Effects of parenteral phosphorus dose restriction in preterm infants

Abbreviations

1 Introduction

The majority of hospitals in United States have been affected by the rapidly growing incidence of drug shortages [1–3]. Almost every component of parenteral nutrition (PN) has been in limited supply since 2010 [4]. Many hospital systems are implementing conservation strategies developed by the American Society for Parenteral and Enteral Nutrition (ASPEN) [4–6]. In response to a national shortage of parenteral phosphorus solutions (March-October 2013), a hospital-wide phosphorus dose conservation strategy was implemented which included restricted use of phosphorus in preterm infants <1250 g birth weight and no parenteral phosphorus in preterm infants >1250 g. Phosphorus was replaced only if serum level reached ≤2.0 mg/dL. Human milk fortification was introduced earlier when the enteral feeds reached 50 ml/kg/day instead of 80 ml/kg/day as per the unit protocol to compensate for the parenteral phosphorus shortage. The usual intent of neonatal parenteral phosphorus administration is to match the rate of fetal accretion, promote bone mineralization, and prevent metabolic bone disease [7–9]. Fetal accretion rates of calcium and phosphorus are maximal in the third trimester, and so premature infants are born with smaller mineral stores [10, 11]. Premature infants are at increased risk for metabolic bone disease because of smaller mineral stores, increased mineral demand, use of medications such as corticosteroids and diuretics, and immobilization [12–17]. In addition to metabolic bone disease, hypophosphatemia could lead to intracellular phosphate depletion and a fall in intracellular adenosine triphosphate [18, 19]. Adenosine triphosphate depletion impairs diaphragmatic contractility, prolonging ventilator dependence, and impairs myocardial contractility [20–23]. Recently Moltu SJ et al. suggested a relationship between hypophosphatemia and septicemia in very low birth weight infants [24]. The objective of this study was to monitor the effects of the parenteral phosphorus dose restriction in premature infants.

2 Methods

In a retrospective case-control study, preterm infants (≤35 weeks birth gestation and ≤2500 g birth weight) admitted at Rainbow Babies and Children’s Hospital neonatal intensive care unit (NICU), who received PN, survived longer than one week, and had no major congenital anomalies were studied. The content of our standardized NICU PN during the study period included dextrose 10–12.5%, Trophamine 3-4 g/kg/day, calcium 10–15 mEq/L, phosphorus 8-9 mmol/L with no cysteine additives. Clinical and laboratory data in the first four weeks of life of all eligible preterm infants admitted consecutively during the parenteral phosphorus dose conservation from March 1, 2013 to April 30, 2013 (cases) was compared to the data of infants admitted consecutively six months prior to the shortage from September 1, 2012 to December 31, 2012 (controls). Demographic data observed included gender, ethnicity, size per gestational age, antenatal steroid use, mode of delivery and APGAR scores. Nutritional data collected included duration of PN, components of PN, type of enteral feed, days to reach full enteral feeds (150 ml/kg/day), enteral feed fortification, multivitamin use, days to regain birth weight, and weekly average nutritional intake (calories, protein, calcium, phosphorus, and vitamin D). Laboratory data collected included average and peak serum calcium level, average and nadir serum phosphorus levels, days of serum phosphorus level <3 and <4 mg/dL (Usual unit practice was to consider additional phosphorus supplementation if the level is <4 mg/dL), average and peak serum alkaline phosphatase (ALP, to trend metabolic bone disease severity though it has significant limitations [25]) level and average and peak direct bilirubin levels. The protocol at the Rainbow Babies and Children’s NICU includes weekly monitoring of these labs in addition to monitoring per clinical concern. Outcomes observed included days of ventilatory and oxygen support, days of inotropic support, postnatal steroid use, radiographic evidence of rickets or osteopenia (metabolic bone disease), retinopathy of prematurity, necrotizing enterocolitis, culture proven sepsis, length of NICU stay, and mortality. The study was approved by Institutional Review Board at University Hospitals, Cleveland.

3 Statistical analysis

A univariate analysis was performed to identify differences between the case and the control groups. Student’s t test was used for parametric continuous variables, and the Mann–Whitney U test was used for nonparametric continuous variables. Chi-square tests and Fisher exact tests were used for categorical variables. All quantitative data were expressed as the mean ± standard deviation (SD), or median with inter-quartile (IQ) range. A p≤0.05 was considered to be statistically significant. The statistical software IBM SPSS Statistics version 22 (SPSS, Chicago, IL) was used for the statistical analysis of the data.

4 Results

Twenty consecutive cases were compared to 40 consecutive controls. There was no significant difference between the demographics of the two groups (Table 1). Cases had lower average serum phosphorus levels (4.1 ± 1.8 vs. 5.0 ± 1.3 mg/dL, p = 0.04) and days with phosphorus levels <3 mg/dL (median with IQ range, 0 (0, 0) vs. 0 (0, 3), p = 0.02) compared to controls (Table 2). Cases had significantly more patients with phosphorus level <3 mg/dl compared to controls [6/20 (30%) vs. 3/40(7.5%), p = 0.03]. Cases had a higher average ALP level (420 ± 170 vs. 320 ± 134 IU/dL, p = 0.02) and peak ALP level (median with IQ range, 515(409,595) vs. 358(305,481), p = 0.02) compared to controls (Table 2). The average weekly intake of phosphorus was lower in cases, compared to controls, in week 1(22.1 vs. 29.3 mg/kg/day, p = 0.051) and week 2 (66.9 vs. 78.2 mg/kg/day, p = 0.38), however, it did not reach statistical significance (Table 3). There were no significant differences in average weekly intake of calories, protein, calcium (Table 3) or vitamin D. Cases received more inotropic support (15.0 vs. 0.0%, p = 0.03) compared to controls. The 3 infants in the case group who have received inotropic support had longer days with phosphorus levels <3 mg/dL (5.3 days) and days with phosphorus levels <4 mg/dL (13.3 days) - both were above the 3rd quartile of the case group. There was no statistically significant difference in days of mechanical ventilation, chronic lung disease, diuretic use, sepsis, length of stay or mortality between the two groups (Table 2). However, the difference in average and peak ALP was present only in preterm infants with birth weight >1250 g, the group who received more parenteral phosphorus dose restriction while the clinical and laboratory characteristics were similar in infants with birth weight <1250 g. There was also a trend for lower serum phosphorus levels and need for higher inotropic support and mechanical ventilation in infants with birth weight >1250 g (Table 2).

5 Discussion

This study demonstrates that preterm infants developed significantly higher biochemical markers of metabolic bone disease after PN phosphorus dose restriction despite being exposed to a relatively short duration of PN. We speculate the modest effects of phosphorus dose restriction may become moreclinically important if shortage is prolonged or severe, or if it involves extreme preterm infants.

Drug shortages including PN products continue to increase nationwide, which is adversely affecting patient care and is a significant public health threat [5, 6]. It was reported that over 20 PN products have been on shortage lists since 2010 [5, 6]. Similarly during 2010–12 period, availability of every parenteral nutrition product, with the exception of dextrose and sterile water was affected by national shortage [26]. The common reasons for PN product shortages are limited number of manufacturers, product quality issues, or discontinuation of injectable medications [5].

PN product shortages compromise patient care, especially for vulnerable populations like preterm infants, when hospitals are forced to restrict or modify the regimens. Published reports of harm associated with parenteral drug shortages in preterm infants include, zinc deficiency dermatitis [27], dry scaly skin caused by selenium deficiency [28], copper deficiency associated neutropenia or hypochromic anemia [28] and potential for epigenetic ramification from early nutrient deficiencies [28].

While awaiting for the replacement medications, many hospital systems follow published conservation recommendations from key organizations like ASPEN to cope up with the shortages [4, 26]. Applying these considerations for each individual drug shortage in preterm infants require complex decision making process to achieve the best use of available resources [29]. The immature gut of the preterm infant further limits the use of alternate enteral therapies. The reported challenges for hospital system to deal with drug shortages include lack of advanced warnings, limited access to alternative appropriate medications and the significant time and resources and resulting negative financial effects [26].

The main limitations of this study include its retrospective design, single-center site and small sample size; however, these data suggest the potential consequences in preterm infants if phosphorus supply is not stabilized in the first weeks of life when the demand for minerals is high. Although our hospital has prioritized the delivery of parenteral phosphorus to this population via a reasonable rationing protocol, these findings indicate consequences of suboptimal phosphorus delivery. Every hospital system should have a multidisciplinary team for strategic planning and to address drug shortages in a manner that ensures patient safety and effective treatment while minimizing the financial impact according to the published guidelines by key organizations like ASPEN [26]. Parenteral drug shortages need to be addressed systematically at the manufacturer and supply level to reduce or eliminate these preventable clinical consequences.

Disclosures

None.

Funding and conflict of interests

None to report. Kera McNelis wrote the first draft of the manuscript. No honorarium, grant, or other form of payment was given to anyone to produce the manuscript.