Poor reproducibility of percentage of normally shaped sperm using the World Health Organization Fifth Edition strict grading criteria

ObjectiveTo determine the reproducibility of the World Health Organization Fifth Edition (WHO5) strict grading methodology by comparing the percentage of morphologically normal sperm (PNS) recorded by the core laboratory with results obtained at the fertility centers participating in a multisite clinical trial.DesignSecondary cohort analysis of data from the Males, Antioxidants, and Infertility trial.SettingFertility centers.PatientsSemen values of 171 men participating in a multicenter, double-blind, randomized, placebo-controlled trial evaluating the effect of antioxidants on male fertility.InterventionsNot applicable.Main outcome measuresStrict morphology expressed as PNS as determined at each fertility center and the core central laboratory for the same semen sample.ResultsNo correlation was found in the PNS values for the same semen sample between the core laboratory and fertility center laboratories either as a group or by individual site. Interobserver agreement was similarly low (κ = 0.05 and 0.15) between the core and fertility laboratories as a group for strict morphology, categorized by the WHO5 lower reference limits of 4% and 0, respectively. Moderate agreement was found between the core and 2 individual fertility laboratories for the cutoff value of 0 (κ = 0.42 and 0.57). The remainder of the comparisons demonstrated poor to fair agreement.ConclusionsStrict morphology grading using the WHO5 methodology demonstrated overall poor reproducibility among a cohort of experienced fertility laboratories. This lack of correlation and agreement in the PNS values calls into question the reproducibility, and thereby the potential applicability, of sperm strict morphology testing