A comparison of the clinical utility of several published formulae for estimated osmolality of canine serum

Objective To determine the normal osmole gap for 18 previously published formulae used to estimate serum osmolality in dogs. Design Prospective study. Setting University veterinary medical teaching hospital. Animals Two hundred and fifty client‐owned dogs. Interventions None. Measurements and Main Results Serum samples were saved and frozen at −80°C after routine biochemical analysis as ordered by attending clinicians. An Advanced Micro Osmometer 3300 was used to measure serum osmolality. Eighteen distinct formulae previously reported in the medical literature were used to calculate the osmolality from the biochemical analysis results. The calculated osmolality was then subtracted from the measured osmolality to determine the osmole gap. Osmole gaps for azotemic and hyperglycemic dogs were compared to those of dogs without azotemia or hyperglycemia using each formula. The median measured osmolality for all dogs in the study was 302 mOsm/kg (interquartile range 297–307). The osmole gaps varied widely depending on the formula used to calculate osmolality and the presence or absence of hyperglycemia or azotemia. Eleven formulae led to calculated osmolality and osmole gaps that were not statistically different when hyperglycemia or azotemia was present. Four out of these 11 formulae resulted in osmole gaps near zero. Conclusions and Clinical Relevance Multiple formulae reported to calculate serum osmolality can be used in the clinical setting, but they result in significantly different normal osmole gaps. Clinicians should be aware of the specific reference interval for the formula being used. The authors recommend the formula 2(Na+) + [glucose/18] + [BUN/2.8] because it is easy to use and is reliable even when hyperglycemia or azotemia are present.