The medical management of central nervous system infections in Uganda and the potential impact of an algorithm-based approach to improve outcomes.

Published

Journal Article

BACKGROUND: In sub-Saharan Africa, HIV has increased the spectrum of central nervous system (CNS) infections. The etiological diagnosis is often difficult. Mortality from CNS infections is higher in sub-Saharan Africa compared to Western countries. This study examines the medical management of CNS infections in Uganda. We also propose a clinical algorithm to manage CNS infections in an effective, systematic, and resource-efficient manner. METHODS: We prospectively followed 100 consecutive adult patients who were admitted to Mulago Hospital with a suspected diagnosis of a CNS infection without any active participation in their management. From the clinical and outcome data, we created an algorithm to manage CNS infections, which was appropriate for this resource-limited, high HIV prevalence setting. RESULTS: Only 32 patients had a laboratory confirmed diagnosis and 23 of these were diagnosed with cryptococcal meningitis. Overall mortality was 39%, and mortality trended upward when the diagnosis was delayed past 3 days. The initial diagnoses were made clinically without significant laboratory data in 92 of the 100 patients. Because HIV positive patients have a unique spectrum of CNS infections, we created an algorithm that identified HIV-positive patients and diagnosed those with cryptococcal meningitis. After cryptococcal infection was ruled out, previously published algorithms were used to assist in the early diagnosis and treatment of bacterial meningitis, tuberculous meningitis, and other common central nervous system infections. In retrospective comparison with current management, the CNS algorithm reduced overall time to diagnosis and initiate treatment of cryptococcal meningitis from 3.5 days to less than 1 day. CONCLUSIONS: CNS infections are complex and difficult to diagnose and treat in Uganda, and are associated with high in-hospital mortality. A clinical algorithm may significantly decrease the time to diagnose and treat CNS infections in a resource-limited setting.

Full Text

Duke Authors

Cited Authors

  • Trachtenberg, JD; Kambugu, AD; McKellar, M; Semitala, F; Mayanja-Kizza, H; Samore, MH; Ronald, A; Sande, MA

Published Date

  • November 2007

Published In

Volume / Issue

  • 11 / 6

Start / End Page

  • 524 - 530

PubMed ID

  • 17512773

Pubmed Central ID

  • 17512773

International Standard Serial Number (ISSN)

  • 1201-9712

Digital Object Identifier (DOI)

  • 10.1016/j.ijid.2007.01.014

Language

  • eng

Conference Location

  • Canada