A study was undertaken to determine the relationship of presenting features and outcome in primary childhood meningitis at the Paediatric Department, University Hospital, Kuala Lumpur from January 1980 to December 1989. A total of 177 cases of primary childhood meningitis admitted during the study period were analysed retrospectively. One hundred and nine cases (61.6%) had complete recovery, 49 cases (27.7%) recovered with residual complications and 19 cases (10.7%) died. The mean duration of illness of those with complete recovery, residual complications and acute mortality was 4.8 days, 9.6 days and 7.6 days respectively (p < 0.05). Neurological signs were more common in those with mortality and morbidity i.e. 89% and 78% respectively as compared to those who had complete recovery (58%). There was significant correlation between outcome and haemoglobin level on admission (complete recovery 11 gm%, residual complications 10.4 gm%, mortality 9.6 gm% with p < 0.05), mean CSF sugar content (complete recovery 2.2 mmol/l, residual complications 1.5 mmol/l, mortality l.1 mmol/l, p < 0.00001), and mean CSF protein level (complete recovery 168 mg%, residual complications 321 mg%, mortality 344 mg%, p < 0.001). Gram positive organisms were associated with higher mortality and morbidity i.e. 24% and 40%, as compared to gram negative organisms 12% and 31% respectively. There was no significant correlation' between age, sex, peripheral and CSF white cell count with outcome illness. Our data show that prolonged duration of illness, presence of neurological signs, low haemoglobin level, low CSF sugar content, high CSF protein level, and gram positive organisms were associated with poorer outcome of childhood meningitis.
A descriptive study was carried out on patients admitted for febrile convulsion over a two-year period to determine rate of lumbar puncture (LP) refusal, factors associated with LP refusal and outcome of such patients. From 77 patients indicated and requested for LP, 19 (25%) patients refused the procedure. Refusal of LP was significantly more common among the Malay ethnic group (p = 0.01) but not significantly associated with age,gender or whether the patient was admitted for a first or recurrent febrile convulsion. Half of the patients who refused LP had to be started empirically on antibiotics for meningitis. Patients who refused LP were also 8.5 times more likely to discharge themselves "at own risk" (AOR), compared to other patients with febrile convulsion (p = 0.004). In conclusion, LP refusal is a common problem in the local setting and is a hindrance to the proper management of patients with fever and seizure. Appropriate measures must be carried out to educate the public, particularly those from the Malay ethnic group on the safety and usefulness of the procedure. Reasons for patients discharging AOR following LP refusal also need to be addressed and problems rectified.