Objective: To test a scheme for quality control of semen analysis.
Design: The reproducibility of assessment of sperm concentration, motility, and morphology was obtained for the same sample measured by different technicians (between or intertechnician variation) and for different samples assessed by each technician with time (within or intratechnician variation).
Setting: Andrology Laboratory.
Patients: Semen samples were analyzed from all patients attending the clinic.
Interventions: None.
Main outcome measures: Within technician and between technician coefficients of variation for concentration, motility, and morphology.
Results: When 100 sperm were routinely assessed, both intratechnician variation, as assessed from the precision of duplicate measurements, and intertechnician variation revealed hyperbolic curves with increasing variation at low percentages (less than 20) of motile or morphological forms. When these low values were excluded, mean intratechnician variations were 5.0%, 8.4%, and 2.8% for concentration, motility, and morphology, respectively, and mean intertechnician variations were, respectively, 6.1%, 5.6%, and 5.6%. Similar mean intertechnician variation for morphology was obtained for repeated assessment of prestained (7.3%) or presmeared (5.9%) slides. The use of cryopreserved semen to monitor longitudinal changes in the technicians' assessments revealed variations of 8.1% to 12% in concentration and 9.7% to 14% in motility. Computing the monthly means for sperm concentration, motility, and morphology over a 4.5-year period revealed a marked reduction in percentage of normal morphological forms, unrelated to the sperm count or mean age of the men attending the clinic. This was shown to be a result of a shift in the assessment by technicians.
Conclusions: Quality control is necessary and possible in an andrology clinic.