[Determination of urine density: comparison of refractometry and measurement of the period of oscillation using the protocols of the National Committee for Clinical Laboratory Standards].

Two methods for the determination of urine density (relative density or specific mass) have been evaluated and compared using NCCLS protocols: the refractometric method and a new method based on the measure of shifts in harmonic oscillation (SHO). Within the range investigated (from 1,000 up to 1,034) the methods were linear. Total imprecision, including the carry-over, was the same for both: the 95% confidence limits, at the level of 1,003, 1,013 and 1,030, were 1,001-1,005, 1,011-1,015 and 1,028-1,032 respectively. In analysing urines having known density, the SHO method was always accurate, while the refractometric method was affected by a significant bias at protein and haemoglobin concentrations greater than 4 and 2 g/L respectively. A direct comparison of the results obtained by the two methods was made on 300 untimed specimens of urine with density ranging from 1,002 to 1,034; the equation of regression line (refractometric method on the x axis, SHO on the y axis), calculated with a non-parametric statistical method, was SHO = refractometric method + 0,001. Despite some urine specimens (24, i.e. 8%) showed density differences (absolute value) of more than 0,003 (between 0,004 and 0,007 relative density units), the good mean agreement observed allows to conclude that both methods give clinically useful analytical results.