Effect of Romanowsky-Stained Concentrated Preparations versus Direct Smears on Veterinary Students’ Ability to Identify Bacterial Sepsis in Fluid Cytology Samples from Dogs, Cats, and Horses

Abstract

Veterinary students’ accuracy, confidence, and time required to diagnose bacterial sepsis in fluid cytology samples was evaluated using two different slide preparation methods: direct smears and cytocentrifuged concentrated preparations. We hypothesized veterinary students would diagnose fluids as septic on concentrated preparations more accurately and quickly than on direct smears. Thirty third- and fourth-year students who had previously participated in a clinical pathology course completed a survey regarding general cytology experience and reviewed 40 randomized Romanowsky-stained slides via microscopy. Slides consisted of 10 septic and 10 non-septic samples with matched direct and concentrated slides, prepared from fluids from dogs, cats, and a horse. Participants’ slide evaluation time, diagnosis, confidence, and slide photographs of areas considered septic were recorded. No difference in diagnostic accuracy between direct and concentrated samples was identified (area under the curve: 57% for both preparations, p = 0.77), although students agreed with pathologist-determined diagnoses more often when viewing concentrated samples ( M = 63%, SD = 11% for concentrated; M = 56%, SD = 21% for direct, p = .012). A positive relationship existed between accuracy of diagnosis ( R2 = .59) and senior status ( p = .002), comfort interpreting cytology slides ( p < .03), and if the student had taken the senior pathology rotation ( p = .02). Only 38% (121/319) of participant photographs correctly identified sepsis. Under experimental conditions, concentrated preparations did not increase the accuracy of veterinary students’ bacterial sepsis diagnosis; however, since accuracy did increase with cytology experience and comfort level, additional pre-clinical and clinical cytology training may benefit students before entering practice.