FEVER SENSITIVITY AND ITS ATTRIBUTABLE FRACTION IN MALARIA DIADGNOSIS AMONG CHILDREN IN MAKURDI NIGERIA
Fever (axillary temperature ≥ 37.5 oC) is no more recommended for clinical diagnosis of malaria for treatment, but in practice, the use of fever to diagnose malaria is not completely abandoned. This study investigated the sensitivity of fever in malaria diagnosis and its attributable fractions in children. Some 1738 children aged 0-14 years, half with fever and another half without fever were enrolled in a case control study in Makurdi. Fever, Giemsa stained thick blood films, and expert microscopy were used to diagnose malaria, and determine the sensitivity of fever, and its attributable fractions among the children. The sensitivity of fever, specificity, and positive predictive values decreased as the children’s age rose, while the negative predictive value increased across the age groups. Sensitivity of 71.3 %, (95 % CI: 64.4 % - 78.2 %) in < 5 years old children decreased to 50.0 %, (95 % CI: 57.1 % - 77.1 %) in 10 – 14 years old, while specificity of 58.5 %, (95 % CI: 53,7 % - 63.3 %) declined to 52.9 %, (95 % CI: 48.5 % –59.3 %). Malaria attributable fraction of fever (MAFF) and population attributable fraction (PAF) also declined from 70.5 % to 56.3 %; and 29.3 % to 11.0 % respectively. These results suggest that the continuous use of fever for malaria diagnosis may identify some cases of the disease among children, but a large proportion of fevers seen in children may not be due to malaria.