Clinical indicators of Pneumocystis jiroveci pneumonia (PCP) in South African children infected with human immunodeficiency virus

Abstract

Pneumocystis pneumonia (PCP) is an important cause of morbidity and mortality amongst HIV-infected children in Africa. Definitive diagnostic resources for PCP in Africa are limited due to their expense and technical difficulty, however recognising and treating children at risk is essential. As management decisions for children with pneumonia are made primarily on a clinical basis in many African regions, it is important to attempt to define a valid clinical diagnostic technique for PCP that could be used by clinicians to determine the use of correct empirical antibiotic therapy. The objectives of this study were to identify clinical features (associated with PCP in HIV-infected children hospitalised with pneumonia, to determine the combination of features that best predicts PCP in these children, and to calculate the diagnostic accuracy of these features. This study was a re-analysis of a database of a prospective study. Consecutive children below ten years of age, with a primary diagnosis of pneumonia or severe pneumonia, and who were known to be HIV-infected or were suspected of having HIV infection, were included prospectively over a 12 month period. Clinical data and diagnostic testing for PCP were obtained on admission. Bi­ and multivariate analysis of associations of the clinical variables with PCP were performed using logistic regression, to identify the combination of variables that best predicted PCP. The diagnostic accuracy of the best predicted features were calculated.