Bayesian accuracy estimates and fit for purpose thresholds of cytology and culture of endometrial swab samples for detecting endometritis in mares.

This research sought to analyse the effectiveness and potential inaccuracies in testing for endometritis in horses. The primary tools of examination were endometrial swab samples, and the methods included cytology and culture tests. The aim was to understand the diagnostic accuracy of these methods, assess their impact and establish measures for suitable thresholds.

Research Objectives and Methodology

• The goal of this study was to ascertain the consequences of misdiagnosis in the regular screening and testing for endometritis in mares by using Bayesian Latent Class Models (BLCM).
• The paper sought to determine the diagnostic precision of cytology and culture of endometrial swab samples, to weigh the effects of different cytology thresholds on test accuracy and misdiagnosis costs, and to gauge the sensitivity and specificity of a diagnostic approach incorporating these tests.
• Data for the study was gathered retrospectively from software recordings of swab samples taken from mares for diagnostic or pre-breeding purposes, based in South East England from 2014 to 2020. The research involved a total of 3,448 mares.
• To classify the culture results as positive, the researchers used three different case definitions. They also graded endometrial smears based on the percentage of polymorphonuclear cells (PMN) per high power field (HPF). By employing the use of BLCM, the researchers analysed the collective results of the three culture case definitions with a cytology threshold preset at > 0.5% PMN.

Findings and Conclusions

• The study concluded that sensitivity of cytology results varied marginally between different scenarios, with values falling notably below those of previous reports that used histology as the standard test.
• Specificity estimates varied at a greater rate but were comparably high across all scenarios, showing the tests’ high potential to rightly identify a healthy mare.
• The sensitivity of culture tests showed a continuous increase across the scenarios, indicating that these tests were more likely to correctly identify mares with endometritis. The specificity estimates showed a descending trend across the scenarios, however, they resided at a high range similar to the cytology results, showing their high potential to rightly identify a healthy mare.
• Across all scenarios and endometritis prevalence contexts, increasing the cytology threshold reduced the misclassification costs associated with false positives, thereby contributing to better antimicrobial stewardship.
• The research, therefore, suggested some limitations in the procedures of testing for endometritis in mares. It showed that raising the cytology threshold could improve diagnostic performance and minimize the costs of misclassification, thus assisting the better management of antimicrobial stewardship.