Results of digitised blood smear differentiations by veterinary students using item analysis

Familiarisation with manual blood examination methods and the morphologies of leukocytes in peripheral blood contributes to routine veterinary practice. It enables veterinarians to verify automated analysis results and to examine blood cell morphology using the microscope. Third-year students therefore participated in an online module including 10 clinical cases of various mammal species with a haematological focus. Each case required the differentiation of 100 leukocytes using digitised cell images (= items) photographed from corresponding blood films. The study aims to provide insights into student difficulties with different leukocyte morphologies by calculating the Difficulty Index (DI) values. Out of 247 participating students, 96% completed the course in full, contributing 2197 differential white blood count (dWBC) responses for evaluation. The mean DI for all items (n = 1033) was 0.95 (± 0.09 SD), indicating overall low difficulty. Nucleated red blood cells (nRBC) (DI 0.98 ± 0.03 SD), segmented neutrophils (0.98 ± 0.07), and lymphocytes (0.97 ± 0.05) obtained high scores, whereas DIs for myelocytes (0.72 ± 0.14) and monocytes (0.82 ± 0.20) posed a greater challenge for the students to recognise these types of cells. Basophils, metamyelocytes, band neutrophils, platelets, and eosinophils ranged between DIs of 0.83 (± 0.12) to 0.94 (± 0.08). In contrast to hands-on microscopy, this digital format provided valuable training to gain routine in leukocyte differentiation and presentation, particularly of uncommon cell types. These should, however, be reliably distinguished by the examiner from the more common cell types, as they usually have a relatively high clinical significance even if they occur in small numbers. Nevertheless, the lack of dynamic manual adjustments during the microscopic examination emphasises the need for hands-on microscopy in combination with a digital format.