Estimation of white blood cell and platelet counts in ovine blood smears, and a comparison with the ADVIA 120 hematology analyzer

Background Manual evaluation of blood cell counts on stained blood films is a common procedure in resource-limited laboratories of farm animal clinics. Moreover, settings for sheep blood cell counts are not provided on most veterinary hematology analyzers. Objectives We aimed to (a) compare the results of white blood cell (WBC) counts evaluated microscopically on ovine blood smears with those obtained by the ADVIA 120 hematology analyzer and validate appropriate correction factors for the manual technique; and (b) assess the two suggested factors to calculate platelet counts on blood smears in sheep. Methods The blood samples of 57 sheep were used to generate a regression equation between the average WBC count per field and the WBC count determined using the ADVIA 120 analyzer. Thirty-one new ovine samples were used to assess the agreement between the calculated WBC counts based on a generated equation and those obtained by the analyzer using the Passing-Bablok test and Bland-Altman plots. Similarly, agreements between platelet counts using two different factors for platelet calculation were assessed using the Bland-Altman plot. Results The average bias of calculated WBC counts was 0.4%, with precision and accuracy being over 95%. Regarding calculated platelet counts, Bland-Altman plots revealed a bias of 26.4% and 1.4% when the average number of platelets per field was multiplied by 15 000 and 20 000, respectively. Conclusions Microscopic WBC counting in ovine blood is a reliable alternative to automated analyses using the generated equation. A better agreement between the two methods was observed when a factor of 20 000 was used to calculate platelet counts in ovine blood smears.