Objective To investigate how infants and children with suspected viral encephalitis are currently managed in a UK tertiary children’s hospital.
Methods Case notes of all infants and children who received intravenous aciclovir for suspected encephalitis over a 6-month period were reviewed. Suspected viral encephalitis was defined as a child with fever or history of febrile illness and a reduced level of consciousness, irritability or a change in personality or behaviour or focal neurological signs.
Results Fifty one children were identified. Two had proven herpes simplex encephalitis (HSV) and two had clinically diagnosed viral encephalitis with no cause identified. Forty children had cerebrospinal fluid (CSF) analysis, but basic results were incomplete in 13 cases. CSF was sent for the detection of HSV DNA by PCR in 27 cases. The initial dose of aciclovir was incorrect in 38 cases. The median (range) length of intravenous aciclovir treatment was 4 (1–21) days. Six children were given a full course of aciclovir (10 or more days). For 14 children, there appeared to be no real indication for starting aciclovir. Case note documentation was generally inadequate.
Conclusions The management of children with suspected viral encephalitis appears haphazard in many cases. Guidelines for the management of children with suspected viral encephalitis are needed.
Statistics from Altmetric.com
Fever, seizures, irritability or reduced level of consciousness are common childhood symptoms and signs. Children with these features may have a central nervous system (CNS) infection including viral encephalitis. The commonest diagnosed cause of encephalitis in the United Kingdom (UK) is herpes simplex virus (HSV) type 1.1 Untreated, the mortality of HSV encephalitis is approximately 70%. This can be reduced significantly to 20% if intravenous aciclovir is given.2 3 Because it usually takes several days to confirm the diagnosis, aciclovir treatment is started if there is a strong suspicion of HSV encephalitis; this will usually be based on appropriate clinical features and cerebrospinal fluid (CSF) or imaging findings consistent with viral encephalitis.4 Earlier treatment is associated with a better outcome.5 However, initiating treatment too early, without appropriate investigations, or in patients for whom there is no clear indication, can make the subsequent management difficult. The standard treatment for HSV encephalitis is now 14–21 days of intravenous aciclovir, because of apparent relapses following shorter courses.6,–,8 But maintaining this treatment course in patients in whom it was never proven whether or not they had HSV encephalitis is difficult, leading to tough decisions about when to stop the treatment.4 The aims of this study were to investigate how children admitted with suspected encephalitis were managed, by looking at their investigations and treatment.
What is already known about this topic
Children with proven herpes simplex encephalitis or clinically diagnosed viral encephalitis require at least 10 days of treatment with intravenous aciclovir.
The results from cerebrospinal fluid analysis are very helpful in the management of children with suspected central nervous system infection including viral encephalitis.
What this study adds
Children with suspected viral encephalitis are often inadequately investigated and managed.
Guidelines for the management of children with suspected viral encephalitis are needed.
The study was conducted at the Alder Hey Children’s NHS Foundation Trust, a large secondary and tertiary paediatric centre serving a population of approximately 3.5 million people, with more than 60 000 new attendances to the emergency department every year. We retrospectively studied all children admitted over a 6-month period (February–November 2005) who were treated for suspected viral encephalitis. To identify such children we obtained from pharmacy records all children prescribed intravenous aciclovir; after reviewing the case notes we excluded those who had localised skin infections or immunocompromised patients (mainly oncology cases) treated for cutaneous varicella zoster (chickenpox) infection. Details of the presenting clinical features, and the relevant investigations and treatments were obtained from the case notes.
Children were considered to have “suspected viral encephalitis” (and therefore have an indication for treatment with intravenous aciclovir) if they had a fever or history of febrile illness and a reduced level of consciousness, irritability or a change in personality or behaviour or focal neurological signs.9 Children were subsequently classified as having a diagnosis of “proven viral encephalitis” if a viral cause was identified, or “clinically diagnosed encephalitis” if their clinical picture, CSF findings and other investigations were consistent with viral encephalitis, but no virus was identified. The initial management of children with seizures or a febrile illness or both was assessed for whether NICE guidelines were followed.10 11 We considered the dose of intravenous aciclovir given to be correct if it agreed with the doses quoted in the British National Formulary for Children: 20 mg/kg three times a day for infants aged 1–3 months; 500 mg/m2for children aged 3 months to 12 years; 10 mg/kg three times a day for children of 12 years or over. The study was registered with the Trust audit department and all data were handled according to national guidance.
Fifty six children receiving intravenous aciclovir were identified of whom 51 were included (five had received the drug for other indications). Twenty five (49%) children were admitted via the emergency department and 26 (51%) were transferred to the intensive care department from local district general hospitals.
The clinical features at presentation are summarised in table 1, and initial diagnoses, investigations, results and final diagnoses are given for each patient in table 2. One child was immunocompromised through treatment for acute lymphoblastic leukaemia. Thirty one children had a history of seizures or had witnessed seizures (19 generalised tonic–clonic, 12 complex partial). Fourteen children had new focal neurological signs: four had increased tone and reflexes, four had pupil abnormalities, two had facial weakness, two had abnormal posturing, one had dysarthric speech and one had a lateral rectus palsy. Documentation of symptoms and signs was inadequate in many cases (table 1), for example, only two thirds had the presence or absence of a rash noted, less than two thirds had a coma score recorded and only one half had the presence or absence of neck stiffness documented. Thirty seven children (72%) met the case definition for suspected viral encephalitis, or had aciclovir started in accordance with NICE febrile illness guidelines. But for 14 children there appeared to be little rationale for starting aciclovir. Initial diagnoses in these children included: new onset afebrile seizure(s) in four children, sepsis/meningitis in infants less than 3 months old (four children), afebrile status epilepticus in children with epilepsy and a pre-existing severe neurological disorder (two children), acute cerebellitis (one child), malfunction of a ventriculo-peritoneal shunt (one child), afebrile seizures following head injury (one child), and deliberate drug overdose (one child).
Seventeen children had a lumbar puncture and two had CSF sampled from their ventriculo-peritoneal shunt on the day of admission. Twenty had a lumbar puncture subsequently and one had CSF sampled during surgery for a temporal lobe abscess, but 11 never had a lumbar puncture. The median (range) day of illness CSF was analysed was 3 (1–21). Fifteen children had a CSF pleocytosis. CSF analysis was incomplete in 13 cases: in three this was because CSF protein or glucose was not sent and in 10 because no paired plasma glucose was sent. All children who had CSF analysis undertaken had bacterial culture performed and three (8%) were positive (two Mycobacterium tuberculosis, one Streptococcus pneumoniae). However, only 27 (53%) children had CSF PCR sent for HSV DNA (in seven this was before aciclovir was started); one patient was PCR positive (HSV type 1). Lumbar puncture results confirmed the clinical diagnosis in five (10%) children and changed management in 31 (79%) cases, for example, by allowing discontinuation of aciclovir and/or antibiotics. One neonate, with blistering lesions on the scalp, had vesicle fluid positive for HSV type 2 by tissue culture and electron microscopy and had both a CSF picture and MRI findings consistent with encephalitis, although HSV PCR was not requested on the CSF. Seven children (14%) had bacterial meningitis proven microbiologically by CSF or blood culture, and one child had a pneumococcal brain abscess.
In 38 children (74%), the initial dose of aciclovir prescribed was incorrect; in 20 it was sufficiently incorrect to require recalculation. Surface area was used for dose calculation in 17 of the 32 children (53%) in whom it was indicated. Renal function was tested in 47 (92%) of the children on admission. No children developed renal impairment during their admission. The median (range) length of intravenous aciclovir treatment was 4 (1–21) days. Six children were given a full course of aciclovir (10 or more days). Three received 21 days’ aciclovir – one intravenously and two completed their course orally (after 18 and 19 days of intravenous treatment, respectively). No children were given oral valaciclovir. The reasons for discontinuing aciclovir before 10 days were documented in 15 (29%) cases: negative HSV PCR in CSF (seven patients), rapid clinical improvement (five patients), loss of venous access and clinical improvement or alternative diagnosis suspected (two patients) and tuberculous meningitis confirmed (one patient).1
Antibiotics were prescribed in 50 (98%) cases, with cefotaxime given in 45 cases. The median (range) length of antibiotic treatment course was 7 (1–14) days. The only child who did not receive antibiotics had a diagnosis of acute para- or post-infectious cerebellitis. There were no deaths and in the children with clinically diagnosed or proven encephalitis no clinical relapses occurred.
Viral encephalitis is an uncommon, but serious and sometimes fatal, disorder. The incidence in the UK is unknown, and there are many possible causes, but the commonest proven cause is HSV type 1, which is treatable with the intravenous antiviral drug aciclovir.2 Although proven HSV encephalitis is rare, children presenting with features consistent with suspected viral encephalitis are not, and so presumptive treatment is often started. In our study, over a 6-month period 51 children were treated with aciclovir, for four of whom viral encephalitis was finally diagnosed clinically, including two with an identified viral cause – one HSV1 and one HSV2. Our study was limited by its retrospective nature, and the fact that it would have missed children with suspected encephalitis not given aciclovir, although the existence of such children seems unlikely given the readiness with which the drug appears to be given.
Nevertheless there are some clear messages from our data. For 14 (27%) of the 51 children treated with aciclovir, there was no real indication for this treatment in the first place: seven presented with a single or multiple seizures which were not associated with a febrile illness, and which are thus not an indication for aciclovir10 – indeed two of these children were already known to have epilepsy, and one had an acute head injury. However of those that presented with seizures, one that subsequently developed diarrhoea ultimately proved to have a rotavirus encephalopathy, and another had a cerebral abscess, so it might be argued aciclovir treatment was justified in these patients. The other children treated with aciclovir but with no clear indication included one with a problematic ventriculoperitoneal shunt, one who had taken a drug overdose, and four young infants with clinical features of sepsis; although antibiotics are indicated in such children, there are no guidelines or suggesting aciclovir should be used.11 12 Notably, investigations, including lumbar puncture, were incomplete in five of these 14 children. Interestingly, in almost all the cases where aciclovir was given without a good indication, the initial diagnosis considered by the clinician (which was not encephalitis) proved to be correct.
The results confirm our suspicion that in recent years a tendency has developed among many paediatricians to start all children who are acutely unwell and who may have a CNS infection on intravenous aciclovir as well as broad spectrum antibiotics, even though there is no evidence base for the aciclovir usage. The reasons for this trend are unclear, but may be because the well publicised guidelines for the management of meningococcal meningitis and septicaemia in children13 have been inappropriately extended to include giving aciclovir to all children with any suspected CNS infections. Although aciclovir is generally considered to be a safe drug, serious side effects are reported including renal impairment (especially when given with ceftriaxone), hepatitis and bone marrow failure.14 15 A more common problem with the overuse of aciclovir is the decision about when to stop treatment, especially in the child who appears to be recovering, and who has not been investigated fully with lumbar puncture. Prolonged treatment and the associated lengthy hospital stay adds to the economic burden on the NHS.16 In most cases a complete course of aciclovir was not given, but the rationale for stopping was not explained in the notes. Given that HSV encephalitis can occasionally relapse, the reason for stopping the drug should be clearly documented. In some patients treatment was changed from intravenous to oral aciclovir, presumably because of problems with venous access. Our practice, should this be necessary, is to give at least 10 days of intravenous treatment if possible and then use oral valaciclovir, because of its better bioavailability than oral aciclovir.17
Our results suggest that in this group of children, aciclovir was given too readily without proper consideration of the most likely diagnosis and without undertaking the necessary investigations. This is in contrast to a recent report of suspected encephalitis in adults in the USA where there were inappropriate delays before starting treatment.18 An earlier report from the UK suggested that guidelines would be helpful.16
Previously, we showed that only 53% of children with suspected CNS infection had a lumbar puncture.19 Following this, lumbar puncture guidelines were published on our hospital intranet, and there has been regular teaching of junior doctors. In the current series, 37 children (72%) with suspected encephalitis had a lumbar puncture, which appears to indicate an improvement. However, 32 children did not have an immediate lumbar puncture, although in most cases there was no contraindication to it; 20 of these had an lumbar puncture subsequently, but in 11 the investigation was not performed at all. Even a week after starting treatment lumbar puncture can be useful, because HSV DNA may be detectable up to this time.20 Surprisingly, in our study, even when a lumbar puncture was undertaken, CSF HSV PCR was not always requested, despite the child being treated for suspected HSV encephalitis. If encephalitis is suspected strongly enough on clinical grounds and CSF analysis findings and a decision is made to start aciclovir, then HSV PCR should always be requested on the CSF.
If the initial CSF results are normal, the diagnosis of a CNS infection (and in particular encephalitis) is unlikely and it is possible that a decision to start aciclovir could be considered but deferred as long as the child continues to be observed in hospital. Although HSV encephalitis has occasionally been reported in patients without a CSF pleocytosis, this is rare and tends to occur on only the first or second day of illness with a repeat lumbar puncture usually revealing cells.21 22 This does not appear to be current practice; we found that 18 of the 40 children who had CSF analysis had a normal CSF, and therefore may not have needed aciclovir treatment at all, if the treatment had been guided by the CSF findings. For these children, a prolonged course of intravenous aciclovir seems unnecessary. We would recommend treating all children in whom encephalitis is a strong clinical possibility with intravenous aciclovir, even if the initial CSF HSV PCR is negative, for a minimum of 10 days. In this situation, the lumbar puncture should be repeated and CSF HSV PCR sent again. However, if the child has made a full clinical recovery and CSF analysis and brain imaging are normal or if a definite alternative diagnosis has been reached, then aciclovir can be discontinued providing a clear rationale has been documented in the case notes. If the child subsequently shows any sign of a possible relapse, then appropriate investigations should be repeated, including a lumbar puncture and brain imaging, and treatment restarted.
We have shown that the management of children with suspected encephalitis appears to be haphazard. In almost a third of patients, aciclovir was started without a rational basis and the decision to start treatment was frequently made without using appropriate investigations. Even when lumbar punctures are performed, HSV PCR is not always requested. Aciclovir is often given at the wrong dosage, and the drug subsequently stopped without explanation. Guidelines on the initial management of suspected encephalitis in children are needed.
We thank Mrs Pat McCarrick for case note retrieval.
Competing interests None.
Funding TS is an Medical Research Council senior fellow.
Provenance and peer review Not commissioned; externally peer reviewed.
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.