Appendix E2 hour time limit for an urgent face-to-face consultation following remote assessment: GDG reasoning and justification in the absence of data to inform a formal economic analysis

Publication Details

E.1. Background

The GDG was asked to produce a guideline to aid healthcare professionals in identifying children with serious bacterial illness (SBI) in an attempt to reduce mortality and morbidity in young children. During the guideline development process, the GDG identified evidence-based symptoms and signs that indicate whether a child has a high risk of having SBI. It also identified symptoms and signs that indicate that a child is at very low risk of SBI and can be looked after at home. Current practice is not evidence based and is variable. It is likely that referral patterns from some healthcare providers will change when the guideline is implemented. It is anticipated that some children who would previously not always have been recognised as needing specialist attention (a very small proportion of children who present with fever) will in the future be referred for consultation with a specialist. Furthermore, a number of children for whom referral is not indicated (the far larger proportion) and who would previously have been referred for consultation or unnecessary investigations, will now not be referred unnecessarily under this new guidance. The focus of the guideline is that the right children should be getting the right treatment at the right time and adverse health outcomes (including death) will therefore be avoided. The GDG noted the evidence that problem-based guidelines with care pathways for children with medical problems reduce invasive investigations, and lead to more appropriate treatment and reduced time spent in accident and emergency (A&E) services.246

E.2. GDG justification of the 2 hour waiting time for an urgent referral

An important feature of this clinical guideline on children with feverish illness is the introduction of a ‘traffic light’ system to identify children with varying degrees of risk of serious illness. The guideline makes clear recommendations on which children are unlikely to require medical attention beyond information and reassurance (children with ‘green’ features) and who can thus be confidently managed at home. The guideline identifies children who require an urgent face-to-face consultation with a healthcare professional (‘red’) and those who may require a face-to-face consultation or require a healthcare ‘safety net’ to be put in place (‘amber’).

Because of the limited information that can be obtained from a remote assessment, the GDG originally recommended that all children with ‘red’ or ‘amber’ features should be referred for urgent face-to-face assessment. The GDG felt it was necessary to make a recommendation on the maximum time a child should have to wait to be first assessed by a healthcare professional if they were classified as requiring an urgent consultation during a remote assessment. The aim of this was to recommend a time frame within which action taken will make a difference to the outcome for the child.

Despite an extensive search of the published and grey literature, no clinical data could be identified to define this limit. The GDG debated the issue among themselves and decided that it was such an important question that wider consensus was required. Accordingly, the question went out as part of the Delphi consultation exercise as agreed in the guideline methods protocol. A high level of agreement was reached for a maximum wait of 2 hours following referral for urgent face-to-face assessment (83% agreement). 2 hours was chosen as one of the time periods for the Delphi exercise because it is an existing Department of Health standard for urgent referrals for out-of-hours health care.247

It was recognised by the GDG that children with one or more ‘amber’ signs included children who may not require an urgent referral. It was agreed to make a recommendation on specific waiting times only for children with ‘red’ features, and to recommend that a child with one or more ‘amber’ features is seen face-to-face by a healthcare professional, but that the timing of the consultation for these children could be carried out within a longer time frame which could be based on the clinical judgement of the person carrying out the initial remote assessment.

The GDG believes that a 2 hour maximum wait for an urgent consultation does not represent an uplift in care and is a cost-effective use of NHS resources. The reasons for this conclusion are outlined here. First, there is audit data to suggest that this is already accepted routine practice for children at a high risk of SBI. Second, the GDG strongly believes that a wait longer than 2 hours could potentially increase mortality and morbidity. Finally, the GDG believes that by using a traffic light system to classify children according to their risk of having a serious illness, health-care professionals will have a clearer indication as to which children do genuinely require an assessment by a healthcare professional within 2 hours. By excluding the children with ‘green’ features and most of the children with ‘amber’ features from this urgent referral group, the GDG believes the number of children who are referred for a face-to-face assessment by a healthcare professional within 2 hours will be reduced.

Evidence was presented to the GDG to show that the Department of Health has already set a national standard for response to urgent calls as part of the National Quality Requirements in the Delivery of Out-of-Hours Services.247 This specifies a maximum 2 hour wait for a face-to face urgent consultation for out-of-hours care: ‘Face-to-face consultations (whether in a centre or in the patient’s place of residence) must be started within the following timescales, after the definitive clinical assessment has been completed:

  • Emergency: Within 1 hour.
  • Urgent: Within 2 hours.
  • Less urgent: Within 6 hours’.

Further evidence was presented from NHS Direct that, in line with the out-of-hours Quality Requirements, currently recommends a time frame of less than 2 hours for a child requiring an urgent face-to-face assessment. Audit data from NHS Direct was presented to the GDG to show that, of those who contact NHS Direct via the 0845 telephone number, 31.8% of children under 5 years with a primary diagnosis of fever were referred on for an urgent face-to-face clinical assessment within 2 hours, following detailed nurse assessment (Figure E.1). Also, 47% of out-of-hours calls for the same patient group were referred for a face-to-face clinical assessment within 2 hours. (It is important to note that during the course of these assessments a focus for the fever may be identified which in itself justified the referral within this time period.)

Figure E.1. NHS Direct audit data covering the period 1 January 2006 to 31 December 2006; this data equates to a coverage of the whole of the population of England for the 0845 46 47 calls and a population coverage of 708,500 for the out-of-hours calls.

Figure E.1

NHS Direct audit data covering the period 1 January 2006 to 31 December 2006; this data equates to a coverage of the whole of the population of England for the 0845 46 47 calls and a population coverage of 708,500 for the out-of-hours calls.

One stakeholder comment suggested that a 2 hour time limit for an urgent referral would be very difficult to implement in an A&E care setting where the 4 hour waiting time directive is the current target for the NHS. The guideline is clear that primary care should continue to be the first point of contact for a child with fever (as validated by the NHS Direct data presented here showing that children with fever are referred to the GP within 2 hours, 6 hours or for a routine appointment). The GDG clarified that the new recommendation means that a child with ‘red’ features should be offered an initial assessment (for example, by an A&E triage nurse) within 2 hours, and that the current target of 4 hours for A&E is the time limit for initial assessment, treatment and discharge. The promise to patients derived from the NHS Plan in 2000 set out in Your Guide to the NHS stated that, on arrival in A&E, ‘you should be assessed by a nurse or doctor, depending on how urgent your case is, within 15 minutes of arrival ...’.248

These two waiting time targets are therefore compatible and in keeping with the Department of Health NHS Plan and Quality Requirements. Other stakeholders who commented on the 2 hour time frame felt that it was too long a wait for children requiring an urgent referral.

The GDG believes that, if the traffic light system is adhered to, the recommendation for a 2 hour urgent referral will apply to a smaller but more relevant proportion of children with fever than are currently referred for an urgent assessment. A GDG member who is a GP presented evidence to the GDG from a survey of children presenting with fever as their predominant symptom and the prevalence of ‘amber’ features in this patient group. The practice has 9518 patients, with 633 children aged 5 years and under.

There were 157 consultations in this age group, involving 77 children with 83 episodes of acute fever with no other symptoms that worried the parent. Fifty-three episodes were telephone triage, and in 24 of these cases a face-to-face consultation was advised (45.2%). In thirteen of these cases, an ‘amber’ symptom was noted. The rest (104) were all face-to-face consultations without telephone triage, and in 18 consultations, ‘amber’ symptoms were recorded, with a diagnosis made in nine cases. Six of these children were referred for a paediatric assessment unit for specialist advice, which represents 3.8% of children presenting with fever as their primary symptom. During the period of the survey, there were no children who would have been classified as ‘red’ under the traffic light system.

Only 13 of those assessed remotely and 18 of those assessed face-to-face showed ‘amber’ features, and thus potentially none of these children fell into the urgent referral group. The absence of either ‘red’ or ‘amber’ features would have allowed at least some of these children to be confidently managed at home, and those with ‘amber’ features only could have been referred within a longer time frame of safety netting, which could have been put into place following face-to-face assessment. The data suggests that the proportion of children who require an urgent face-to-face referral following remote assessment would potentially be reduced and is very small compared with the far greater number of children who have either ‘amber’ symptoms and require assessment within a longer time frame by a healthcare professional or have self-limiting illness (who can be confidently managed at home).

Having reviewed the data and based on their own experience, the GDG consensus was that an individual GP in a group practice such as the one surveyed would be unlikely to see more than one or two cases of SBI a year, and for some of the more rare conditions would be unlikely to see one case in their professional career. During the period of the survey there were no children who would have been classified as ‘red’ under the traffic light system. This is because urgent referrals would only be needed for children with ‘red’ features and a proportion of children with ‘amber’ features. This assertion is supported by the data in the GP survey referred to above where no children were classified as ‘red’ and 19% were ‘amber’.

Further evidence of the number of children likely to present to secondary care with ‘red’ symptoms was considered. An American study of 6611 febrile children presenting to an emergency department found that 3.3% of children had a Yale Observation Score greater than 10.101 A YOS score of 10 means the child has symptoms that are ‘red’ signs and symptoms on the proposed traffic light system. It is important to note that the 3.3% is a small fraction of the total number of children with fever but it still may be an overestimate because the data do not indicate how many of the 3.3% of children with a YOS score over 10 have other symptoms which are ‘red’ features in the traffic light system. Also, the study was done in a hospital setting and it is based on the American healthcare system. Furthermore, the GDG’s recommendation would only apply to children referred from remote assessment in this context and not all children with ‘red’ symptoms, many of whom will present for a face-to-face clinical assessment as their first point of healthcare contact.

E.3. Cost-effectiveness of a 2 hour referral for face-to-face assessment

The GDG did not identify any data on the likely cost or cost savings from recommending a 2 hour time limit for an urgent face-to-face assessment or the likelihood of this leading to an increase in referrals to specialist care. The issue was discussed in detail during a number of GDG meetings. The main point that was agreed was that the GDG believes that the guideline’s recommendations will support the identification of those children requiring urgent assessment, referral and initiation of management which in some cases will be life-saving and certainly prevent unnecessary long-term morbidity. There is a cost-effectiveness threshold under which any intervention that saves lives or prevents serious morbidity is generally seen to be cost-effective. If we assume that a life-saving intervention that prevents one death in a very young child is worth around 25 QALYs (75 years discounted at 3.5%), then an intervention that costs £500,000 (25 × £20,000) and saves one life is within the threshold for cost-effectiveness.

The GDG found it impossible to guess how many children with ‘red’ symptoms who were seen face-to-face urgently from a remote assessment (within 2 hours) would be saved from death or serious morbidity. The argument for cost-effectiveness is that £500,000 (to save one child’s life) could be spent on additional face-to-face assessments for it to be cost-effective if it saved one life. The cost of additional face-to-face assessment is hard to estimate if it is within surgery hours, but it costs around £35–40 for an out-of-hours consultation*249 or £70 for a home visit.250 Therefore if an additional 7,100 (£500,000/£70) patients could be seen for face-to-face assessment, this would be cost-effective if it saved one additional child’s life.

This does not take into account the potential savings from preventing the health and social care costs of serious morbidity in children which would make the intervention more cost-effective. Nor does it take into account that the carers of children with ‘red’ symptoms will contact health services somehow, and the guideline emphasises the fact that this should almost always be primary care in the first instance. This is a less expensive option than A&E services which cost £77–105 per visit for 2005/06, depending on the cost of investigations.250

This very brief analysis of cost-effectiveness assumes that at least three children’s deaths are prevented every year in the district general hospital by putting in place a 2 hour assessment in a population of 250,000, and there are children are currently at risk of death and serious morbidity who are not currently being urgently assessed and referred for specialist advice. It also assumes that all children at risk of death from SBI are seen eventually by a healthcare professional, and do not die at home without any health service contact. It is assumed that deaths can be prevented by more timely referral to specialist services for those children who urgently need it, and that the cost of investigations and initial management once reaching a specialist care unit would be the same at whatever stage they were referred (that is, a standard package of investigations and management of a child with suspected SBI would be initiated).

Clearly there are costs around diagnosis and initial management of a child with suspected SBI once they reach specialist services, but the GDG was not clear that these would be any different (whether higher costs if a child is referred urgently or higher if referred after a delay of more than 2 hours). Without empirical data, these assumptions cannot be verified, but the GDG members believe that these are conservative assumptions that reflect the real world closely enough to make the assertion that the 2 hour face-to-face referral is very likely to be cost-effective.

E.4. Conclusion

The aim of this guideline is to improve the identification of those children who are genuinely at a high risk of serious illness and require urgent assessment and treatment to prevent death and serious morbidity. Using the traffic light system, those children in the ‘red’ category have been identified as being at a high risk of serious illness and the GDG believes that it is already established best clinical and cost-effective practice for this small group to be seen urgently within 2 hours and this guidance will reinforce that practice. The guideline will also reduce unnecessary assessment (urgent and routine) and diagnostic testing of children who are at low risk of serious illness.



Annual cost or provision of out-of-hours care in England was £316 million in 2004–05, and the number of people using the service in England was 9 million.