Just like molarity, normality and mass units such as mg/dL, osmolality is a way of expressing concentration. In fact, osmolality is the concentration, of dissolved ions/particles of all types in a solution. It is expressed as milliosmoles per kilogram of water (mOsm/kg of H2O) and it takes into account everything dissolved in that solution, regardless of size or charge. In a solution of serum, plasma or urine, that’s a LOT of dissolved particles!
The most commonly used osmometers measure this combined concentration by means of a technique called freezing point depression. Any time you dissolve a solute in a solution, you lower the solution’s freezing point below that of the pure solvent. Thus ocean water freezes at a lower temperature than pure water. Serum, which is a water-based solution, also has a lower freezing point than pure water. Measuring how much the temperature is lowered is an accurate means of measuring the number of particles present.
On the other end of the spectrum, dissolved particles also change the temperature at which a solution becomes a vapor. There have been osmometers on the market that measure osmolality by vapor pressure measurements. These osmometers have the drawback of not being capable of detecting volatile solutes which may be present in the solution, as they would boil off before the serum reached its boiling point. This would include such compounds as ethanol, methanol, isopropanol and ethylene glycol. A freezing point osmometer detects and includes these solutes in its measurement.
Osmolality can also be estimated from the concentrations of the major solutes present. This is referred to as a calculated osmolality. Many main chemistry analyzers currently in use will give a calculated osmolality, or the LIS can be programmed to calculate it. There are multiple formulas in use for calculating osmolality, but they all use the concentrations of the major solutes contributing to osmolality, sodium (and chloride), glucose and urea nitrogen (BUN). In the US the formula for a calculated osmolality is some version of:
(2 X Na+ in mEq/L) + (glucose in mg/dL ÷ 18) + (BUN in mg/dL ÷ 2.8)
Two times the sodium accounts for the chloride also. If you are using SI units, this formula is simply 2 times the sodium + glucose (in mmol/L) + BUN (in mmol/L).
A calculated osmolality cannot be used to replace a measured osmolality. Since the calculated osmolality uses only sodium, glucose and urea nitrogen, this method will give falsely low osmolalities whenever there is a significant excess of solutes other than these three. The presence of the volatile solutes mentioned earlier will not be detected. In addition, more common solute excesses such as severe lactic acidosis or ketoacidosis will also cause an elevated osmolality that would be missed with a calculated value. The primary clinical utility of a calculated osmolality is in conjunction with a measured osmolality. Subtracting the calculated from the measured value can tell you whether there is an “osmolal gap” present. The difference in the two values is caused by solutes other than the three included in the calculation, and may indicate the presence of an agent that should be followed up on, like alcohol or ethylene glycol.
Measuring urine osmolality is a good way to determine the ability of the kidneys to retain water and concentrate the urine.
-Patti Jones PhD, DABCC, FACB, is the Clinical Director of the Chemistry and Metabolic Disease Laboratories at Children’s Medical Center in Dallas, TX and a Professor of Pathology at University of Texas Southwestern Medical Center in Dallas.
Lablogatory 
Very good article