Several methods have been employed in the diagnosis of LF. Complement Fixation (CF), Indirect Immuno-Fluorescent Antibody Test (IFAT) and Enzyme Linked Immuno-Sorbent Assay (ELISA) and Reverse ELISA for IgG and IgM antibodies are all methods with which the diagnosis has been made.106, 107 Others include PCR, viral isolation and virus detection by Tissue culture antigen.110 Amongst these methods however, virus isolation is the accepted gold standard. 22 Viral isolation from blood or serum is possible for up to 2 weeks post onset of illness. 13 Isolation of the virus in blood, urine or throat washings, demonstration of IgM antibodies in patient’s blood, or a 4 fold rise in IgG titre between acute and convalescent phase sera are all
recognised criteria for diagnosis.22,90
42
In Nigeria, serologic tests have documented LF in various populations.52 However, Serologic tests carry a drawback such that confirmation is frequently required using viral isolation.22 Moreover, viral isolation is risky and expensive, requiring a Bio-safety level 4 laboratory 89 ; an expense beyond the reach of most developing countries.
Although IFAT detection of LF has also been documented in West Africa, 90 ELISA and reverse ELISA have demonstrated higher levels of sensitivity, specificity and capability for early diagnosis compared with IFAT. 90 Both IFAT and ELISA have problems in detecting low amounts of
specific antibodies. 90 IFAT is also technically challenging even for well trained personnel. 90
The early appearance of antibodies in LF is a rationale for the use of antibody detection tests.108 However, only approximately 50% of patients develop LF antibodies in the first week of illness90 and it is evident that early patient isolation may be missed with reliance on antibody tests alone.
Use of LV antigen and IgM ELISA testing together have a 95% sensitivity and 90%
specificity, but result interpretation may be difficult. Other methods such as Tissue biopsy for viral antigen isolation may be difficult, except at autopsy, because of bleeding tendencies associated with the disease.69,107
Reverse Transcription Polymerase Chain Reaction (RT-PCR) methods for detecting LV RNA have been described and thus rapidly emerged as the dominant method of diagnosis.21,68,69,109,110 It’s demonstrated ease, accuracy and successful field use in other emergent infectious viral illnesses such as HIV encouraged the testing of it’s suitability in LF diagnosis.PCR has become a powerful tool for diagnosis with proven reliability, safety and sensitivity for the rapid diagnosis of LF. 110
However, strain variation is a potential source of error and produces false negative results in PCR diagnosis. 110 This is because LV comprises of a highly genetically diverse group of strains and limited strain specificity of primer pairs used in PCR may thus affect diagnosis.
69 In spite of these, the use of RT-PCR in rapid diagnosis of clinical cases of LF has become well
established.111
43
OUTCOME OF CONVULSIONS ASSOCIATED WITH FEVER
Febrile convulsions (FC) are generally innocuous but may be associated with significant morbidity and mortality. Sequelae of febrile seizures may include recurrence of febrile
convulsions, 10,28,112 anoxic brain damage following repeated prolonged convulsive episodes10,30, aspiration of vomitus and/or traditional medications 10 and death. 10 In addition,
choreoathetosis, cortical blindness, and severe spasticity are neurologic sequelae documented
in Nigerian children following febrile convulsions.10, 31, 39
Recurrent febrile convulsions are the most frequent sequelae following an initial febrile seizure.112 However, the risk of recurrent afebrile convulsions and the risk of neurological sequelae, also exists. 112 Early age at onset,a family history of afebrile seizures, complex febrile seizures, abnormal neurological development, and a family history of epilepsy are factors that
may increase the risk of epilepsy.28, 112
The prognosis of febrile convulsions in temperate countries 28, 112 is in sharp contrast with the prognosis in the tropics, where febrile convulsions have significant morbidity and mortality.10,16,31 Febrile convulsions are significant contributors to emergency room deaths in the tropics. Familusi and Sinnette reported a mortality of 6.4% at Ibadan.10 Febrile convulsions account for 3.5% in Benin 31, 2.5% in Ilorin 5 and 1.8% in Jos 16 of emergency room deaths.
Contributory factors to the relatively poor prognosis of febrile convulsion in the tropics are
44
multifactorial. A major factor is administration of traditional medicinal therapy.5, 10 In Ibadan, 28 (72%) of the 39 deaths out of 590 children had traditional concoctions administered.5
Similarly the deaths in the studies at Benin 31, Ilorin 5, and Jos 16 were related to administration of traditional concoctions, especially that containing cow urine concoction (CUC). CUC, commonly used in the south-western region of Nigeria, has been reported to cause
hypoglycaemia and cerebral oedema 10, 113, 114 and may be associated with prolonged seizures 5, 10, 36
Other remedies like, burning of feet and scarification marks may worsen the prognosis following febrile convulsions. Cerebral malaria, bacteraemia, and perhaps a missed diagnosis of meningitis may also be contributory factors.
JUSTIFICATION OF THE STUDY
CAWF has been studied extensively.5, 10, 28, 30, 36, 93 While studies from temperate countries 28, 30 noted that viral illnesses accounted for the majority of cases, reports from the tropics implicate malaria as the main cause, 5, 10, 16, 31, 39, 93 However, there is little information on the contribution of LF.
The majority of the studies on the prevalence and clinical features of LF are from other West African countries namely Sierra Leone, Liberia and Guinea. 22, 78, 90 In addition, although convulsions have been described in LF as a possible marker for disease severity,
45
the true prevalence of LF in childhood CAWF has not been determined, neither have the determinants of convulsions in LF.
LF presentations, especially the contribution to aetiology of convulsions and pyrexia, amongst paediatric populations in Nigeria, require to be studied. Undertaking this study in Irrua, an area endemic for LF, should add to our understanding of the manifestations of this emerging illness with respect to CAWF.
For an illness which presents with non specific features mimicking common infections, but is associated with a high mortality rate, it is important to determine the prevalence and clinical features in children. The clinical features thus identified could aid clinical diagnosis and intervention, especially in resource poor settings as appropriate laboratory support is frequently unavailable.
AIMS AND OBJECTIVES Broad Objective
To determine the prevalence and clinical presentation of Lassa fever in children ≥ 1 month to
≤15 years old with convulsion associated with fever in a Lassa fever endemic area.
Specific Objectives
1. To determine the prevalence of Lassa fever in children with convulsion associated with fever.
2. To determine the characteristics of convulsions in children with Lassa fever compared to
those without Lassa fever.
46
3. To compare the clinical features of children with Lassa fever at presentation with those of children with other common infections.
4. To determine the outcome in children with Lassa fever presenting with convulsion associated with fever.
PATIENTS AND METHODS
STUDY SETTING
The study was carried out in the Children’s Emergency Room (CHER), ISTH, Irrua, Edo State from the 1st of December 2009 to the 31st of November 2010. The CHER is a 15 bed facility.
The Hospital serves as a referral centre for Edo, Delta, Kogi and Ondo states.
Irrua is the head-quarters of Esan-Central Local Government Area of Edo state. The population is predominantly sub-urban and rural with Esan as the predominant ethnic group, but there are also other Nigerians in the area. Lassa fever is endemic in the area and due to cultural practices like bush burning, it is thought that the rodent reservoir may be displaced from it’s natural habitat to dwell in households thus increasing the likelihood of human infection.
47 STUDY DESIGN
This was a prospective case-control study of consecutive febrile children with convulsions presenting to the CHER, ISTH. Blood, urine and CSF samples for parasitologic, bacteriologic and clinical chemistry determinations were processed at the respective service laboratories of the hospital. LV-RT-PCR was carried out at the Research and Diagnostic Laboratory of the Institute for Lassa Fever Research and Control (ILFRC), ISTH.
SAMPLE SIZE DETERMINATION
A retrospective pilot study of febrile children admitted at ISTH between 1st September 2008 and 31st August 2009 was undertaken prior to commencement of the study. This was pertinent, to make up for the lack of data on Lassa fever in children with CAWF, as the study aimed at determining the prevalence and pattern of clinical presentation of LF in children with CAWF.
Amongst 97 children admitted with a febrile illness and in an age range of ≥ 1 month to ≤ 15 years, 66/97 (68%) had CAWF. Overall, LF was suspected in 13/97 (13.4%) and confirmed in 6/97 (6.2%). LF was confirmed in 3 (4.5%) of the 66 children with CAWF and in 3 (9.7%) of the 31 without CAWF (fisher’s exact p=0.577).
Since this study aimed at determining both the prevalence and presentation of LF, a sample size optimal for both the prevalence as well as case-control aspects of the study in which all subjects were exposed to the same factors was required.
A. SAMPLE SIZE FOR A CROSS SECTIONAL (PREVALENCE) STUDY