The identification and quantitation of the intensity of the acute-phase response at the point of care might be of clinical relevance.
To report the possibility of automatic screening of unstained peripheral blood slides by using a 3-dimensional image analysis system.
Peripheral venous blood was obtained from children with acute inflammation/infection and examined by an automatic 3-dimensional image analyzer to detect the number of white blood cells as well as to reveal the degree of erythrocyte aggregation, a marker of the humoral phase response.
We included 66 children with acute bacterial infections and 59 with nonbacterial inflammation/infection; mean ages of the 2 groups were 4.3 +/- 3.9 years and 4.2 +/- 3.7 years, respectively (P = .91). The percentages of correct classifications based on discriminant analysis in predicting between bacterial and nonbacterial inflammation/infection were 61.3% by using the white blood cell count, 64.5% by using the percentage of granulocytes, 61.6% by using the degree of erythrocyte aggregation, and 59.2% by using the number of leukocytes counted on the unstained slides. The results of the receiver-operated characteristic curve analysis yielded an area under the curve of 0.714 (P < .001) for the number of granulocytes, 0.699 (P < .001) for the white blood cell count, 0.685 (P < .001) for the number of leukocytes on the slides, and 0.685 (P = .001) for the degree of erythrocyte aggregation. The correlation between the number of leukocytes by the electronic cell analyzer and the number of cells counted on the slides was highly significant (r = 0.85, P < .001).
It is feasible to use an automatic 3-dimensional image analyzer to reveal the different intensities of the acute-phase response between a group of children with an acute bacterial infection and another with nonbacterial inflammation/infection. These findings might be relevant for potential application at the point of care.