Are the increasing clinical demands for osmolality measurements and their associated electrolytes appropriate?

An audit of urine and plasma osmolalities and their associated urea and electrolytes over a 4 week period found that there were 124 plasma and 96 urine osmolality requests from 67 patients. In 21 patients (31.3 per cent), the osmolality results were useful in reaching a more precise diagnosis. In a further 11 cases, urine osmolality rather than plasma would have been appropriate. Seventy-one per cent originated from the Intensive Therapy Unit and were largely requested reflexly by the hospital computer order communication system. Plasma osmolal gaps could be calculated on 80 occasions (65 per cent). The formula 1.89 Na + 1.38 K + 1.03 urea + 1.08 glucose + 7.45 proved to be more accurate than the formula [Na+K] x 2 + urea+glucose (in mmol/L) with the latter showing a positive bias when compared to measured values. The osmolal gap was > 10 mOsm/Kg using the more complex formula on 23 occasions in 16 patients but only twice using the simpler calculation. These 16 patients usually had organ failure and were very ill. Urine sodium and potassium were measured on 72 occasions in 27 of these patients but urine chloride was never requested. Urine sodium < 20 mmol/L was found in 7 patients all of whom had relative or absolute hypovolaemia. Urine sodium was measured in 73 per cent of patients investigated for SIADH in general wards. Data was available to calculate the urine osmolal gap on 52 occasions. The value was > 100 mmol/L in 10 cases and this may be used as an index of the renal ammonium response to acidosis. Much potential derived information from simple indices is unused. As a result of this study, there was an approximate halving of the subsequent request volume.