Sodium glycerophosphate and prescribed calcium concentrations in pediatric parenteral nutrition: a retrospective observational study and economic evaluation

The study protocol was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Thailand (certicate of approval no. 797/2015) following the principles of the Declaration of Helsinki and its contemporary amendments. This retrospective cross-sectional study reviewed the parenteral nutrition orders from September 2014 to August 2015 for pediatric patients including neonates and children aged <18 years who were admitted to King Chulalongkorn Memorial Hospital, Bangkok, Thailand, a public general and tertiary referral teaching hospital with approximately 1,500 beds. A requirement for informed consent was waived due to the retrospective nature of the study and anonymous nature of the data reported. Exclusion criteria included orders without a phosphate concentration or with an unspecified route of administration. Unclear handwritten orders were also excluded. Demographic data of the pediatric patients who required parenteral nutrition including age, sex, medical conditions, and nutritional assessment according to the opinion of the pediatricians were recorded. Route of administration and intravenous calcium and phosphate concentrations were the main outcomes to analyze the effect of NaGlyP, dipotassium phosphate, or mixed phosphate salts in a parenteral nutrition bag, on calcium concentrations. Calcium gluconate was the source of calcium in all parenteral nutrition solutions. Other factors that could be involved in calcium–phosphate precipitation including concentrations of amino acids, sodium, potassium, magnesium, and trace elements were also recorded. Trace elements refer to minerals in Peditrace (Fresenius Kabi, Austria), a commercial product used at King Chulalongkorn Memorial Hospital. The direct costs of the phosphates used in pediatric parenteral nutrition were calculated according to the prices of NaGlyP (Fresenius Kabi) and dipotassium phosphate (Thai Otsuka Pharmaceutical Co.) provided at the Department of Pharmacy, King Chulalongkorn Memorial Hospital.

Data were analyzed using IBM SPSS Statistics for Windows (version 22.0) with a defined level of significance at P< 0.05 for tests of statistical inference. Demographic data are shown as frequencies and percentages. A Kolmogorov– Smirnov test was used to determine the normality of data. There were 3 independent groups for analysis according to the phosphate salts used: NaGlyP or dipotassium phosphate or mixed phosphate salts. A one-way analysis of variance (ANOVA) was used to compare the means of data with normal distribution, and a Kruskal–Wallis test was used for nonparametric data.

We examined 3,535 parenteral nutrition orders made from September 2014 to August 2015; 2,192 of these orders were included in the present study. Demographic data of pediatric patients recorded in parenteral nutrition orders are summarized in Table 1. Most patients were aged ≤ 1 year (84.8%), of whom, about half were mildly malnourished (55.3%) with complex medical conditions (50%).

The reviewed pediatric parenteral nutrition orders were analyzed separately according to the route of administration (Table 2). Kolmogorov–Smirnov test results were significant; therefore, nonparametric statistics were used and all data are shown as median and ranges of 25th percentile to 75th percentile. Overall, 1,347 pediatric parenteral nutrition orders were administered through a central line (61.5%) and 845 orders were administered through a peripheral line (38.5%).

In central parenteral nutrition orders, median phosphorus concentrations between the 3 groups were not significantly different (P= 0.11), while calcium concentrations were significantly higher in the parenteral nutrition solutions containing NaGlyP than in those containing dipotassium phosphates or mixed phosphate salts (P< 0.001). Bags with pediatric parenteral nutrition containing NaGlyP had higher sodium concentrations than those containing dipotassium phosphate (P< 0.001) as a result of the sodium content of NaGlyP and other sodium sources such as sodium chloride and sodium acetate. One order (0.1%) containing NaGlyP required calcium concentration reduction from 10.09 to 4.13 mmol/L. By contrast, almost half (44.2%) of the solutions containing dipotassium phosphates required calcium concentration reductions from 1.10– 130.34 to 0.50–10.27 mmol/L with the range of reduction at 0.43–122.20 mmol/L. Concentrations of amino acids in parenteral nutrition solutions with mixed both phosphate salts were lower than those in the parenteral nutrition bags containing either NaGlyP (P= 0.02) or dipotassium phosphate (P= 0.02) alone. Although trace element content was significantly different between the 3 groups, zinc, copper, manganese, selenium, fluoride, and iodine (from Peditrace) were only added in very small concentrations and would most likely not have had significant influence on the risk of calcium–phosphate precipitation.

In parenteral nutrition orders administered peripherally, median calcium and phosphorus concentrations were significantly higher in the parenteral nutrition solutions containing NaGlyP than they were in bags containing dipotassium phosphates (P<0.001) or mixed phosphate salts (P= 0.03). Two orders (67%) containing dipotassium phosphate required calcium concentration reductions from 3.66–4.29 to 1.95–3.33 mmol/L with the range of reduction at 0.33– 2.34 mmol/L. By contrast, no calcium concentration reduction was needed in peripheral parenteral nutrition containing either NaGlyP or mixed phosphate salts.

Subgroup analysis of mineral and amino acid concentrations in parenteral nutrition solutions was further performed in preterm infants (Table 3). Only NaGlyP was used in both central and peripheral lines. The results showed high parenteral calcium and phosphorus prescription in preterm infants. Interestingly, in parenteral nutrition orders for preterm infants to be administered centrally, median calcium and phosphorus concentrations were

obviously higher than those in orders for all cases (P< 0.001). The median calcium and phosphorus concentrations in parenteral nutrition orders for preterm infants to be administered peripherally were slightly higher than those in orders for all cases. The other components were not significantly different.

According to the prices of NaGlyP and dipotassium phosphate provided at the Department of Pharmacy, King Chulalongkorn Memorial Hospital, 1 mmol of phosphate as NaGlyP cost 0.40 USD (13.80 Thai baht [THB]) and 1 mmol of phosphate as dipotassium phosphate cost 0.21 USD (7.30 THB; 1 USD equivalent to 34.241 THB, U.S. Federal Reserve Bank G5.A annual average rate 2015). Median costs of phosphate used per parenteral nutrition bag were higher for mixed phosphate salts than for either NaGlyP or dipotassium phosphate alone (P= 0.03 and P= 0.04, respectively; Table 4). However, a parenteral nutrition bag containing NaGlyP had a high cost for phosphate of up to 32.77 USD (1,122 THB). A parenteral nutrition bag containing dipotassium phosphate had relatively low cost (up to 14.34 USD or 491 THB).

In the present study, we reviewed 2,192 parenteral nutrition orders made over 1 year. A maximum calcium gluconate concentration at 139.91 mmol/L was mixed with NaGlyP 83.64 mmol/L and 5.6% amino acid in 55 mL of central parenteral nutrition solution. By contrast, a maximum calcium gluconate concentration at 12.21 mmol/L was mixed with dipotassium phosphate 3.87 mmol/L and 3.1% amino acids in 400 mL of central parenteral nutrition solution. There was no record of any visible particle appearing in a bag containing the parenteral solution either when mixing or at any time during administration. Calcium concentrations mixed with NaGlyP in pediatric parenteral nutrition solutions reported in the present study were considerably higher than those reported in published studies of intravenous calcium and phosphate stability [6, 9, 10]. The use of NaGlyP allowed greater concentrations that were compatible with calcium, which may be necessary for preterm infants who require high amounts of calcium and phosphate [1]. Other organic salts of phosphate such as sodium glucose-1-phosphate, sodium d-fructose-1,6-diphosphate, and fructose-1,6-bisphosphate are also properly compatible with calcium, even under conditions that produce a high risk of precipitation [8, 16, 17]. Due to the relatively low dissociation characteristics of phosphate from the organic salts, the risk of calcium–phosphate precipitation may be considered lower than that produced by inorganic salts. The absence of precipitation enables parenteral nutrition formulations, reduces the admixture wastage, and avoids clinical complications [1, 4].

The present study also investigated the costs of different phosphate salts used in pediatric parenteral nutrition solutions. At King Chulalongkorn Memorial Hospital pharmacy, 1 mL of NaGlyP solution costs 0.40 USD and contains 1 mmol of phosphate and 2 mmol of sodium. One milliliter of dipotassium phosphate costs 0.11 USD and contains 0.5 mmol of phosphate and 1 mmol of potassium. Considering the similar amount of phosphorus in the product, NaGlyP is approximately 2 times more expensive than dipotassium phosphate in our pharmacy. The price differential is less than the 5–10 times mentioned by Pereira-da-Silva et al. [3]. Even though the present study did not find a significant difference between the costs of NaGlyP and dipotassium phosphate used in pediatric parenteral nutrition solutions, NaGlyP should only be prescribed for patients who require high amounts of calcium and phosphorus or with suspected high risk of hazards related to calcium–phosphate precipitates. The responses of patients to the parenteral nutrition solutions containing organic phosphate should be carefully monitored, and blood chemistry should be measured to avoid any clinical complications. For example, especially in patients with kidney insufficiency, the aggressive replacement of phosphate may disturb the phosphate homeostasis resulting in clinical symptoms such as changes in mental state, vascular calcification, ventricular dysfunction, and premature death [18]. Apart from the cost, organic phosphate products are sometimes unavailable or not universally approved [12, 14]; therefore, dipotassium phosphate remains the main source of intravenous phosphate supply in many settings.

Limitations of this study include its retrospective design; therefore, the reason for calcium concentration reduction in the parenteral nutrition orders was unclear. The concentration changes may have been a response to changes in the conditions of patients, other than the risk of calcium–phosphate precipitation. Moreover, our data represented only the findings from a single hospital, and the cost analysis was based on local hospital billing rates, without considering other indirect costs. It may not be possible to generalize the results to populations in other hospitals or in other health care systems. The calcium and phosphate compatibility in parenteral solutions in the hospital has only relied on precipitates detected macroscopically; however, microprecipitates may occur without precipitates detected macroscopically [13, 19]. Some particles from microprecipitation have a diameter larger than a human capillary [13]; therefore, risks from microprecipitation cannot be excluded; and caution must be applied when using the mineral concentrations reported from this study.

The calcium concentrations in the mixtures with NaGlyP were higher than the calcium concentrations mixed with dipotassium phosphate. Despite the higher cost, NaGlyP is a useful source of phosphate for pediatric patients who need high amounts of calcium in parenteral nutrition. Further studies of the compatibility of calcium and NaGlyP in pediatric parenteral nutrition solutions in various storage conditions and the clinical outcomes of patients using these solutions are warranted.