Evaluation of the bedside Quikread go® CRP test in the management of febrile infants at the emergency department.

ABSTRACT

Recently C-reactive protein (CRP) point-of-care tests have been developed. We aimed to validate a bedside CRP test (QuikRead go® CRP), to compare it with the laboratory CRP (ARCHITECT c8000 Abbott, Germany) test in children with fever without source (FWS), and to evaluate the optimal CRP cut-off value to identify those patients at a high risk for serious bacterial infection (SBI). The CRP bedside test was prospectively performed in capillary blood samples concurrently with the laboratory CRP testing for 283 well-appearing infants aged 1 to 24 months with FWS attending the emergency department (ED) between May 2013 and August 2015. The mean difference between the laboratory CRP and the QuikRead go CRP values was 0.71 mg/L (p = 0.444). Pearson's correlation coefficient between the CRPs was r = 0.929 (p < 0.001). SBI was diagnosed in 34 patients (12.0%). The area under the receiver operating characteristics (ROC) curve obtained was 0.87 (95%CI 0.82-0.90) for an optimal CRP cut-off value of > 10 mg/L (sensitivity 94.1%, specificity 49.0%, positive predictive value 20.1%, negative predictive value 98.4%), as a predictor of SBI. Nearly 45% of the patients were at a low risk for SBI according to CRP value; thus, additional laboratory tests would have been hypothetically avoided. There was a very strong, positive correlation between the QuikRead go CRP test and laboratory CRP determination. The QuikRead go CRP test provides reliable results to rule out SBI. Its implementation at the ED would improve the management of infants with FWS.