The diagnosis of typhoid fever (TF) in Kinshasa (DR Congo) was assessed by on-site surveys, external quality assessment (EQA) of the Widal test and a microbiological blood culture surveillance study. In 331/536 (61.8%) health facilities, clinicians diagnosed TF by clinical picture and the Widal test. An EQA on the Widal test consisting of three samples revealed correct scores by respectively 27.1%, 65.6% and 3.1% of 125 participating laboratories. Most (80.9% of 152 laboratories) performed <100 Widal tests per month, with a median sample positivity rate of 32.6% (range 0-90.7%). The Widal test was mostly performed on a single sample and by slide agglutination (89.5% and 97.0% respectively); errors in cold chain and procedures were recorded (not making serial dilutions, estimating titres by the intensity of agglutination). Among 293 prescribers, 52.2% and 40.8% requested the Widal test for treatment follow-up and detection of chronic carriers respectively. Salmonella Typhi was recovered from the blood in 2.4% of 3820 patients suspected as having TF, with non-typhoid Salmonellae and other Enterobacteriaceae accounting for the majority of organisms. In conclusion, clinicians rely highly on the Widal test for the diagnosis of TF and the Widal test is poorly performed and interpreted.
Copyright © 2012 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.