High mortality among patients with bacterial meningitis in a rural hospital in Tanzania

W. J. Wiersinga, Q. M. van Dellen, L. Spanjaard, H. J. M. van Kan, A. L. Groen, J. C. F. M. Wetsteyn

Research output: Contribution to journalArticleAcademicpeer-review

7 Citations (Scopus)

Abstract

Although the disease is an important cause of mortality in the region, most published reports on bacterial meningitis in East Africa are from urban referral hospitals. Poor laboratory facilities make diagnosis difficult in the area and treatment is limited to inexpensive antibiotics. The case-fatality 'rate' in one rural hospital in Tanzania, the Ndala Mission Hospital (NMH), appears to have increased dramatically over recent years, perhaps as the result of increasing resistance to ampicillin and chloramphenicol. The aim of the present study, which was partially retrospective and partially prospective, was to review the number, characteristics and outcome of children admitted to this hospital with bacterial meningitis and to investigate possible resistance of the causative micro-organisms to the antibiotics used. Data from the 181 children who were admitted with bacterial meningitis [confirmed by the examination of Gram-stained smears of cerebrospinal fluid (CSF)] between 1999 and 2002 were retrospectively reviewed. The overall mortality among these children was 51%. No seasonal pattern was observed in the number of cases. In a 2-month prospective study in 2002, CSF samples from 19 consecutive cases were collected in Trans-Isolate medium and shipped to the Academic Medical Center in Amsterdam for culture and analysis of antibiotic susceptibility. For only eight (42%) of the cases was there agreement between the species of bacterium identified, by Gram-staining, in Tanzania and that identified, by culture, in The Netherlands. As there was no evidence of resistance to ampicillin and the antibiotics used in the NMH were found to be of good quality, the cause of the high mortality in the NMH remains uncertain. Poor laboratory testing, long doctor patient delays and/or poor drug administration on the wards may all be contributory factors. Attempts will now be made to address each of these problems
Original languageEnglish
Pages (from-to)271-278
JournalAnnals of tropical medicine and parasitology
Volume98
Issue number3
DOIs
Publication statusPublished - 2004

Cite this