Finding a Formula for Calculating Osmolality

In clinical practice, serum osmolality is often used to evaluate patients. The osmole gap is determined by subtracting the calculated osmolality from the measured osmolality. In a healthy dog, osmolality is mostly determined by Na+, K+, Cl-, HCO3-, glucose, and urea in serum; the expected osmole gap in dogs is <10 mOsm/kg. Increases in the osmole gap can occur with some intoxications (eg, ethylene glycol, propylene glycol, ethanol).

In this prospective study, 18 previously published formulae for calculating osmolality were evaluated to determine the reference range for the osmole gap for each and to determine which formulae were most accurate regardless of the patient’s BUN or glycemic status. Of the 250 study dogs, 30 were hyperglycemic, 22 were azotemic, and 200 were neither; 2 dogs were both hyperglycemic and azotemic. Serum osmolality was measured using an Advanced Micro Osmometer 3300. Osmolality was then calculated from biochemical analysis results using each of the 18 formulae. The authors found that the closer an osmole gap is to 0 mOsm/kg, the more accurate the formula and the better it estimates osmotically active solute in the serum. Of the 18 formulae evaluated, the authors recommended the following for calculating osmolality: 2(Na+) + [glucose/18] + [BUN/2.8]. Using this formula, the median osmole gap for the 250 dogs was -2 mOsm/kg. The authors found that this formula was accurate and simple, and it uses readily available serum values.

Commentary

This intriguing study puts to rest a question asked by many criticalists: what is the best formula to use for calculating serum osmolality? Needed to determine the osmole gap, the numbers can depend on what formula you pick. While definitely not an everyday calculation for most clinicians, understanding osmolality and the osmole gap is helpful for diagnosis and management of some toxins, as well as guiding fluid therapy in the ICU. I have long used the formula that the authors recommend. There are easier ones, but the recommended formula takes enough variables into account to be consistently accurate on those rare occasions when it’s needed.—Tony Johnson, DVM, DACVECC