OBJECTIVE: Enteric fever remains a leading cause of febrile illness, especially in developing countries. Due to limited resources, all investigations are not available in every health facility, but simpler investigations and clinical examinations may help in the early diagnosis of the disease. In this study, we evaluated the usefulness of eosinopenia and clinical features in predicting enteric fever in children.

PATIENTS AND METHODS: A cross-sectional study was conducted from March 2018 to March 2019. A total of 100 children with a fever of a minimum of 5 days duration and without any focus of infection were included in the study. Investigations such as hemogram with absolute eosinophil count, liver function test, Widal test, and blood culture were performed. Absolute eosinopenia was defined as the eosinophilic count of 0 μg/dl. Sensitivity, specificity, PPV, and NPV of absolute eosinopenia and other clinical indicators were calculated.

RESULTS: We analyzed data from 100 children, including 75 cases in the enteric fever group (culture/Widal positive) and 25 cases in the suspected enteric fever group. The presence of eosinopenia was statistically significant in the enteric group (p-value = 0.02). Absolute eosinopenia was seen in 52% of confirmed enteric cases, compared to 4% of the suspected enteric group (p-value = 0.001). The positive predictive values for relative bradycardia and coated tongue were 84.3% and 91.4%, respectively. Absolute eosinopenia had a specificity of 96% and PPV of 97.5%. Relative bradycardia, eosinopenia, and absolute eosinopenia were significantly more present in the culture-positive group (p-value = 0.02, 0.04, 0.02, respectively).

CONCLUSIONS: Absolute eosinopenia can be used as a strong predictor of the disease and can help in the early institution of appropriate therapy, especially in a resource-limited setting. Relative bradycardia, coated tongue, along with low TLC, thrombocytopenia, and eosinopenia can be used as other predictors.