 107K)

It was recognised that children with fever may currently be assessed by healthcare professionals who either have or do not have recognised training and/or expertise in the management of children and childhood diseases. In this guideline, professionals with specific training and/or expertise are described as paediatric specialists and they are said to be working in specialist care. Those without specific training and/or expertise are described as non-paediatric practitioners although it is acknowledged that such practitioners may be managing children and their illnesses on a regular basis. Non-paediatric practitioners are said to be working in non-specialist care.

For most children with feverish illness, the initial contact will be in non-specialist care. These contacts will mostly be in primary care but some non-specialist contacts may also be made in secondary care, for example in a general emergency department. A minority of these patients will then be referred on to specialist care, for example in a paediatric assessment unit.

The GDG recognised that assessments of children with feverish illness can take place in three main situations. These are represented by the shaded boxes on the care pathway in Figure 1.3. Broadly, assessments can take place in two ways in non-specialist care. The first is a traditional face-to-face encounter where the child undergoes a full clinical assessment, including history and physical examination. This usually occurs in general practice but it could equally occur in a walk-in centre or a hospital emergency department. Alternatively, the first point of contact could be with what has been described as a remote assessment. This is where the child is assessed by a healthcare professional who is unable to examine the child because the child is geographically remote from the assessor. Remote assessments are becoming increasingly important in the health service and they are used both in and out of normal working hours. Examples include NHS Direct and other telephone advice services. In some circumstances, although the child is not geographically remote from the assessor, it may not fall within the scope of practice for a particular healthcare professional to carry out a physical examination of the child, for example a pharmacist. In these circumstances, the healthcare professional may choose to follow the remote assessment guidance rather than the face-to-face guidance that takes into account signs found on physical examination. In specialist care, the clinical assessment will be undertaken by individuals trained in the care of sick children and the assessment may take place in a paediatric assessment unit, on a children’s ward or in a dedicated paediatric emergency department.

The care pathway demonstrates a number of possible outcomes from each type of encounter with the health services. From a remote assessment, parents and carers will either be advised how to care for their child at home with appropriate advice as to when to seek further attention, or they will be advised to bring the child in for a formal clinical assessment. For the small number of children who have symptoms suggestive of an immediately life-threatening illness, the parents or carers will be advised to take the child for an immediate specialist assessment, for example by calling an ambulance. From a clinical assessment in non-specialist care, a child may again be returned home with appropriate advice. Alternatively, the child may be discharged with a ‘safety net’ that ensures that the child has some kind of clinical review or planned further contact with the health services (see Chapter 6). If the child is considered to be sick or potentially at risk of serious illness, the child will be referred to specialist care. In many cases, a firm diagnosis will be made by the non-paediatric practitioner and the child will be managed and treated accordingly. In these circumstances, the child progresses beyond the scope of this guidance and it is expected that the child would be treated according to relevant national or local guidelines.

In specialist care, a diagnosis may also be made promptly and the child will also leave the remit of this guideline. Some children will be discharged with advice. Others will require immediate admission to the children’s ward and a minority will require intensive care (ITU). This will leave a group of children in whom there is uncertainty as to whether they require admission or not. Increasingly, these children are observed for a number of hours on an assessment unit and then re-evaluated. It is hoped that this practice can help distinguish children with serious illnesses from those with self-limiting conditions.

Literature search strategy

Initial scoping searches were carried out to identify relevant guidelines (local, national, international) produced by other development groups. The reference lists in these guidelines were checked against subsequent searches to identify missing evidence.

Systematic searches to answer the clinical questions formulated and agreed by the GDG were carried out using the following databases via the OVID platform: MEDLINE (1966 onwards), Embase (1980 onwards), Cumulative Index to Nursing and Allied Health Literature (1982 onwards) and PsycINFO (1967 onwards). The most recent search conducted for the three Cochrane databases (Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and the Database of Abstracts of Reviews of Effects) was Quarter 3, 2006. Searches to identify economic studies were undertaken using the above databases and the NHS Economic Evaluations Database (NHS-EED).

Relevant published evidence to inform the guideline development process and answer the clinical questions was identified by systematic search strategies. The clinical questions are shown in the relevant sections. Additionally, stakeholder organisations were invited to submit evidence for consideration by the GDG, provided it was relevant to the clinical questions and of equivalent or better quality than evidence identified by the search strategies. GDG members also contributed evidence under the same conditions.

Search strategies combined relevant controlled vocabulary and natural language in an effort to balance sensitivity and specificity. Both generic and specially developed methodological search filters were used appropriately. Unless advised by the GDG, searches were not date specific.

There was no systematic attempt to search grey literature (conferences, abstracts, theses and unpublished trials). Hand searching of journals not indexed on the databases was not undertaken. Ongoing trials were identified and the principal investigators asked to share their research proposals and outcomes, if available.

Although search strategies were devised for children under the age of 5 years, evidence beyond this age group was considered when no other evidence was available for children under 5 years. Refer to the evidence tables outlining these studies on the accompanying CD-ROM.

Searches were updated and re-run 6–8 weeks before the stakeholder consultation, thereby ensuring that the latest relevant published evidence was included in the database. Any evidence published after this date was not included. For the purposes of updating this guideline, 1 September 2006 should be considered the starting point for searching for new evidence.

Further details of the search strategies, including the methodological filters used, are provided on the accompanying CD-ROM.

Synthesis of clinical effectiveness evidence

The NICE Guidelines Manual was largely abided by. However, because this is a symptom-based guideline with un-established methodology, the methodology used is stated where it was not covered in the NICE Guidelines Manual. Evidence relating to clinical effectiveness was reviewed using established guides^17–24 and classified using the established hierarchical system shown in Table 1.2.²⁴ This system reflects the susceptibility to bias that is inherent in particular study designs.

Table 1.2

Levels of evidence for intervention studies.

The type of clinical question determines the highest level of evidence that may be sought. In assessing the quality of the evidence, each study receives a quality rating coded as ‘++’, ‘+’ or ‘−’. For issues of therapy or treatment, the highest possible evidence level (EL) is a well-conducted systematic review or meta-analysis of randomised controlled trials (RCTs; EL = 1++) or an individual RCT (EL = 1+). Studies of poor quality are rated as ‘−’. Usually, studies rated as ‘−’ should not be used as a basis for making a recommendation, but they can be used to inform recommendations. For issues of prognosis, the highest possible level of evidence is a cohort study (EL = 2) since this is the most appropriate methodology to address prognosis. There are no specific ELs for prognosis and therefore all the prognostic studies were rated according to Table 1.2.

For each clinical question, the highest available level of evidence was selected. Where appropriate, for example if a systematic review, meta-analysis or RCT existed in relation to a question, studies of a weaker design were not included. Where systematic reviews, meta-analyses and RCTs did not exist, other appropriate experimental or observational studies were sought, such as diagnostic studies, which examined the performance of the clinical test if the efficacy of the test was required (see Table 1.3). Where an evaluation of the effectiveness of the test in the clinical management of patients and the outcome of disease was required, evidence from RCTs or cohort studies was used.

Table 1.3

Levels of evidence for studies of the accuracy of diagnostics tests.

The system in Table 1.2 covers studies of treatment effectiveness. However, it is less appropriate for studies reporting diagnostic tests of accuracy. In the absence of a validated hierarchy for this type of test, NICE suggests levels of evidence that take into account the factors likely to affect the validity of these studies (see Table 1.3).

Prognostic studies

A substantial part of the evidence for this guideline was derived from prognostic studies. It is worth noting that there is very limited research on prognostic studies and on methods for assessing their quality. The NICE Guidelines Manual currently contains virtually no advice on how to assess such studies. These limitations were recognised from the outset and the NICE methodology was adapted to account for these deficiencies, as outlined in Table 1.3.

For searching, a highly sensitive evidence-based prognostic study search strategy developed by McMaster University was adopted. Searches for this evidence utilised a prognostic search filter by Wilczynski et al.²⁵ full details of the search strategy are provided on the accompanying CD-ROM.

The search identified 3151 prognostic studies. After filtering double references, 300 different abstracts were screened for inclusion.

Studies were appraised using the checklist for cohort studies in Appendix D of the NICE Guidelines Manual, and the evidence level was allocated using the hierarchy described in Table 1.2. According to this system, the best quality evidence would usually be of evidence level 2 because RCTs are not usually used to address questions of prognosis. Prospective cohort studies are generally the preferred type of study. Lower evidence level studies were included on an individual basis if they contributed information that was not available in the higher evidence level studies but yielded important information to inform the GDG discussions for formulating recommendations.

Delphi consensus

In areas where important clinical questions were identified but no substantial evidence existed, a two-round Delphi consensus method was used to derive recommendations that involved the participation of over 50 clinicians, parents and carers from appropriate stakeholder organisations. The participants rated a series of statements developed by the GDG using a scale of 1–9 (1 being strongly disagree, 9 being strongly agree). Consensus was defined as 75% of ratings falling in the 1–3 or 7–9 categories. Results and comments from each round were discussed by the GDG and final recommendations were made according to predetermined criteria. Full details of the consensus process are presented in Appendix A.

For economic evaluations, no standard system of grading the quality of evidence exists. Economic evaluations that are included in the review have been assessed using a quality assessment checklist based on good practice in decision-analytic modelling.²⁶ Evidence was synthesised qualitatively by summarising the content of identified papers in evidence tables and agreeing brief statements that accurately reflected the evidence. Quantitative synthesis (meta-analysis) was not performed in this guideline due to methodological and statistical heterogeneity of the studies identified.

Summary results and data are presented in the guideline text. More detailed results and data are presented in the accompanying evidence tables. Where possible, dichotomous outcomes are presented as relative risks (RRs) with 95% confidence intervals (CIs), and continuous outcomes are presented as mean differences with 95% CIs or standard deviations (SDs). Moreover, RRs were also calculated as positive predictive values (PPV)/(1 - negative predictive value (NPV)) in diagnoses and prognoses when appropriate.

The quality of cohort studies was appraised based on Appendix B in the NICE Guideline Manual, and Appendix F for diagnostic studies.

Health economics

The aim of the economic input into the guideline was to inform the GDG of potential economic issues relating to fever in children. The health economist helped the GDG by identifying topics within the guideline that might benefit from economic analysis, reviewing the available economic evidence and, where necessary, conducting economic analysis. Where published economic evaluation studies were identified that addressed the economic issues for a clinical question, these are presented alongside the clinical evidence. However, this guideline addressed only assessment and initial management of fever in children. Economic evaluation requires assessment of healthcare resources (costs) alongside health outcomes, preferably quality-adjusted life years (QALYs). Since clinical outcomes of treatment were outside the scope of the guideline, it was anticipated that the economic literature that addressed the guideline questions would be very limited.

Apart from the review of the literature, additional health economic analysis was undertaken for specific questions in the guideline which the GDG identified as requiring economic evaluation. Specifically, health economic analysis was undertaken on the cost of thermometers, and the cost-effectiveness of specific investigations in specialist care (C-reactive protein versus procalcitonin). Additional economic models were developed to assess the impact of changing the pattern of referrals to secondary care but the lack of data prevented any meaningful analysis and conclusions to be drawn from this.

For the analysis that was undertaken, clinical data reported in the guideline were used, and UK cost data were collected. The perspective adopted is the NHS and cost data are reported for 2005/06.

Health economic analysis carried out as part of the guideline development is presented within the relevant clinical chapter, with readers being referred forward to appendices which provide more detailed explanation of methods and results.

Health economic statements are made in the guideline in sections where the use of NHS resources is considered.

Forming recommendations

For each clinical question, the recommendations were derived from the evidence statements presented to the GDG as summaries from the studies reviewed. The link between the evidence statements and recommendation were made explicit in the translation of the evidence statement. The GDG agreed the final recommendation through informal consensus. In the first instance, informal consensus methods were used by the GDG to agree evidence statements and recommendations. Additionally, in areas where important clinical questions were identified but no substantial evidence existed, formal consensus methods were used to identify current best practice (see the section above). Shortly before the consultation period, five to ten key priorities were selected using a nominal group technique for implementation (details available at the NCC-WCH). To avoid giving the impression that higher grade recommendations are of higher priority for implementation, NICE no longer assigns grades to recommendations.

External review

This guideline has been developed in accordance with the NICE guideline development process. This has included giving registered stakeholder organisations the opportunity to comment on the scope of the guideline at the initial stage of development and on the evidence and recommendations at the concluding stage. This involved reviewing by two independent reviewers as part of NICE’s external expert review process for its guidelines. The developers have carefully considered all of the comments during the stage of the consultation by registered stakeholders and expert external reviewers and validation by NICE.

Schedule for updating the guideline

Clinical guidelines commissioned by NICE are published with a review date 4 years from date of publication. Reviewing may begin earlier than 4 years if significant evidence that affects guideline recommendations is identified sooner. The updated guideline will be available within 2 years of the start of the review process.