A CASE-CONTROL STUDY OF THE PREVALENCE AND PRESENTATION OF LASSA FEVER IN FEVER ASSOCIATED WITH CONVULSIONS IN CHILDREN, IN IRRUA

Abstract

Convulsions associated with fever ( CAWF ) are frequent medical emergencies in children. In the
tropics, malaria and bacterial infections are well established causes of CAWF with their role and
contributions extensively researched. However, few reports, if any, are available of the role of
endemic viral illnesses such as Lassa fever ( LF ) in this neurological emergency. It was
hypothesised that LF may be a yet unrecognised significant contributor to CAWF in children
resident in areas endemic for the disease.

The study aim was to determine the prevalence of LF in children with CAWF and to compare the
characteristics of convulsions, clinical features and outcomes in these patients with those of non
LF cases of CAWF and other febrile controls.
In a prospective case-control study of emergency room admissions, consecutive febrile children
≥ 1 month to ≤ 15 years were recruited from 1st of December 2009 to 30th November 2010. At
presentation all patients admitted with a history of fever and a temperature of ≥ 380 C were
included as subjects. Patients who refused consent or failed to meet the inclusion criteria were
excluded. Evaluation of the children was by history and physical examination, followed by Lassa
virus reverse-transcription polymerase chain reaction (LV-RT-PCR) test. Other laboratory
investigations done included, blood film examination for malaria parasites, PCV, WBC and CSF
analysis.
A total of 913 children were seen in the Children’s Emergency Room of ISTH during the study
period. Three hundred and seventy three ( 373 ) febrile children were recruited. These included
108 children with CAWF ( Cases )and 265 children with fever but no convulsions ( Controls ). The
aetiology of fever among the Cases vs Controls respectively were LF ( 3 vs 1 ), malaria ( 47 vs 36
), focal extra cranial infections ( 12 vs 56 ), meningitis( 10 vs 10 ), miscellaneous ( 9 vs 33 ) and
undetermined ( 27 vs 120 ).
The prevalence [ 95% Confidence Interval ( 95% CI ) ] of LF among the cases was 2.8 % [ 0.7%,
7.4%] while the prevalence [ 95% CI ] among the controls was 3.8% [ 1.9%, 6.6% ] . The difference
was not statistically significant ( p = 0.903 ). There was also, no significant difference between the
prevalence of LF in Cases with febrile convulsions and Cases with convulsions associated with
fever outside the age group associated with febrile convulsions compared to their age matched
controls. Bleeding was significantly ( Fisher’s exact p = 0.008 ) associated with LF among febrile
children. No organism was isolated on CSF, blood or urine culture. No patient with LF had CSF
evidence of meningitis.
It is concluded that LF is an important cause of CAWF in endemic areas, although the prevalence
is low ( 2.8% ) and is similar to the 3.8% prevalence in children with fever but no convulsions. It
is also concluded that although malaria ( with a prevalence of 43.5% ) is the commonest cause of
CAWF without localizing signs, meningitis with a prevalence of 9.3% is also an important problem.
It is further concluded that the presence of bleeding is predictive of LF in febrile children.
It is recommended that LF be considered in children presenting with fever, convulsions and
bleeding in LF endemic areas. It is also recommended that whereas the presumptive treatment
of malaria in children with CAWF without localizing signs may be justified, the presence of malaria
parasitaemia does not rule out LF.

A CASE-CONTROL STUDY OF THE PREVALENCE AND PRESENTATION OF LASSA FEVER IN FEVER ASSOCIATED WITH CONVULSIONS IN CHILDREN, IN IRRUA