Typhoid is over-reported in Embu and Nairobi, Kenya.

The paper looks at the usefulness of the Widal agglutination test in the context of variable normal antibody titres in two different populations in Kenya, and in comparison to the blood culture method of diagnosis. It presents a prospective case-control study. We examined 846 blood cultures and an equal number of serum samples, and 782 stools from adults who presented at two study sites; Kenyatta National Hospital and one hospital and 3 clinics in Embu District, with symptoms similar to typhoid. Examined also were 360 serum samples and stools from adults who were apparently healthy (controls) who sought routine medical examination at the study sites. From blood cultures, isolation rates for typhoid for Embu (3% ) and Nairobi (2.2%) were not significantly different (p>0.01). In addition the control population from the two study sites did not show any significant background O antibody titre levels characteristic of typhoid endemic areas. All the 7 commonly available Widal test kits including Murex, Europath, Biotech, Humatex, Biosystems, Microsystems and Typhex, that were evaluated for efficacy were equally specific in diagnosis of typhoid by Widal agglutination methods. However, there were minor differences in the sensitivities of the kits. The Widal test method gave a lower sensitivity (81.3%) than specificity (93%) when compared to the culture of blood for diagnosis of typhoid. Going by the reports of typhoid outbreaks in Embu and Nairobi (ca. 20-25% reported prevalence) we conclude that there has been over-reporting probably due to poor methodologies of performing the Widal test. We recommend adequate clinical examination in suspected cases of typhoid in addition to proper Widal in order to improve typhoid diagnosis. Newer improved methods that are more specific and sensitive than the Widal test need to be evaluated in improving laboratory diagnosis of typhoid.