A simple equation to estimate the osmolarity of neonatal parenteral nutrition solutions

A predictive equation of osmolarity that correlates closely with the measured osmolality was determined. Taking into account that an osmometer is not available in most clinical settings, the proposed equation appears to provide a quick and simple osmolarity calculation of neonatal parenteral nutrition solutions.

We measured the osmolalities of neonatal parenteral nutrition (PN) solutions to determine if these values may be predicted by a simple equation for calculation of their osmolarity values.

The osmolalities of 101 consecutive different final PN admixtures, prepared for 36 neonates, were measured by the freezing point depression method. The respective intra-assay and interassay coefficients of variation were always <2.1%. Linear multivariate regression analysis was used to determine a predictive equation of osmolarity that correlates closely with the value of measured osmolality.

The mean (SD) osmolality of the final PN admixtures was 749.7 (165.4) mOsm/kg. The best-fitted equation, with a coefficient of discrimination R2 = .95 (R2 = .90 for samples between 500 and 1000 mOsm/L) is osmolarity (mOsm/L) = (nitrogen x 0.8) + (glucose x 1.235) + (sodium x 2.25) + (phosphorus x 5.43)-50, with the concentration of components in mmol/L. Adapting the equation in our daily practice, using g/L for glucose and amino acids, mg/L for phosphorus, and mEq/L for sodium, the equation is osmolarity (mOsm/L) = (amino acid x 8) + (glucose x 7) + (sodium x 2) + (phosphorus x 0.2) - 50, with a similar R2.

Taking into account that an osmometer is not available in most clinical settings, the proposed equation appears to provide a quick and simple osmolarity calculation of neonatal PN solutions, thus allowing more accurate decisions to be taken regarding the choice of route and rate of administration of PN solutions.