Osmolality and Osmolarity: Narrowing the Terminology Gap

Brian L. Erstad, Pharm.D., FCCP


Pharmacotherapy. 2003;23(9) 


In this issue of Pharmacotherapy, two articles[1,2] concerning propylene glycol toxicity raise an important issue for clinicians, particularly for those practicing in the intensive care unit setting. Perhaps a less obvious issue, but one of impor-tance in a broader range of practice settings, concerns the osmotic activity calculations and measurements discussed in these articles.

The terminology associated with calculated and measured osmotic activity is often confusing and is not consistent in the medical literature. Osmotic concentration determinations are typically expressed as either milliosmoles/ kilogram (mOsm/kg) of solvent -- referred to as osmolality, or milliosmoles/liter (mOsm/L) of solution -- referred to as osmolarity. The selection of which term to use (osmolality or osmolarity) depends on how the concentration was derived. When derived by an osmometer in clinical laboratories that use a method such as freezing point depression of water (or less commonly, the vapor pressure technique), the concentration is expressed in terms of solvent and is appropriately referred to as osmolality.[3,4] Bedside calculations of osmotic activity by clinicians (using the patient's laboratory data), however, are usually expressed in terms of solution, and hence the term osmolarity is appropriate.[4] Therefore, when evaluating published literature, the reader must refer to the study methodology to determine which term is appropriate, since the investigators may have converted osmolar units to osmolal units in an attempt to increase the accuracy of calculated values,[5] or they may have used a term such as osmolality to describe both measured and calculated values.

Osmotic activity measurements and calculations are used to describe fluid movement between body compartments and to detect foreign substances in the blood, as might be expected in certain types of toxicologic problems.[3] The difference between the measured value derived from a clinical laboratory and a calculated value derived at the bedside is referred to as the "gap" and is commonly used as part of the differential diagnostic evaluation of patients. There are two major sources of confusion when evaluating this gap in the clinical setting or published literature: the method(s) used for calculating osmolality or osmolarity, and how to account for differences in units when subtracting values for calculated osmolarity from measured osmolality to determine the gap.

Numerous equations have been used to calculate osmolality or osmolarity,[5,6] and controversy exists in the literature as to which is most accurate.[6] In general, more accurate equations, particularly those that convert calculated osmolarity values into calculated (or estimated) osmolality values, involve more complex computations.[5,7,8] The value of more complicated computations in the clinical setting is questionable, as illustrated by an investigation that evaluated different equations for determining the gap in a contrived clinical scenario involving ethanol ingestion.[4] Large variations in gaps were found and the resultant standard deviations involving the different equations were similar (~6 mOsm). The results led the authors to question the clinical utility of gap evaluation in this scenario, since small gaps would not exclude toxic ingestions and large gaps could be due a number of common disease states other than toxic ethanol ingestion. Additionally, the gap determination does not tell the clinician about the specific substance involved or if more than one substance is accounting for an abnormal value. Given these inherent limitations in gap determination, there seems to be general agreement that less complicated equations [e.g., serum osmolarity = (2 x serum sodium [mEq/L]) + (BUN [mg/dl]/2.8) + (glucose [mg/dl]/18)] will suffice for use at the bedside.[4,6,7] In this case, the calculated osmolarity is considered roughly equivalent to osmolality (i.e., 1 L ~ 1 kg) since human serum is a dilute aqueous solution with a specific gravity of 1.01.[4]

The need to account for differences in the units of the calculated and measured osmotic activity values when estimating the gap is another source of confusion. In the clinical setting, the amount of error introduced by not adjusting for the differences in the osmolality and osmolarity units is small and is usually not performed.[7] Therefore, the gap is actually a hybrid value between the measured osmolality, with units of mOsm/kg, and the calculated osmolarity, with units of mOsm/L, which explains the variable terminology (osmolal gap vs osmolar gap) found in published literature. When the gap is derived in this manner, the issue seems to be less a function of which term is chosen to describe the gap and more related to the consistent use of whatever term is chosen. For example, in a recent article in a critical care journal,[8] the authors subtracted measured and calculated (with adjustment for water content) osmolality values to determine the gap, yet "osmolar gap" was used to describe the resulting gap values in both the article title and text. The term "osmol gap" increasingly has been appearing in published literature[8,9] and seems to be a useful way of avoiding the sometimes misleading implications of terms such as "osmolal gap" and "osmolar gap."

As I have illustrated in this editorial, the terminology associated with osmotic activity measurements in the clinical arena and the medical literature is confusing and inconsistent. It is, indeed, apparent that within the discipline of clinical medicine, both clarification, as well as standardization, of the terminology still needs to occur. At this juncture, however, the weight of scientific data appear to support the use of the term "osmol gap" when subtracting values for calculated osmolarity (or estimated osmolality) from measured osmolality. The methodology of articles should explain how the calculated values were derived and the laboratory technique used to measure osmolality.