Prevalence and predictors of malaria among febrile children attending the rural health center Jhangara in Sindh

Date of Award


Document Type


Degree Name

Master of Science in Epidemiology & Biostatistics (MSc Epidemiology & Biostats)


Community Health Sciences


Prevalence and Clinical Predictors for Malaria among Children of 0.5 to 10 years of age in Jhangara, Dadu, Sindh In rural Sindh, all children with fever or history of fever are considered to have malaria, and thus many undergo unnecessary anti-malarial treatment with risk of side effects and in addition other diagnoses would not be fully considered. This study was conducted to estimate prevalence of malaria and develop a clinical algorithm to more accurately identify malaria from non-malaria cases. Four hundred thirty eight children age 0.5 to 10 years attending the rural health center in Jhangara, district Dadu, Sindh were recruited. A Sindhi-translated standard questionnaire was used to record symptoms and duration of child's illness. Each child was physically examined, had their axillary temperature measured, and blood samples were collected from which Giemsa stained thick and thin blood films were prepared and examined for presence of Plasmodium parasites. Multiple logistic regression was used to identify the significant predictors of a plasmodium positive blood slide for all study subjects and then for the children aged 24 months or more. The sensitivity and specificity of several candidate algorithms for parasitaemia were evaluated using various combinations have identified predictors. Of 438 children, 26 (5.9%) were Plasmodium slide positive 17 (4%) had P. falciparum. and 9 (2%) had P. vivax. Among the P. falciparum positive children, 15 (887o) had mild, and 2 (127o) had moderate density of the parasite. For children aged 24 months or more fever > 3 days [adjusted odds ratio (aOR) = 5.0; 95Vo confidence interval (CI) =1.2, 26.0), rigors (aOR = 9.0; 95% CI= 3.0, 22.0) and headache (aOR =2.0, 95% CI = 0.9, 6.0), were independent predictors for a plasmodium positive blood slide. The developed clinical algorithm using the above variables had 100% sensitivity and 28% specificity. Similarly, for all study subjects; duration of fever > 3 days (aOR = 2.0; 95% CI =1 .2, 26.0), and pallor (aOR = 1.3; 95% CI= 7.0, 3.70) were independent predictors for a plasmodium positive blood slide. The clinical algorithm combining these variables for all study subject was 85% sensitive and 31% specific. We developed alternative malaria case definitions that remained highly sensitive but were somewhat more specific in low endemic areas. If validated prospectively, these algorithms could prevent over-treatment in addition, it would encourage a more complete evaluation of a sick child's illness.

This document is available in the relevant AKU library