A matter of agreement: The effect of the technique and evaluator on the analysis of morphologic defects in stallion sperm.

Analysis of sperm morphology is an important part of the stallion breeding soundness evaluation since it provides an objective measure of a stallion's sperm quality and is one of many factors that estimate a stallion's fertility potential. To describe the effect of sperm quality level on the technique (Differential Interference Contrast - DIC; Phase-contrast - PH; Dip-Quick staining - DQ; and eosin-nigrosin staining - EN; semen samples fixed in buffered-formal saline) and evaluator (three evaluators; using only DIC), stallions were categorized based on sperm quality into three categories: High: >57% normal sperm, Moderate: 23-56% normal sperm, or Low: <23% normal sperm (four stallions per category). The data were analyzed using three different statistical methods: Analysis of Variance (ANOVA), correlative analysis, and Bland-Altman method (agreement). A higher level of agreement among techniques was observed between DIC and PH for morphologically normal sperm. The agreement between the alternative methods (EN, DQ, or PH) and the standard method (DIC) varied, depending on the sperm quality level (High, Moderate, or Low). Some morphological defects (e.g., AH, AMP) were constantly underestimated with the staining methods (DQ, EN) compared to DIC and PH, particularly in ejaculates with low sperm morphology. Underestimation of some abnormalities, due to the technique or the evaluator, has the potential to alter the clinical interpretation of stallion fertility.