[Diagnostic performance of graphical anomalies in the detection of large platelets and platelet clumps].

UNLABELLED

Thrombocytopenia is a current situation for making a blood smear in routine practice in a medical analysis laboratory. Recent automated hematology analyzers enumerate platelets and generate histograms and specific flags. Operators must be aware of the characteristics of their analyzer in order to avoid spurious results in the case where microscopy review is not possible.

OBJECTIVE

We evaluated the diagnostic performance of various graphical anomalies in the detection of large platelets and platelet clumps.

PATIENTS AND METHODS

Three hundred cases of thrombocytopenia were included in the study on the basis of a platelet count less than 150 × 10(9)/L. This evaluation is expressed by the results of the sensitivity, specificity, positive predictive value and negative predictive value compared to the microscopic review of blood smear.

RESULTS

Graphical performances are variable according to microscopic review of blood smears. Indeed, a not fitted curve is the most sensitive change on platelet histogram to the presence of large platelet. A high specificity to the presence of platelet clumps is announced when the platelet curve fails to return to the baseline. Moreover, characteristic findings on the DIFF scattergram are very specific to the presence of platelet clumps.

CONCLUSION

A normal platelet histogram can validate with great confidence thrombocytopenia in cases where a blood smear cannot be read immediately.