Background: Febrile encephalopathy, defined as fever, seizures and/or altered consciousness, is a common presentation in children in tropical developing countries. Outcomes range from complete recovery through varying degrees of neurological disability which slowly resolve or remain permanent to death from either the acute illness or complications. Whilst bacterial meningitis accounts for a proportion of children affected, the aetiology in many remains unclear but includes malaria and probably viral encephalitis.
Aim: To understand the aetiology, presentation and outcome of febrile encephalopathy in children in Papua New Guinea.
Methods: Children aged between 1 month and 12 years presenting to Port Moresby General Hospital with febrile encephalopathy were studied prospectively. A detailed history and examination and the following laboratory investigations were undertaken as appropriate: cerebrospinal fluid (CSF) microscopy and bacterial culture, gram stain, measurement of protein and glucose and latex agglutination testing for Haemophilus influenzae, Streptococcus pneumoniae and Neisseria meningitides; Ziehl-Neelsen staining and india ink examination on selected samples; IgM for Japanese encephalitis, dengue, rubella and measles; PCR testing and mycobacterial culture for Mycobacterium tuberculosis. Blood was tested for flavivirus, measles and rubella IgM and IgG.
Results: 149 children were enrolled in the study. 129 had a lumbar puncture and CSF examination; 66 had a normal CSF white cell count. A clinical or laboratory-based diagnosis was possible for 140 children, but a definite pathogen was identifiable for only 55 (37%). The diagnoses included bacterial meningitis in 33 (S. pneumonia 16, H. influenza 13 and N. meningitides 4), tuberculous meningitis (5), probable tuberculous meningitis (18), malaria (10), cryptococcal meningitis (1), flavivirus encephalitis (5), rubella encephalitis (1), hepatic encephalopathy (1) and HIV encephalopathy (1). There were 28 cases of meningitis of unspecified aetiology. Of the five children with IgM-confirmed flavivirus encephalitis, one had dengue serotype 1 and two had Japanese encephalitis. Twenty-five children (including three of the five children with CSF flavivirus IgM) had serological IgG evidence of previous flavivirus infection. A history of multiple convulsions, the presence of neck stiffness and use of the Glasgow coma score (GCS) and TB score chart helped to identify children with bacterial meningitis and an adverse outcome and those with febrile convulsions.
Conclusion: The study confirms the importance of S. pneumonia and H. influenza as major causes of febrile encephalopathy in children in Papua New Guinea. Flaviviruses including Japanese encephalitis are a cause of the febrile encephalopathy syndrome, as is Mycobacterium tuberculosis. All children with febrile encephalopathy should have their GCS and TB scores recorded and should be examined for neck stiffness, and a history of the frequency of convulsions should be recorded. These basic clinical data can help to discriminate aetiology, to guide treatment and monitoring and to identify the children at highest risk of adverse outcome.