Uncertainty associated with assessing semen volume: are volumetric and gravimetric methods that different?

The World Health Organization laboratory manual for the examination of human semen suggests that an indirect measurement of semen volume by weighing (gravimetric method) is more accurate than a direct measure using a serological pipette. A series of experiments were performed to determine the level of discrepancy between the two methods using pipettes and a balance which had been calibrated to a traceable standard. The median weights of 1.0ml and 5.0ml of semen were 1.03 g (range 1.02-1.05 g) and 5.11 g (range 4.95-5.16 g), respectively, suggesting a density for semen between 1.03g and 1.04 g/ml. When the containers were re-weighed after the removal of 5.0 ml semen using a serological pipette, the mean residual loss was 0.12 ml (120 μl) or 0.12 g (median 100 μl, range 70-300 μl). Direct comparison of the volumetric and gravimetric methods in a total of 40 samples showed a mean difference of 0.25ml (median 0.32 ± 0.67ml) representing an error of 8.5%. Residual semen left in the container by weight was on average 0.11 g (median 0.10 g, range 0.05-0.19 g). Assuming a density of 1 g/ml then the average error between volumetric and gravimetric methods was approximately 8% (p < 0.001). If, however, the WHO value for density is assumed (1.04 g/ml) then the difference is reduced to 4.2%. At least 2.4-3.5% of this difference is also explained by the residual semen remaining in the container. This study suggests that by assuming the density of semen as 1 g/ml, there is significant uncertainty associated with the average gravimetric measurement of semen volume. Laboratories may therefore prefer to provide in-house quality assurance data in order to be satisfied that 'estimating' semen volume is 'fit for purpose' as opposed to assuming a lower uncertainty associated with the WHO recommended method.