新生儿复合肠外营养渗透浓度的新模型
静脉输液渗出是新生儿重症监护病房（NICU）最常见的可预防并发症之一，发生率约为2.7%。肠外营养（PN）液体渗透浓度是导致液体渗透的主要因素。有研究表明对于新生儿复合PN制剂，依据营养制剂成分通过现有的商业化软件计算出的理论渗透浓度往往低于实际测出的渗透浓度值。美国雪松西奈医疗中心发现在363例PN案例中，理论渗透浓度均低于实测渗透浓度值，且理论渗透浓度的计算偏差随PN制剂浓度的提高而加大。此外，该中心设计出一种新的运算模型（最佳拟合多项式回归）可以缩小计算偏差，更加准确地预测新生儿肠外营养制剂的渗透浓度。
JPEN J Parenter Enteral Nutr. 2017;41(2):291-292.
A New Model for Compounded Neonatal Parenteral Nutrition Osmolality.
Sivan Borenstein, Ellen Mack, Katherine Palmer, Tram Cat, Meenu Sandhu, Charles F. Simmons.
Cedars-Sinai Medical Center, Los Angeles, California, USA.
PURPOSE: Intravenous (IV) infiltrates are one of the most frequent preventable harm events (up to 2.7/100 patient admissions) in the neonatal intensive care unit (NICU). Probability of infiltration is a function of many factors, including osmolality of the parenteral nutrition (PN) solutions. Preliminary studies suggest that calculated theoretical osmolarity (mOsm/L) underestimates measured osmolality (mOsm/kg) in individually compounded PN. We sought to improve prediction of measured osmolality in compounded PN using a broader range of basic nutrient solutions with different chemical compositions and from a variety of manufacturers. This study tested the hypothesis that calculated theoretical osmolarity underestimates measured osmolality and derived a new, more accurate algorithm that predicts measured osmolality in compounded PN.
METHODS: Theoretical osmolarity (mOsm/L) was calculated from individual components of neonatal basic nutrient solutions using a commercial software tool. Osmolality of final compounded PN was determined by averaging 3 consecutive measurements obtained from a freezing point depression micro-osmometer (Advanced Model 3320; Advanced Instruments). Photoprotected PN bags were stored at room temperature. The stability of compounded PN osmolality was determined using paired samples before and after PN administration (n = 23 pairs). The osmolal gap, defined as the difference between calculated theoretical osmolarity and measured osmolality, was determined in compounded PN solutions from postadministration samples (n = 363) and from the individual basic nutrient solutions. The relationship between calculated theoretical osmolarity and measured osmolality was derived by linear or polynomial regression using the least squares method (SAS). Regression models were developed with various combinations of 10 common components of neonatal PN, in addition to theoretical osmolarity calculated from a commercial PN software tool. The best-fit regression model was tested to minimize the osmolal gap between calculated theoretical osmolarity and measured osmolality.
RESULTS: PN osmolality was stable at room temperature for ≥24 hours, thus validating analysis of postadministration samples. Calculated theoretical osmolarity underestimated measured osmolality in all PN samples (n = 363). The deviation between calculated theoretical osmolarity and measured osmolality increased with higher PN concentrations and underestimated PN osmolality as much as 20% when dextrose concentrations were >10 g/dL. Dextrose and amino acids contributed the majority of total PN osmolarity and osmolality, with dextrose contributing the most to the osmolal gap error. We determined a new, best-fit polynomial regression that effectively corrects for systematic osmolality error (R2 = 0.986).
CONCLUSIONS: Calculated theoretical osmolarity (mOsm/L) systematically underestimated the measured osmolality (mOsm/kg) in compounded PN made from solutions across a wide variety of neonatal basic nutrient solutions. We developed a new algorithm to more accurately predict PN osmolality from neonatal basic nutrient solutions commonly used in North America. Although expert consensus thresholds for excessive peripheral PN osmolality exist, these study results suggest that prior clinical studies likely incorporated a systematic underestimate of measured PN osmolality. Future research should reassess the contribution of PN osmolality to adverse PN administration events in neonates.
FINANCIAL SUPPORT: This study was funded by the Ruth and Harry Roman Chair of Neonatology in honor of Larry Baum, Department of Pediatrics at Cedars-Sinai Medical Center.
DOI: 10.1177/0148607116686023