Comparison of manual and automated leukocyte counts for determination of the absolute neutrophil count: application to a pediatric oncology clinic.

BACKGROUND

Determination of the absolute neutrophil count (ANC) is a critical test prior to initiation of chemotherapy and is a standard component of cancer therapy protocols. Automated determination of this parameter potentially shortens the turnaround time necessary between specimen phlebotomy and chemotherapy infusion in an outpatient setting. However, there are certain factors that can lead to spuriously elevated or lowered ANCs, possibly leading to inappropriate dosage. We therefore compared ANC results in a series of samples in which both automated and manual results were available.

PROCEDURE

Sets of 111 specimens, tested over a 1-month period, had matched automated and manual ANC results available for initial retrospective analysis. An additional set of 35 specimens with ANCs of <1.5 x 10(9)/L were subsequently analyzed in a similar fashion. Automated ANC results were obtained with a Cell-Dyn 3500 (Abbott Diagnostics, Santa Clara, CA, USA) automated hematology analyzer, and manual ANC results were obtained using 100 cell differentials performed by 1 of 13 medical technologists. Results were tabulated and analyzed using standard linear regression and scatter plot analyses.

RESULTS

Of the initial 111 specimens, automated ANC values ranged from 0.16-14.2 x 10(9)/L (median=2.6 x 10(9)/L), as compared with 0.24-13.9 x 10(9)/L (median 3.0 x 10(9)/L) for manual ANC values (R(2) = 0.99; SE=0.49). Differences between the ANC values ranged from -55 to +33% (SD=14%) of the manual value. Of the second set of 35 specimens, regression analysis yielded an R(2) value of 0.92, with a SE of 0.11. Both data sets yielded acceptable degrees of variation on scatter plot analyses.

CONCLUSIONS

Automated ANC values appear adequate for determining suitability for chemotherapy and lessen the turnaround time between specimen phlebotomy and result verification.