Purpose: Vapor pressure osmometers are currently designed to handle sample volumes as small as 2.0 microliters (microl) but smaller sample volumes are desirable in tear fluid studies. We determined the minimum sample size required for adequate validity and repeatability of osmolality measurements.
Methods: A standard saline solution (290 mmol/kg) was sampled by a variable, calibrated pipette. Forty samples were processed with a Wescor 5520 vapor pressure osmometer at each of the following volumes: 2.0, 1.6, 1.2, 1.0, 0.8, 0.7, 0.6 and 0.5 microl. Prior to each test series, the instrument was calibrated with the identical volume of 290 mmol/kg saline. Relevant descriptive statistics were computed and an analysis of variance (ANOVA) was performed on the resulting data.
Results: The mean osmolalities of the eight 40-sample groups ranged from 288.42 to 290.68 mmol/kg and were not significantly different from 290 mmol/ kg or each other (p>0.05). The standard deviations were inversely correlated with the sample volumes, gradually increased to approximately 1% as sample volume was reduced to 0.8 microl, then more rapidly increased as the sample volume was lowered still further.
Conclusion: Sample microvolumes as small as 0.8 microl can be collected for accurate and repeatable results with the Wescor 5520 vapor pressure osmometer when a standard deviation of approximately 1% is acceptable. Microvolumes from 0.7 to 0.5 microl may also be used if the expanded spread of data can be offset by multiple repeated readings. Using a 2.0 microl sample volume, the ultimate accuracy and repeatability of the Wescor vapor pressure osmometer was +/-2% at 290 mmol/kg: 99%, of all readings (+/-3 standard deviations) should fall within +/-6 mmol/kg of the true value.