Comparison of two platelet count estimation methodologies for peripheral blood smears

Area of Science:

• Hematology
• Clinical Pathology
• Laboratory Science

Background:

• Manual estimation of platelet counts from peripheral blood smears is a common laboratory procedure.
• Accurate platelet counts are crucial for diagnosing and managing various bleeding and clotting disorders.
• Automated hematology analyzers provide rapid platelet counts, but manual methods are used for verification and in resource-limited settings.

Purpose of the Study:

• To compare the accuracy and agreement of two manual platelet count estimation methods with each other and with automated counts.
• To assess the influence of varying platelet counts (high, low, normal) and hemoglobin levels (low vs. normal/high) on the agreement between methods.

Main Methods:

• Wright's stained peripheral blood smears from 184 patients were analyzed.
• Two manual methods were used: traditional (platelets/OIF x 20,000) and alternate (platelets/OIF x hemoglobin x 1,000).
• Manual estimates were compared to automated counts (Coulter LH750) using paired T-tests and correlation coefficients.

Main Results:

• The traditional method showed strong agreement with automated counts (r=0.90, p=0.87 for mean difference).
• The alternate method demonstrated significantly lower agreement with automated counts (r=0.81, p<0.0001 for mean difference).
• These agreement patterns persisted across different platelet count and hemoglobin level subsets.

Conclusions:

• The traditional manual method for estimating platelet counts from blood smears provides better agreement with automated counts than the alternate method.
• The traditional method serves as adequate quality assurance for verifying automated platelet counts.
• The alternate method, despite recommendations for low hemoglobin, did not show superior agreement in this study.