Search

Background Many clinical laboratories require that specimens for serum and urine osmolality determination be processed within 3 h of sampling or need to arrive at the laboratory on ice. This protocol is based on the World Health Organization report on sample storage and stability, but the recommendation lacks good supporting data. We studied the effect of storage temperature and time on osmolality measurements. Methods Blood and urine samples were obtained from 16 patients and 25 healthy volunteers. Baseline serum, plasma and urine osmolality measurements were performed within 30 min. Measurements were then made at 3, 6, 12, 24 and 36 h on samples stored at 4–8℃ and room temperature. We compared baseline values with subsequent measurements and used difference plots to illustrate changes in osmolality. Results At 4–8℃, serum and plasma osmolality were stable for up to 36 h. At room temperature, serum and plasma osmolality were very stable for up to 12 h. At 24 and 36 h, changes from baseline osmolality were statistically significant and exceeded the total allowable error of 1.5% but not the reference change value of 4.1%. Urine osmolality was extremely stable at room temperature with a mean change of less than 1 mosmol/kg at 36 h. Conclusions Serum and plasma samples can be stored at room temperature for up to 36 h before measuring osmolality. Cooling samples to 4–8℃ may be useful when delays in measurement beyond 12 h are anticipated. Urine osmolality is extremely stable for up to 36 h at room temperature.

A new method of separating serum or plasma from whole blood (Sure-Sep, W. R. Warner) has been evaluated with particular regard to the suitability of the serum/plasma samples for radioimmunoassay (R.I.A.) and competitive protein binding (C.P.B.) techniques. Sure-Sep did not affect the results obtained for the following assays: cortisol, thyroxine, triidothyronine, insulin, and growth hormone. In addition it was also found that contact of the serum or plasma samples with Sure-Sep for 24 hours at 4°C was without effect on the results obtained.