Mineral balance studies in sick preterm intravenously fed infants during the first week after birth. A guide to fluid therapy

A.S.P.E.N. clinical guidelines: parenteral nutrition ordering, order review, compounding, labeling, and dispensing

BACKGROUND

Parenteral nutrition (PN) is a high-alert medication available for patient care within a complex clinical process. Beyond application of best practice recommendations to guide safe use and optimize clinical outcome, several issues are better addressed through evidence-based policies, procedures, and practices. This document provides evidence-based guidance for clinical practices involving PN prescribing, order review, and preparation.

METHOD

A systematic review of the best available evidence was used by an expert work group to answer a series of questions about PN prescribing, order review, compounding, labeling, and dispensing. Concepts from the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) format were applied as appropriate. The specific clinical guideline recommendations were developed using consensus prior to review and approval by the American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) Board of Directors. The following questions were addressed: (1) Does education of prescribers improve PN ordering? (2) What is the maximum safe osmolarity of PN admixtures intended for peripheral vein administration? (3) What are the appropriate calcium intake and calcium-phosphate ratios in PN for optimal neonatal bone mineralization? (4) What are the clinical advantages or disadvantages of commercially available premade ("premixed") multichambered PN formulations compared with traditional/customized PN formulations? (5) What are the clinical (infection, catheter occlusion) advantages or disadvantages of 2-in-1 compared with 3-in-1 PN admixtures? (6) What macronutrient dosing limits are expected to provide for the most stable 3-in-1 admixtures? (7) What are the most appropriate recommendations for optimizing calcium (gluconate) and (Na- or K-) phosphate compatibility in PN admixtures? (8) What micronutrient contamination is present in parenteral stock solutions currently used to compound PN admixtures? (9) Is it safe to use the PN admixture as a vehicle for non-nutrient medication delivery? (10) Should heparin be included in the PN admixture to reduce the risk of central vein thrombosis? (11) What methods of repackaging intravenous fat emulsion (IVFE) into smaller patient-specific volumes are safe? (12) What beyond-use date should be used for (a) IVFE dispensed for separate infusion in the original container and (b) repackaged IVFE?

Determinants of urea production and mineral retention in parenterally fed preterm infants

OBJECTIVE

To determine how weight for gestational age affects urea and mineral excretion by preterm infants receiving total parenteral nutrition (TPN).

STUDY DESIGN

Daily urine samples were collected from all preterm infants given high calcium TPN, providing 30 kcal/g amino acids, during its first 44 months of use, and from all those given standard TPN, providing 25 kcal/g amino acids, over the previous 24 months. Urine urea and mineral excretion were measured as follows: Urea excretion mmol/kg/day = Urine urea/urine creatinine X creatinine production Creatinine production μmol/kg/day = -2.07 + 2.34 X gestational age in weeks Results: High calcium TPN was evaluated in 52 infants. Urea excretion did not rise with increasing TPN intake. During the first week, urea excretion increased with weight for gestational age, with higher rates in above average than below average weight infants. It also increased with gestational age in above average but not below average weight infants. Below average weight infants had lower potassium and phosphate excretion than those above average. Standard TPN was evaluated in 20 infants. Urea excretion increased with TPN intake to higher levels than on high calcium, and also increased with weight for gestational age.

CONCLUSION

Urea excretion was simple to measure, with remarkably consistent daily results in individuals. Below 30 weeks gestation infants on TPN providing 30 kcal/g amino acids had urea excretion < 0.1 g urea N/kg/day, < 3.5 mmol/kg/day if below average weight, and < 0.12 g urea N/kg/day, < 4.3 mmol/kg/day if above average weight. Below average weight infants retained more potassium and phosphate during the first week than those above average, and their greater requirements were provided by the TPN.

History of medical understanding and misunderstanding of Acid base balance

To establish how controversies in understanding acid base balance arose, the literature on acid base balance was reviewed from 1909, when Henderson described how the neutral reaction of blood is determined by carbonic and organic acids being in equilibrium with an excess of mineral bases over mineral acids. From 1914 to 1930, Van Slyke and others established our acid base principles. They recognised that carbonic acid converts into bicarbonate all non-volatile mineral bases not bound by mineral acids and determined therefore that bicarbonate represents the alkaline reserve of the body and should be a physiological constant. They showed that standard bicarbonate is a good measure of acidosis caused by increased production or decreased elimination of organic acids. However, they recognised that bicarbonate improved low plasma bicarbonate but not high urine acid excretion in diabetic ketoacidosis, and that increasing pCO2 caused chloride to shift into cells raising plasma titratable alkali. Both indicate that minerals influence pH. In 1945 Darrow showed that hyperchloraemic metabolic acidosis in preterm infants fed milk with 5.7 mmol of chloride and 2.0 mmol of sodium per 100 kcal was caused by retention of chloride in excess of sodium. Similar findings were made but not recognised in later studies of metabolic acidosis in preterm infants. Shohl in 1921 and Kildeberg in 1978 presented the theory that carbonic and organic acids are neutralised by mineral base, where mineral base is the excess of mineral cations over anions and organic acid is the difference between mineral base, bicarbonate and protein anion. The degree of metabolic acidosis measured as base excess is determined by deviation in both mineral base and organic acid from normal.

Feeding problems in preterm infants of preeclamptic mothers

AIM

Maternal disease can cause prematurity and neonatal complications, notably feeding problems. To determine the relationship between maternal disease and the nature and severity of neonatal feeding problems, we compared feeding profiles, time to demand feeding and length of hospital stay between preterm infants of preeclamptic mothers, mothers with amniotic infection and mothers with other disease causing prematurity.

METHODS

The retrospective study used labour ward data collected from 2002 to 2005 in a tertiary university centre to analyse three groups of singletons born at <32 completed gestational weeks to mothers with preeclampsia (n = 61), amniotic infection (n = 55) and non-preeclamptic non-amniotic infection controls (n = 55). The groups were similar in gestational age, birthweight and sex ratio; all infants received enteral feeding according to departmental guidelines. Feeding profiles and enteral/oral nutrition were compared.

RESULTS

Feeding problems occurred in 46% of the preeclamptic group, 11% of the amniotic infection group and 13% of controls. Full oral demand feeding was established at 36 0/7 weeks postmenstrual age, 35 3/7 weeks (P = 0.03) and 35 2/7 weeks (P < 0.0001), respectively. Feeding problems were the main cause of delay (7-10 days) in hospital discharge in the preeclamptic group (P = 0.0002).

CONCLUSIONS

Feeding problems are greater, and hospital stay longer, in preterm infants of preeclamptic mothers than in other preterm infants.

Antenatal steroid treatment prevents severe hyperkalemia in very low-birthweight infants

BACKGROUND

Hyperkalemia is seen quite often in very low-birthweight (VLBW) infants and concentrations sometimes become high enough to cause cardiac arrhythmia. The purpose of the present study was to identify factors that increase serum concentrations of potassium in VLBW infants.

METHODS

Retrospective comparative analysis was performed on 140 VLBW infants who had been admitted to the Toho University Perinatal Center between January 1993 and December 1999 and needed mechanical ventilation for respiratory distress. Serum concentrations of potassium at 24 and 48 h of age were compared in two groups of infants, those whose mothers did and did not receive antenatal steroid treatment. Risk factors for severe hyperkalemia were analyzed by multiple linear regression models and Pearson's partial correlation analysis.

RESULTS

Antenatal steroid treatment reduced serum potassium concentrations significantly at 24 and 48 h, as well as the incidence of cardiac arrhythmia and necessity for glucose insulin treatment for severe hyperkalemia. Multiple linear regression showed the serum potassium concentration at 24 h of age was associated with antenatal steroid hormone treatment, 24 h fluid intake volume, serum sodium concentrations at 24 h, gestational weeks and sampling site. Serum concentration of potassium at 48 h of age was associated with blood urea nitrogen, gestational week, serum sodium concentration at 48 h of age and fluid intake between 24 and 48 h of age. Urine output volume and serum creatinine concentrations were not correlated with potassium concentrations at either age.

CONCLUSION

Antenatal steroid hormone treatment can reduce early hyperkalemia in VLBW infants and also the incidence of cardiac arrhythmia and the use of glucose insulin treatment.

Internal potassium shift in premature infants: cause of nonoliguric hyperkalemia

To study the pathophysiology of nonoliguric hyperkalemia, we measured serum potassium concentration and external K balance (intake and excretion), and estimated internal K balance (a shift from intracellular space to extracellular space) in 24 nonoliguric premature infants during the first 72 hours after birth. Data were analyzed from two aspects: gestational age (group 1, 24 to 28 weeks, n = 9; group 2, 29 to 32 weeks, n = 9; group 3, 33 to 36 weeks, n = 6) and postnatal age (0 to 72 hours). Serum K concentration rose from baseline (0 hour) to 24 hours in groups 1 and 2 (p < 0.01) but did not rise in group 3. The external K balance was negative in all groups during the study period, and was more negative in the more premature infants (group 1 > group 2 > group 3) during the second 24 hours. There was a significant difference (p < 0.01) between the internal K balance of the three groups during the first 24 hours (group 1 > group 2 > group 3), and the K shift decreased significantly (p < 0.05) during the study period in groups 1 and 2. The more premature the infants, the larger the K shift and the larger the rise in serum K concentration during the first 24 hours, and the more negative the external K balance after 24 hours. These data indicate that K loading caused by the K shift associated with prematurity produces a rapid rise in serum K concentration, resulting in an increase in urinary K excretion. We conclude that an internal K shift inversely proportional to gestational and postnatal age is the primary cause of nonoliguric hyperkalemia in very premature infants.

Role of plasma phosphate measurements in detecting rickets of prematurity and in monitoring treatment

Twice weekly plasma and urine measurements were made in 24 very low birth weight infants. Intravenous feeding was given whilst infants required respiratory support. Subsequently they received breast milk or formula milk with a vitamin D supplement of 400 U/day. Fourteen babies required intravenous feeding for more than 10 days. Six developed radiological rickets or severe osteoporosis, and these infants had plasma phosphate levels < 1.2 mmol/L on breast milk or < 1.8 mmol/L on formula milk. Babies without radiological rickets had plasma phosphate levels > 1.2 mmol/L on breast milk and > 1.8 mmol/L in all but one on formula milk. Successful treatment of rickets was associated with a rise in plasma phosphate to the above levels. Untimed urine calcium and phosphate concentrations expressed as creatinine ratios were not helpful in detecting babies with rickets, but may be useful in monitoring therapy.

Renal function in sick very low birthweight infants: 3. Sodium, potassium, and water excretion

Renal excretion of sodium, water, and potassium was measured on 434 occasions in a sample of 40 infants of 25.5-33 weeks' gestation, birth weight 720-2000 g, between the ages of 0.5 and 36 days. Water excretion varied between 1% and 30% of the glomerular filtration rate, or 15-350 ml/kg/day, and varied widely from day to day in individual infants. Nearly all infants became hyponatraemic before or after the first postnatal week. There were a few instances of hypernatraemia in the first week caused by high insensible water loss. There were high levels of sodium excretion up to 16% of filtered sodium, or 21 mmol/kg/day, in the first two postnatal weeks. Highest levels of sodium excretion were seen in the most immature infants in the first week. In most infants sodium excretion increased either in the first week or later before a subsequent decline. Potassium excretion was often high in the first week, as much as 96% of filtered potassium, or 5 mmol/kg/day, and is associated with early hyperkalaemia.

Renal function in sick very low birthweight infants: 2. Urea and creatinine excretion

Plasma urea and creatinine concentrations and urea and creatinine clearances and excretion were measured in a sample of 40 infants of 25.5-33 weeks' gestation, birth weight 720-2000 g, between the ages of 0.5 and 33 days. Creatinine excretion rate was between 60 and 120 mumol/kg/day in the first five postnatal weeks (mean 90.5) and was independent of sex or growth retardation. This can be used in clinical practice to estimate instantaneous urine flow rate V, if the creatinine concentration is measured in a randomly voided urine sample, from the formula V = 90.5/urine creatinine, with 95% confidence limits +/- 39%. There is a wide range of plasma creatinine at all gestations and ages decreasing from range 75-130 mumol/l in the first two days to 35-80 mumol/l at 3 weeks of age. Plasma urea is a poor indicator of glomerular filtration rate (GFR) in sick preterm infants. GFR (ml/min/kg) can be estimated from plasma creatinine from the formula GFR = 69.2/plasma creatinine but this estimate is imprecise with 95% confidence limits +/- 46%. Urea:creatinine clearance ratio was usually less than 1.0 (range 0.18 to 1.5) and was lower when the urine flow rate was low. Urea excretion was up to 17 mmol/kg/day in the first two weeks, higher in the more immature infants. These high levels were paralleled by a high plasma urea concentration, up to 18 mmol/l. A high plasma urea is not necessarily associated with renal failure or dehydration.

Sodium and potassium in red blood cells of premature infants during the first few days: risk of hyperkalaemia

Erythrocyte sodium and potassium were studied in 64 newborn infants including 21 very low birthweight infants (birthweight less than 1500 g) during the first three days after birth. Erythrocyte sodium showed a positive correlation with gestational age (r = 0.63, p less than 0.01) and birthweight (r = 0.66, p less than 0.01). Erythrocyte potassium was negatively correlated with birthweight (r = -0.33, p less than 0.05). The Na/K ratio in red blood cells showed a positive correlation with gestational age (r = 0.60, p less than 0.01) and birthweight (r = 0.65, p less than 0.01). In VLBW infants plasma potassium rose significantly (p less than 0.01) from 0-6 h to 12-30 h and decreased from 12-30 h to 30-60 h (p less than 0.05). Erythrocyte potassium decreased slightly from 0-6 h to 12-30 h, but not significantly. A new finding that "more immature infants have higher potassium and lower sodium concentration in RBC" may suggest a potential risk of hyperkalaemia in tiny infants.