Accuracy and feasibility of point-of-care white blood cell count and C-reactive protein measurements at the pediatric emergency department

Abstract

Background: Several point-of-care (POC) tests are available for evaluation of febrile patients, but the data about their performance in acute care setting is sparse. We investigated the analytical accuracy and feasibility of POC tests for white blood cell (WBC) count and C-reactive protein (CRP) at the pediatric emergency department (ED).

Methods: In the first part of the study, HemoCue WBC and Afinion AS100 CRP POC analyzers were compared with laboratory's routine WBC (Sysmex XE-2100) and CRP (Modular P) analyzers in the hospital central laboratory in 77 and 48 clinical blood samples, respectively. The POC tests were then adopted in use at the pediatric ED. In the second part of the study, we compared WBC and CRP levels measured by POC and routine methods during 171 ED patient visits by 168 febrile children and adolescents. Attending physicians performed POC tests in capillary fingerprick samples.

Results: In parallel measurements in the laboratory both WBC and CRP POC analyzers showed good agreement with the reference methods. In febrile children at the emergency department (median age 2.4 years), physician performed POC determinations in capillary blood gave comparable results with those in venous blood analyzed in the laboratory. The mean difference between POC and reference test result was 1.1 E9/L (95% limits of agreement from -6.5 to 8.8 E9/L) for WBC and -1.2 mg/L (95% limits of agreement from -29.6 to 27.2 mg/L) for CRP.

Conclusions: POC tests are feasible and relatively accurate methods to assess CRP level and WBC count among febrile children at the ED.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. Bland-Altman and linear correlation plots for comparison of WBC methods in the laboratory.

Bland-Altman plot presenting level of agreement (A) and a scatterplot showing linear correlation (B) between the HemoCue white blood cell (WBC) point-of-care test and Sysmex WBC results in 77 clinical blood samples. All testing was performed in the hospital’s central laboratory.

Fig 2. Bland-Altman and linear correlation plots for comparison of CRP methods in the laboratory.

Bland-Altman plot presenting level of agreement (A) and a scatterplot showing linear correlation (B) between the Afinion AS100 C-reactive protein (CRP) point-of-care test and Modular CRP results in 48 clinical samples. All testing was performed in plasma samples in the hospital’s central laboratory.

Fig 3. Bland-Altman and linear correlation plots for comparison of WBC methods in the emergency department.

Bland-Altman plot presenting level of agreement (A) and a scatterplot showing linear correlation (B) between the HemoCue white blood cell (WBC) point-of-care (POC) test and Sysmex WBC results in 171 pediatric emergency department (ED) visits. Clinicians performed POC testing at the ED in capillary blood samples and Sysmex analysis was done in venous blood samples in the laboratory.

Fig 4. Bland-Altman and linear correlation plots for comparison of CRP methods in the emergency department.

Bland-Altman plot presenting level of agreement (A) and a scatterplot showing linear correlation (B) between the Afinion AS100 C-reactive protein (CRP) point-of-care (POC) test and Modular CRP results in 171 pediatric emergency department (ED) visits. Clinicians performed POC testing at the ED in capillary blood samples and Modular CRP analysis was done in plasma separated from venous blood samples in the laboratory.