期刊
TROPICAL MEDICINE & INTERNATIONAL HEALTH
卷 22, 期 3, 页码 286-293出版社
WILEY
DOI: 10.1111/tmi.12823
关键词
fever; children; infection; C-reactive protein; leucocytes; point-of-care test
资金
- Axis-Shield PoC, Oslo, Norway
- Swedish Research Council for Health, Working Life and Welfare/the European Commission under a COFAS Marie Curie Post-Doctoral Fellowship
- Training Health Researchers into Vocational Excellence in East Africa (THRiVE) - Wellcome Trust
ObjectiveTo assess the role of point-of-care (PoC) assessment of C-reactive protein (CRP) and white blood cell (WBC) count to identify bacterial illness in Tanzanian children with non-severe non-malarial fever. MethodsFrom the outpatient department of a district hospital in Tanzania, 428 patients between 3 months and 5 years of age who presented with fever and a negative malaria test were enrolled. All had a physical examination and bacterial cultures from blood and urine. Haemoglobin, CRP and WBC were measured by PoC devices. ResultsPositive blood cultures were detected in 6/428 (1.4%) children and urine cultures were positive in 24/401 (6.0%). Mean WBC was similar in children with or without bacterial illness (14.0 x 10(9), 95% CI 12.0-16.0 x 109 vs. 12.0 x 10(9), 95% CI 11.4-12.7 x 109), while mean CRP was higher in children with bacterial illness (41.0 mg/l, 95% CI 28.3-53.6 vs. 23.8 mg/l, 95% CI 17.8-27.8). In ROC analysis, the optimum cut-off value for CRP to identify bacterial illness was 19 mg/l but with an area under the curve of only 0.62. Negative predictive values exceeded 80%, while positive predictive values were under 40%. ConclusionWBC and CRP levels had limited value in identifying children with bacterial infections. The positive predictive values for both tests were too low to be used as single tools for treatment decisions.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据