Abstract
This study investigated and compared quantitative and semi‐quantitative coronary lesion evaluation in Atlantic salmon, Salmo salar . A total of 121 immature farmed and 47 sexually mature wild Atlantic salmon were included. Coronary arteries from all fish were morphometrically evaluated using a semi‐quantitative method. A subsample of 76 salmon was additionally evaluated using a quantitative method. Another subsample of 71 salmon was used for reliability testing of the semi‐quantitative method. Ten cross‐sections of coronary arteries located on the ventral surface of the bulbus arteriosus were semi‐quantitatively scored with regard to myointimal coronary lesions. Quantitative measurements comprised maximal intimal thickness and area lesion (%) while semi‐quantitative measurements comprised maximal intimal thickness and lesion degree (%). Scores for individual fish were expressed as the average of lesion score, denoted as mean range lesion. One person performed all evaluations blindly. Two pathologists tested repeatability of the semi‐quantitative lesion score measurements blindly. Both semi‐quantitative and quantitative variables increased significantly with increasing lesion score. Semi‐quantitative estimation gave significantly higher maximal intimal thickness values than quantitative measurement. Semi‐quantitative measurements were found to explain 63.2% of variation in quantitatively measured maximal intimal thickness. Due to lack of agreement, the two methods are not directly comparable and one method cannot replace the other. Repeatability within and between observers with regard to semi‐quantitative classification was very good with Kappa values larger than 61.5%. The semi‐quantitative method was a valid and reliable method for coronary lesion evaluation in Atlantic salmon.