Several guidance documents support best practices across the stages of the parenteral nutrition (PN) use process to optimize patient safety. The critical step of PN order verification and review by the pharmacist requires a contextual assessment of the compatibility and stability implications of the ordered PN prescription. This article will provide working definitions, describe PN component characteristics, and present a wide-ranging representation of compatibility and stability concerns that need to be considered prior to preparing a PN admixture. This paper has been approved by the American Society for Parenteral and Enteral Nutrition (ASPEN) Board of Directors.

INTRODUCTION

Parenteral nutrition (PN) therapy continues to be a valued clinical intervention for many patients across care settings. It is considered a high-alert medication, meaning that its use requires safety-focused policies, procedures, practices, and systems to limit patient risk. From the first use of PN, pharmacists were called upon to design and prepare safe formulations taking into account sterility, compatibility, and stability.¹ Numerous national guidance documents promote the safety and efficacy of PN therapy.^2-5 Among the guidance documents are those that specifically address the review of the PN prescription.³ In this critical step of the PN use process, a pharmacist reviews each PN prescription for completeness, clinical appropriateness, and formulation safety, communicating any concerns with the prescriber.^{6, 7} This is inherently understood to require the pharmacist to apply their knowledge of pharmaceutical science and ensure the likelihood of a safe PN formulation.

This comprehensive review fills an identified void on the topic with a deeper look behind isolated tables or lists of PN compatibility and stability. The article aims to reinforce the pharmacist's appreciation of the complexity of PN formulation review by defining and then describing the major compatibility and stability pitfalls that need to be recognized to limit the risk of infusing a potentially unsafe PN. A pharmacist specializing in nutrition support uses their knowledge, skill, and experiences to verify the safety of PN regimens recommended or prescribed by others. This article also serves as a resource for other clinicians involved in recommending or prescribing a specific PN regimen to better appreciate the pharmacist's foundation and rationale. A follow-up article will provide practical steps to help clinicians recommend or prescribe safe PN formulations with compatibility and stability in mind, and a third article will describe methodology for evaluating PN compatibility and stability.

These recommendations do not constitute medical or other professional advice and should not be taken as such. To the extent that the information published herein may be used to assist in the care of patients, this is the result of the sole professional judgment of the attending healthcare professional whose judgment is the primary component of quality medical care. The information presented is not a substitute for the judgment by the healthcare professional. Circumstances in clinical settings and patient indications may require actions different from those recommended in this document, and in those cases, the judgment of the treating professional should prevail. The American Society for Parenteral and Enteral Nutrition (ASPEN) does not endorse any particular brand of products mentioned herein. This paper has been approved by the ASPEN Board of Directors.

Pharmacy science

The pharmaceutical sciences often refer to the broad, interdisciplinary foundational sciences and the unique knowledge underpinning pharmacy practice.⁸ This is not just a collection of science for its own end, but one that ultimately supports the safe and effective use of medications. This knowledge allows both qualitative and quantitative application of scientific principles to patient care—for example, applying known data on the physical properties and chemical properties of substances to best predict solubility, compatibility, stability, formulation design, and physiologic action, among others. Pharmaceutical science disciplines include pharmaceutics, pharmacokinetics, and pharmacodynamics, which are each incorporated into the clinical care of the patient.⁸ It is the unique responsibility of the pharmacist in general to ensure that prepared medications dispensed for patient use are safe and effective. This obligation includes ensuring the integrity of any prepared parenteral dosage form, not least being the compatibility and stability of PN—arguably the most complex drug in routine clinical use. This requires the pharmacist reviewing their patient's PN formulation to be knowledgeable in pharmaceutics, a responsibility already emphasized in the literature.^9-12

Chemical compounds (eg, drugs, nutrients) are formulated into dosage forms by manufacturers to create a product to be administered to patients. For intravenous (IV) administration, this requires sterile, low-particulate (within compendial limits) dosage forms formulated as solutions—usually aqueous, but they may often include some nonaqueous solvents. The exception are those products formulated as oil-in-water emulsions (eg, lipid injectable emulsions [ILEs]). Physicochemical properties (eg, solubility, ionization constants) are accounted for in the formulation of each manufactured IV product. Each product may include substances beyond the active ingredient(s), including excipients intended for a specific role (eg, to maintain sterility, solubility, stability). Combining individual IV drug products (ie, components) into a single preparation is referred to as an IV admixture. The pharmacy aseptically prepares these IV admixtures (also known as compounded sterile preparations). The most complex such admixture, containing over 50 chemical components, is PN. Throughout this paper, the compounded sterile PN, as a preparation available for administration, is referred to as an admixture (ie, PN admixture) regardless of whether it contains ILEs (ie, 3-in-1, total nutrient admixture [TNA]) or not (ie, 2-in-1).

Inevitably, the admixture is less stable than its individual component products, with risk for incompatibilities. Risk for incompatibility and instability is ever-present and an issue long recognized with PN but less well appreciated in recent years.^{4, 13} Compatibility and stability are at least as important as sterility in determining a beyond-use date for PN admixtures, whether preparing the admixtures for immediate use in acute care or for a 7-day supply in the home care setting.

Compatibility

Compatibility refers to the uneventful physical and chemical coexistence of two or more components over time after being combined.^14-16 Combining substances at any point prior to or during administration holds the potential for physical and chemical reactivity. The term incompatibility therefore indicates an interaction between two or more substances (active ingredients, excipients, and/or materials [eg, packaging, containers, infusion sets]) over time in a specific environment. These may be physical incompatibilities (eg, complexation or leaching or desorption) or chemical incompatibilities. The latter involve reactants of intermolecular and interionic forces as might be seen with the Maillard reaction and calcium and phosphate insolubility. The manifestation of an incompatibility may be a change in color, pH, osmolality or viscosity, formation of gas, or yielding of a precipitate (whether visible or not) altering the admixture. The result may ultimately limit therapeutic effect of one or more active components or increase the risk for toxicity including that from the infusion of precipitates.

Incompatibility is generally preventable, often dependent on proportional concentrations of ingredients, requiring some knowledge of chemical structures, solubilities, pKa values for ionizable ingredients, and the pH of each component to anticipate and avoid incompatibilities. The potential incompatibilities may or may not already be defined in the literature but are less likely described for multicomponent admixtures such as PN. If an incompatibility is already reported, the pharmacist will need to be able to decipher the cause to prevent similar recurrence in formulating a PN admixture. Multiple and overlapping forces may affect a PN admixture. For example, the pH of the final admixture may influence the solubility as well as the stability of ingredients. Consider that a pH difference of 2 units represents a 100-fold change, since the scale is logarithmic not linear. Each PN component product is manufactured at a pH at which the active ingredient is most soluble and stable. The pH of each component can vary considerably from other component products and is a critical characteristic that needs to be considered when combining components in a PN admixture to avoid incompatibility and instability related to the overall pH of the final PN admixture.

Stability

The term stability refers to the maintenance of the chemical integrity of each active ingredient or the physical integrity of the dosage form/system over time in a given environment.^14-16 Instability therefore indicates the irreversible decomposition/degradation of active ingredients or the dosage form/system as a result of pH, temperature, light, oxygen, solvents, and/or reactants. These may be physical (eg, crystallization, adsorption, broken emulsion) or chemical (eg, hydrolysis, oxidation). Oxidation reactions may occur in the presence of oxygen (auto-oxidation) or additionally in the presence of light (photo-oxidation). The environment in which PN preparation—as well as storage and administration—takes place involves exposure to air and light. Instability may manifest as increased particulate matter, a visible haze, visible oil, or other color change but, in most cases, may only be identified using appropriate methodology. The rates of degradation or reaction vary with the reactants and the environment. The ultimate result is that instability can take away from biological activity of the degraded component or generate risk from toxic by-products or particulate matter, including large lipid droplets.

The physical instability of ILE refers to increases in lipid droplet size. However, chemical instability of ILE may include peroxide formation and pH change without necessarily affecting physical stability.¹⁷ Combining substances at any point prior to intravascular infusion holds the potential to alter chemical structure of a substance or the physical state of the admixture. For example, ILE physical instability may occur not only in a TNA but also with Y-site administration.¹⁸ Any change beyond acceptable parameters (eg, maintaining at least 90% of the labeled amount or initial concentration of an active ingredient throughout infusion) indicates instability with the risk of altering therapeutic effect. Although individual minerals—including electrolytes and trace elements—do not undergo decomposition, they do possess properties (ie, ionization, valence states) that invite compatibility concerns or stability concerns for the organic components in the PN admixture or the emulsion itself. Instability is not reversible; the degradation can be slowed but not stopped.

Application to practice

Whether a PN formulation is completely customized to a patient's macronutrient and micronutrient needs or begins with a standardized, commercially available PN product, variation of just one component has the potential to influence the final admixture. The complexity of a PN admixture is quickly recognized when acknowledging the many ingredients that are eventually in the patient's infusion container. Newer products and updated compatibility and stability study methods and data may even supplant older information. Furthermore, with each new product introduced to the market and included as a PN component, or as clinical practice evolves to better customize the PN to meet patient needs, recalibration of our understanding of compatibility and stability is needed. This has happened over time, whether it was the introduction of TNAs into practice,^{19, 20} including the newer ILE products,²¹ or evaluating new drugs administered by Y-site with PN.^{22, 23} The importance of compatibility and stability is heightened for the patients who will receive a 7-day supply of PN admixtures at home.²⁴ Incompatibility or instability of an admixture is seldom obvious to the unaided eye but can be determined experimentally using validated methods or can be predicted based on known physicochemical properties of all the components. Unfortunately, significant harm and fatality can occur when available data are not taken into account or pharmaceutics is not considered.^{25, 26} Pharmacists without the requisite knowledge or not applying principles do so at their own peril and place their patients at risk.

PN COMPONENTS AND RELEVANT PROPERTIES

Numerous product components are used to prepare a PN admixture. As each component has unique physicochemical properties, a summary of these will be presented. Several aspects relevant to compatibility and stability will be noted below prior to the discussion of compatibility and stability in PN admixtures in subsequent sections.

Macronutrients

Amino acids

Commercial amino acid products available in the United States are listed in Tables 1 and 2.^27-36 Most are supplied in plastic containers, with an overwrap to avoid water loss and subsequent altered concentration. Plastic is now used in preference to glass bottles for ease of transport and storage, and very little plasticizer is expected to leach into the amino acid solution from the product's plastic container during storage.

Table 1. Content and characteristics of AA injection products (pediatric)

Product Manufacturer	Aminosyn-PF 10%²⁷ ICU Medical	Premasol 10%³³ Baxter	TrophAmine 10%³⁶ B. Braun
Essential AAs, g/100 ml
Cysteine^a	0	<0.016	<0.016
Histidine	0.312	0.48	0.48
Isoleucine	0.76	0.82	0.82
Leucine	1.2	1.4	1.4
Lysine	0.677	0.82	0.82
Methionine	0.18	0.34	0.34
Phenylalanine	0.427	0.48	0.48
Threonine	0.512	0.42	0.42
Tryptophan	0.18	0.2	0.2
Tyrosine^a	0.044	0.24	0.24
Valine	0.673	0.78	0.78
Nonessential AAs, g/100 ml
Alanine	0.698	0.54	0.54
Arginine	1.227	1.2	1.2
Aspartic acid	0.527	0.32	0.32
Glutamic acid	0.82	0.5	0.5
Glycine	0.385	0.36	0.36
Proline	0.812	0.68	0.68
Serine	0.495	0.38	0.38
Taurine	0.07	0.025	0.025
Electrolytes, mmol/L
Acetate	46	94	96.2
Phosphate	None	NR	NR
Sodium	None	NR	5
Chloride	None	<3	<3
Other
Na Metabisulfite NF, mg/L	0	0	<50^b
Water for injection, USP	qs	qs	qs
pH (range)	5.5 (5.0–6.5)	5.5 (5.0–6.0)	5.5 (5.0–6.0)
Osmolarity, mOsm/L	788	865	875
Total AAs, g/L	100	100	100
Nitrogen, g/L	15.2	15.5	15.5

Abbreviations: AA, amino acid; NR, not reported; USP, United States Pharmacopeia.
^a Conditionally essential.
^b The product packaged in plastic is sulfite-free, compared with the product packaged in glass.
qs: the amount which is enough or sufficient to result in an accurate final volume of the product.

Table 2. Content and characteristics of AA injection products (adult)

Product Manufacturer	Aminosyn II 15%²⁸ ICU Medical	Clinisol 15%²⁹ Baxter	Plenamine 15%³² B. Braun	ProSol 20%³⁴ Baxter	Travasol 10%³⁵ Baxter
Essential AAs, g/100 ml
Cysteine^a	None	None	None	None	None
Histidine	0.45	0.894	0.894	1.18	0.48
Isoleucine	0.99	0.749	0.749	1.08	0.60
Leucine	1.5	1.04	1.04	1.08	0.73
Lysine	1.575	1.18	1.18	1.35	0.58
Methionine	0.258	0.749	0.749	0.76	0.40
Phenylalanine	0.447	1.04	1.04	1	0.56
Threonine	0.6	0.749	0.749	0.98	0.42
Tryptophan	0.3	0.25	0.25	0.32	0.18
Tyrosine^a	0.405	0.039	0.039	0.05	None
Valine	0.75	0.96	0.96	1.44	0.58
Nonessential AAs, g/100 ml
Alanine	1.49	2.17	2.17	2.76	2.07
Arginine	1.527	1.47	1.47	1.96	1.15
Aspartic acid	1.05	0.434	0.434	0.6	None
Glutamic acid	1.107	0.749	0.749	1.02	None
Glycine	0.75	1.04	1.04	2.06	1.03
Proline	1.083	0.894	0.894	1.34	0.68
Serine	0.795	0.592	0.592	1.02	0.50
Taurine	None	None	None	None	None
Electrolytes, mmol/L
Acetate	107.6	127	151	140	88
Phosphate	None	None	None	None	None
Sodium	50	None	None	None	None
Chloride	None	None	None	None	40
Other
Na Metabisulfite NF, mg/L	0	0	300		NR
Water for injection, USP	qs	qs	qs	qs	qs
pH (range)	5.8 (5.0–6.5)	6.0 (5.0–7.0)	5.6 (5.2–6.0)	6.0 (5.5–6.5)	6.0 (5.0–7.0)
Osmolarity, mOsm/L	1270	1357	1383	1835	998
Total AAs, g/L	150	150	150	200	100
Nitrogen, g/L	23	23.7	23.7	32.1	16.5

Abbreviations: AA, amino acid; NR, not reported; USP, United States Pharmacopeia.
^a Conditionally essential.
qs: the amount that is enough or sufficient to result in an accurate final volume of the product.

The products vary in the number and concentration of each crystalline amino acid, total amino acid, and nitrogen content, as well as electrolyte content and pH.^27-36 As a result, the amino acid products and associated compatibility and stability data are not necessarily interchangeable. Therefore, a cautious approach to determining compatibility and stability data is needed to ensure the evidence is specific to the product in question. Historically, these solutions contained chloride salts of some of the amino acids, but they currently contain some acetate salts or acetic acid buffering depending on the amino acid content, with consequent pH differences between products. The different products contain both essential (indispensable) and nonessential (dispensable) amino acids at varying proportions in compatible mixtures from the manufacturer and are expected to remain stable in their original container during storage. Storage of amino acids should avoid light exposure to maintain stability up to the expiration date. Adult products typically do not contain cysteine or taurine or glutamine. Products intended for pediatric patients contain lower concentrations of alanine, arginine, and lysine. Electrolytes may also be found in amino acid solutions, again varying with the product, and need to be considered for compatibility and total clinical dose.

Each crystalline amino acid in a commercial product is considered an active pharmaceutical ingredient and meets standards for purity. Amino acids are amphoteric in nature; therefore, depending on solution pH, they can react as acids and bases, and they are each least soluble at their isoelectric point (at which ionic charges neutralize each other), thereby reducing any ionic attraction to water. For example, tyrosine has low solubility and precipitates at a more acidic pH. The buffering capacity (titratable acidity) of amino acid solutions is determined in large part by the proportion of the formulation that is arginine, histidine, and lysine. This buffering capacity will be important to the overall PN admixture and is more robust at a higher final concentration of amino acids in the PN. Additionally, the ratio of basic to acidic amino acids may be important to optimal emulsion stability.³⁷

Most products no longer include a sulfite (SO₃^2–) antioxidant preservative, with its potential to cause hypersensitivity reactions. To help avoid oxidation, most products are filled under nitrogen atmosphere or under vacuum. However, unlabeled contaminants are present in amino acid solutions and include aluminum, cadmium, chromium, lead, zinc, and possibly manganese.^{38, 39} This contamination reflects the ability of amino acids to easily form complexes with metals. The affinity for metals varies with the amino acid, with different kinetics for each complex formation.³⁸ Nonnutrient metals (eg, cadmium, lead) are most likely to complex with alanine, aspartate, glutamate, glycine, histidine, methionine, phenylalanine, serine, and threonine.⁴⁰ Although this likelihood is based on individual amino acid stability constants, the amino acid concentration in a commercial amino acid mixture remains the major determining factor.

Amino acids can complex with macrominerals such as calcium (with lysine especially) and trace minerals such as copper (with cysteine especially). In most cases, the chelates of trace metals with amino acids are expected to maintain the bioavailability of both components.⁴¹ However, copper-cysteine complexes are likely to precipitate. The dilution of copper (elemental) to a concentration <160 mcg/L is expected to avoid this complexation.^{42, 43} Precipitates (yellow-to-brown) were noted in pediatric PN admixtures.⁴³ The precipitate was attributed to an incompatibility between copper (as cupric sulfate [SO₄^2–]) and a cysteine-containing, lower-pH amino acid solution, which did not recur when using a different amino acid solution.⁴³ No evaluation was made to determine whether the precipitate was copper cysteinate, copper sulfide (S^2–), or both.

Two amino acids are each available as a separate injectable product in glass containers: cysteine and arginine. But only cysteine is intended for PN admixtures when indicated. Cysteine hydrochloride is available as a 50-mg/ml (34.5 mg/ml cysteine) solution at a pH of 1.0–2.5. This amino acid may form the insoluble dimeric form cystine over time. As mentioned above, the complex of cysteine with copper may precipitate out of solution, decreasing amino acid concentration.⁴⁴

Although not an amino acid, l-carnitine is synthesized endogenously from lysine and methionine. This nutrient is available for inclusion in PN admixtures when indicated. The product (200 mg/ml) does not contain a preservative. Hydrochloric acid and sodium hydroxide may be used to adjust the product pH to 6.0–6.5. As a zwitterion (ie, contains both positively charged and negatively charged functional groups) with a pKa of 3.8, this very water-soluble primary amine can behave as a basic amino acid.

Dextrose

Dextrose, or α-d-glucose monohydrate (C₆H₁₂O₆·H₂O), is a natural monosaccharide resulting from the hydrolysis of corn starch and is the predominant source of carbohydrate in PN.⁴⁵ Each gram of dextrose monohydrate provides 3.4 kcal or 14 kilojoules (anhydrous dextrose provides 4 kcal/g). It is a sterile, nonpyrogenic IV solution commercially available in concentrations ranging from 5% to 70% and contains no antimicrobial agent. IV dextrose products have a pH range between 3.2 and 6.5, with most solutions in the range of 4–4.5, since the solution contains no added buffer. Dextrose may convert to anhydrous glucose in temperatures >40°C, so excessive heat exposure is to be avoided and temperatures of 20–25°C are recommended for storage; freezing is also to be avoided.^46-48 Exposure to heat during the sterilization phase of manufacturing results in the degradation products 5-hydroxymethyl-furaldehyde, levulinic acid, and formic acid, which drive down the pH of sterile dextrose solutions over time.²⁰ The small amount of the reactive open-chain form of glucose in solution varies depending on that pH.⁴⁹

Dextrose has a molecular weight (MW) of 198.17 Da and, at a concentration of 5%, is considered isotonic (252 mOsm/L), whereas 70% is hypertonic with an osmolarity of 3532 mOsm/L (∼50 mOsm/L contributed for every 1% dextrose in solution). Isotonic solutions of dextrose 5% have been shown to facilitate growth of Burkholderia cepacia and Serratia marcescens after contamination at 25°C.⁵⁰ Most PN formulations contain dextrose at a final concentration of 10%–25% in present-day clinical practice.

The commercially available concentrations of dextrose contain no more than 25 mcg/L of aluminum or 5 mg/L of di-2-ethylhexyl phthalate (DEHP) (in polyvinyl chloride [PVC] bags) within their expiration periods, which may be of most importance in neonates with chronic exposure.⁵¹ Patients with a known allergy to corn or corn products should avoid use of dextrose.^{47, 48}

Lipid emulsions

ILEs are an essential component in PN as a source of energy and essential fatty acids and as a therapeutic modality. During the manufacturing process, an egg lecithin–based phospholipid emulsifier is used to disperse the oil phase into the aqueous phase to ensure a stable emulsion. ILE products are oil-in-water emulsions consisting of one or more oils containing triglycerides, a phospholipid emulsifier, and glycerol and are manufactured to mimic properties of natural chylomicrons (Table 3).^52-58 Sodium hydroxide is added to maintain a pH range between 6 and 9, and the glycerol is added to improve tonicity and provide additional energy.⁵⁹ Nearly 1 million lipid droplets can be found per milliliter, and the emulsion remains most stable at pH 8 with a surface potential of at least −35 mV in repulsive forces.⁶⁰

Table 3. Composition and properties of lipid injectable emulsion products

Category	Component	OO,SO-ILE⁵³	SO-ILE⁵⁴	SO-ILE⁵⁵	FO-ILE⁵⁶	SO,MCT,OO,FO-ILE⁵⁷
Commercial product concentration		20	20,30	20	10	20
Source oil	SO, %	20	100	100	0	30
	FO, %	0	0	0	100	15
	MCT, %	0	0	0	0	30
	OO, %	80	0	0	0	25
Additives	Egg phospholipid, g/100 ml	1.2	1.2	1.2	1.2	1.2
	Glycerin, g/100 ml	2.25	2.25	2.5	2.5	2.5
	α-Tocopherol, mg/100 ml	3.2	0	0	15–30	16.3–22.5
	Sodium oleate, g/100 ml	0.03	0	0.03	0.03	0.03
ω-3	Linolenic acid, %	0.5–4.2	4–11	4–11	1.1 (mean)	1.5–3.5
	EPA, %	0	0	0	13–26	1–3.5
	DHA, %	0	0	0	14–27	1–3.5
ω-6	Linoleic acid, %	13.8–22	44–62	48–58	1.5 (mean)	14–25
ω-6	Arachidonic acid, %	0	0	0	0.2–2	NR
ω-7	Palmitoleic acid, %	0	0	0	4–10	NR
ω-9	Oleic acid, %	44.3–79.5	19–30	17–30	4–11	23–35
Saturated fatty acids	Caprylic acid, %	0	0	0	0	13–24
	Capric acid, %	0	0	0	0	5–15
	Palmitic acid, %	7.6–19.3	7–14	9–13	4–12	7–12
	Stearic acid, %	0.7–5	1.4–5.5	2.5–5	0	1.5–4
	Myristic acid, %	0	0	0	2–7	NR
	ω-6:ω-3 ratio	9:1	7:1	7:1	1:8	2.5:1
	Phytosterols, mg/L	208.8 ± 39.4	422.4 ± 130.5	NR	0	142.2 ± 15.3
	Phosphate, mmol/L	15	15	15	15	15
Energy	kcal/ml	2	2	2	1.12	2

Abbreviations: DHA, docosahexaenoic acid; EPA, eicosapentanoic acid; FO-ILE, fish-oil lipid injectable emulsion (Omegaven, Fresenius Kabi USA); OO,SO-ILE, olive-oil, soybean-oil lipid injectable emulsion (Clinolipid, Baxter Healthcare Corporation); NR, none reported; SO-ILE, soybean-oil lipid injectable emulsion (Intralipid, Baxter Healthcare Corporation; Nutrilipid, B. Braun Medical); SO,MCT,OO,FO-ILE, soybean-oil, medium-chain triglycerides, olive-oil, fish-oil lipid injectable emulsion (Smoflipid, Fresenius Kabi USA).

The United States Pharmacopeia (USP) Chapter <729> (“Globule Size Distribution in Lipid Injectable Emulsions”) provides explicit pharmacopeial specifications to ensure manufacturer product integrity, including a mean droplet diameter below 500 nm (0.5 µm) and a large globule content (percentage of fat globules >5 µm or ‘‘microns’’ [ie, PFAT₅]) no greater than 0.05%.⁶¹ Additionally, ILEs must possess a free fatty acid content no greater than 0.07 mEq/g.

These emulsions, however, are inherently unstable systems and, over time, may undergo various stages of destabilization such as aggregation, creaming, coalescence, and increased particle size. In aggregation, dispersed lipid droplets come together but do not fuse. These aggregates may be redispersed with gentle agitation and safely administered to patients. Creaming is the initial stage of emulsion destabilization and is one of only two stages that may be detectable by the naked eye, the other being free oil as the final stage of a broken emulsion. The lipid particles present in a cream layer are destabilized, but their individual droplet identities are generally preserved. A cream layer is visible at the surface of the emulsion as a translucent band separate from the remaining dispersion. With gentle agitation, a creamed emulsion may be redispersed and safely administered to a patient. Coalescence occurs when there is a fusion of lipid droplets, leading to a fewer number but an increase in droplet size. This type of emulsion cannot be redispersed and is unsafe to administer to patients or include in a PN admixture.⁶²

The stability of ILE is dictated by its zeta potential and the effectiveness of its emulsifying agent. The emulsifier forms a mechanical barrier around each fat droplet and simultaneously forms an electrostatic (ionic) barrier with their ionized phosphate groups to separate droplets from each other and the surrounding water. Emulsifying agents create a barrier that prevents droplets from coalescing to form unsafe, larger globules. The zeta potential is a negative charge created by the emulsifier that generates repulsive forces that act to separate particles and work against attractive (van der Waal) forces, thus improving the stability of the emulsifier. Compatibility and stability of an ILE are dependent on the emulsifier's strength and the presence of confounders that may counteract the emulsifier's repulsive properties to keep oil droplets apart.⁶³ The zeta potential can be negated with the addition of divalent cations (eg, calcium, magnesium), trivalent cations (eg, iron), or high concentrations of monovalent cations (eg, potassium, sodium) or by decreasing the emulsion's pH. As the pH decreases to <5, the zeta potential becomes less negative and more neutral, thus increasing the chance of droplets coalescing and creating larger, unsafe particles. Optimal ILE stability is achieved at a pH range of 7–8 and when the zeta potential is high (ie, more negative), at least −35 mV, as the repulsive forces will exceed the attractive forces. Conversely, all repulsive forces disappear by a pH of 2.5 and the lipid particles will coalesce, resulting in an unsafe emulsion. Other factors that can disrupt an emulsion's integrity include exposure to oxygen, as it will oxidize the fats within the emulsion. Package design and use of an oxygen absorber help lessen this risk.⁶⁴ Even with these protective measures, ILEs will destabilize over time as the emulsion's pH decreases and particle size increases.

Until recently, only soybean-oil ILEs were available in the United States. In the past decade, ILEs containing olive oil, fish oil, and medium-chain triglycerides (MCTs) have become available (Table 3).⁵² Because of the differences in chain length associated with these sources of triglycerides, the same phospholipid emulsifying agent may behave differently, and emulsion stability may be different depending on the blend of oils used in an ILE.⁶⁵ For example, MCTs can improve the stability of an ILE by displacing long-chain triglycerides (eg, from soybean, fish, or olive oil) at the droplet surface, thus reducing stress on the emulsifier because of its shorter chain length.⁶⁶

For these reasons, compounding a TNA (ie, 3-in-1 PN admixture) or coadministering parenteral medications with an ILE remains a challenge. Any time an ILE has another medication or IV fluid mixed with it, there is a potential for a disruption of the emulsion or coalescence that creates larger fat globules that may be unsafe to infuse. Some literature on ILE stability may not utilize the methods specified by USP Chapter <729> (ie, dynamic light scattering, light obscuration) for determining the stability of an emulsion. Importantly, both methods must be used in a complementary fashion to assess the quality of ILEs. Dynamic light scattering is only a sizing method and provides only semiquantitative information on the relative droplet size distribution; it does not count the number of particles or droplets. In contrast, light obscuration both counts the fat globules present and identifies the size of lipid droplets (1.3–400 µm), providing “reliable” count values at a size threshold of 5 µm.

Micronutrients

The micronutrients include electrolytes, trace elements, and vitamins. Each is a required component in a PN admixture to support a patient's metabolic needs. A parenteral micronutrient product may include multiple electrolytes, multiple trace elements, or multiple vitamins, but several of these nutrients are also available as single-entity products. The physicochemical properties of each active ingredient and accompanying excipients are important to the issue of PN admixture compatibility and stability. Although some of the mineral salts are organic, the remainder are inorganic.

Electrolytes

The electrolytes, found as salts in injectable products, that serve as components for PN admixtures include sodium, potassium, magnesium, calcium, and phosphorus. Generally, electrolyte solutions are mildly acidic with polarity similar to water. This allows the salts to dissociate in aqueous PN admixtures to varying degrees depending on the salt and concentration. The monovalent electrolytes (sodium and potassium in their representative salts) are much less likely than the multivalent electrolytes (calcium, magnesium) to pose compatibility concerns.

Multielectrolyte additives

Products that contain multiple electrolytes may offer some convenience as a component in preparing PN admixtures. Each product will vary in the concentration and specific electrolyte salts, often low in potassium and without phosphorus. As a result, the pH may vary between products in a range of ∼4–8, often requiring hydrochloric acid or acetic acid for pH adjustment, and are hypertonic (Hyperlyte CR [B. Braun] [pH 5.0–5.4], 5500 mOsm/L; TPN Electrolytes [Hospira, Inc] [pH 6.0–7.5], 6200 mOsm/L). These sterile solutions may contain one or more different sodium salts (acetate, chloride, gluconate), potassium chloride, magnesium chloride, and often calcium chloride. Incidentally, magnesium chloride and calcium chloride may be included as part of standardized, commercially available PN products. When these products are used as a component for preparing a final PN admixture for a patient, the difference in any included electrolyte salts is important to note, as chloride salts tend to dissociate to a greater degree and are available for interaction, with an influence on pH.

Bicarbonate, citrate, and lactate salts should be avoided in PN admixtures because of incompatibilities. For example, with a pH of nearly 8, sodium bicarbonate drives precipitation of several salts, including calcium and magnesium. Cardioplegic solutions, intraperitoneal irrigation solutions, or dialysate solutions—which usually contain sodium, magnesium, and calcium as chloride salts, as well as containing potassium chloride in cardioplegic solutions, or other sodium salts (eg, acetate, bicarbonate, gluconate, or lactate) in irrigation and dialysate solutions—are not indicated as components to be used for PN admixtures.

Sodium and potassium chloride or acetate

The chloride salts (sodium chloride, potassium chloride) generally form neutral aqueous solutions available for use in preparing PN admixtures. As a weaker conjugate acid anion, the acetate salts (sodium acetate, potassium acetate) form more alkaline solutions. Despite significant dissociation of the ions from these salts in solution, the monovalent minerals are less likely to interact with other components at typical final admixture concentrations.

Magnesium sulfate or chloride

Aqueous solutions of magnesium are expected to have a pH of 5.5–7.0; this includes magnesium sulfate products adjusted with sulfuric acid and sodium hydroxide. Magnesium chloride is not typically available as a separate additive but has been included in combination products, including the standardized, commercially available PN products.

Calcium gluconate or chloride

Although calcium gluconate is the preferred salt for use in compounding PN because of its lower risk of precipitation, calcium chloride has also been used. The pH for calcium gluconate injections ranges from 6.0 to 8.2 and is adjusted with hydrochloric acid and/or sodium hydroxide. Based upon this pH range, this salt is more likely to be mildly alkaline than acidic. The pH of calcium chloride injections is 5.5–7.5 when diluted with water for injection, and it may contain hydrochloric acid and/or sodium hydroxide for pH adjustment. Additionally, calcium chloride dissociates more extensively in water than does calcium gluconate, rendering more calcium ions free to complex with phosphate and to produce calcium phosphate precipitates. Even if the extent of dissociation was similar, calcium gluconate dissociates into a weak acid (gluconic acid) and a strong base (calcium hydroxide), whereas calcium chloride dissociates into a strong acid (hydrochloric acid) and a strong base (calcium hydroxide).

Sodium or potassium phosphate

Mixtures of inorganic monobasic (dihydrogen) and dibasic (monohydrogen) phosphate salts of sodium and potassium are available for use in compounding PN in the United States. Although the pH of each individual phosphate salt is not identified in USP monographs, the pH of sodium phosphates injection has been reported in the range of 5.0–6.0, whereas the pH of potassium phosphates injection has been reported between 6.2 and 7.8. The degree of dissociation of inorganic phosphate salts into free phosphate anions is largely dictated by pH. In addition, phosphate may occur in three different anionic forms, including H₂PO₄⁻ anion (dihydrogen or monobasic phosphate), HPO₄²⁻ (monohydrogen or dibasic phosphate), and PO₄³⁻ (tribasic phosphate), all of which may generate insoluble calcium phosphate. The content of aluminum contamination is much greater in potassium phosphates injection than in sodium phosphates injection.

Trace elements

In the time since the initial recommendations for including copper, chromium, manganese, and zinc in PN admixtures,⁶⁷ and after hospitals had to prepare their own formulations,⁶⁸ there have been a number of lessons learned in IV trace element requirements as well as an evolution in the available products on the market.^69-72 Currently, there are multi–trace element products as well as single-entity trace element products available to include in PN admixtures or other carrier fluids (Table 4).^73-76 The multi-element products are commonly used because of convenience, whereas the single-entity products become more valuable for long-term use in patients with varying requirements or for individual-element repletion independent of the PN admixture. Although rarely observed, certain trace element solutions may exhibit a slight colored hue.

Table 4. Trace element products

	Multiple-entity product				Single-entity product
	Pediatric			Adult
	Multitrace-4 Neonatal⁷³ (1 ml)	Multrys (1 ml)	Multitrace-4 Pediatric⁷³^a (1 ml)	Tralement⁷⁴^b (1 ml)	Chromium (1 ml)	Copper (1 ml)	Manganese (1 ml)	Selenium (1 ml)	Zinc (1 ml)	Zinc (1 ml)
Trace elements
Chromium (as chloride), mcg	0.85	None	1	None	4	None	None	None	None	None
Copper (as sulfate), mcg	100	60	100	300	None	None	None	None	None	None
Copper (as chloride), mcg	None	None	None	None	None	400	None	None	None	None
Manganese (as sulfate), mcg	25	3	25	55	None	None	100	None	None	None
Selenium (as selenious acid), mcg	None	6	0	60	None	None	None	60	None	None
Zinc (as sulfate), mg	1.5	1	1	3	None	None	None	None	3	5
Other
Water for injection, USP	qs	qs	qs	qs	qs	qs	qs	qs	qs	qs
Preservative	None	None	None	None	None	None	None	None	None	None
pH	2.3–2.7	1.5–3.5	1.5–3.5	1.5–3.5	1.5–2.5	2.0–3.5	2.0–3.5	1.8–2.4	2.0–4.0	2.0–4.0
Osmolarity, mOsm/L	50	NR	34.7	114	308	327	3.64	16	96.5	157.2

Abbreviations: NR, not reported; USP, United States Pharmacopeia.
^a Manufacturer has voluntarily ceased production to develop US Food and Drug Administration (FDA)–approved products that align with American Society for Parenteral and Enteral Nutrition (ASPEN) recommendations.
^b Tralement is indicated for pediatric patients ≥10 kg as well as adults.
qs: the amount that is enough or sufficient to result in an accurate final volume of the product.

Products formulated for neonatal and pediatric patients do not contain a preservative, as the commonly used benzyl alcohol is associated with adverse effects. Most multidose vial products contain no preservative/bacteriostatic agent. Without a preservative, the vials of any trace element product have a limited beyond-use date after initial entry. Of note, the concentration of each trace element is relatively low in the products, made even lower when added to a PN admixture, creating a challenge in analysis when close to the limits of detection. Furthermore, most methods of analysis determine mineral concentration without distinguishing between oxidation states or complexation and precipitation.

Multielement additives

Historically, products have contained up to seven trace elements, but multi–trace element products in the United States currently remain at four elements in fixed-concentration proportions. The available products vary in concentration of each trace element, the excipients, and pH. Sulfuric acid and sodium hydroxide may be used to adjust product pH to 1.5–3.5. Products with excessive copper and manganese content have been transitioned to products with lower concentrations of these minerals in keeping with expert recommendations.^{69, 70} Products containing fluoride, iodine, iron, and/or molybdenum may still be available outside the United States.

Chromium

Chromium can exist in multiple oxidation states, including the toxic hexavalent chromate form. The chromic (Cr³⁺) ion is the most physiologically valuable and the one used in parenteral products, compared with the chromous (Cr²⁺) form. Chromic chloride is a hydrated salt (MW = 266.5 Da) containing just under 20% chromium. Hydrochloric acid and sodium hydroxide may be used to adjust product pH to 1.5–2.5. Given the amounts of chromium present as a contaminant in other PN components (eg, amino acid solutions), the requirement for this trace element as an additive in short-term treatment has been reconsidered. Although chromium concentrations can be measured using neutron activation analysis, atomic absorption spectrophotometry, or inductively coupled plasma with mass spectrometry (ICP-MS), these reflect total chromium and may not adequately distinguish oxidation state of the mineral, which is important to compatibility as well as physiologic activity.

Copper

The oxidized cupric (Cu²⁺) form is more soluble and stable in solution than the cuprous (Cu¹⁺) state. Copper (cupric) sulfate is a hydrated salt (MW = 249.69 Da) containing just over 25% copper. Sulfuric acid and sodium hydroxide may be used to adjust product pH to 2.0–3.5. Copper (cupric) chloride is also available and may contain hydrochloric acid and sodium hydroxide to adjust product pH to 1.5–2.5. Copper can form complexes with organic anions, including some amino acids, depending on the degree of other competing cations available in the admixture. Some of these copper complexes may have limited solubility. Although copper concentrations can be measured using atomic absorption spectrophotometry or ICP-MS, these reflect total copper content and cannot distinguish oxidation state or the free form from the complexed mineral, including insoluble microprecipitates. For example, a PN compatibility study that noted no visually apparent copper precipitate and reported unchanged copper concentration over time using atomic absorption did not necessarily rule out copper microprecipitate.⁷⁷

Manganese

Of the several oxidation states of manganese, the divalent (Mn²⁺) form is most stable in aqueous solutions. This form has been available in both the chloride and sulfate salts. Manganese sulfate is a hydrated salt (MW = 169 Da) containing 32.5% manganese. Sulfuric acid may be used to adjust product pH to 2.0–3.5. Although manganese concentrations can be measured using neutron activation analysis, atomic absorption spectrophotometry, or ICP-MS, these reflect total manganese and cannot distinguish between the free and the complexed mineral.

Selenium

Selenium can exist in a number of oxidation states. Selenious acid (H₂SeO₃) incorporating the selenite ion (SeO₃^2–) and containing 61% selenium has been more commonly used than sodium selenate (SeO₄^2–) for parenteral preparations. Nitric acid may be used to adjust product pH to 1.8–2.4. Although selenium concentrations can be measured using atomic absorption spectrophotometry or ICP-MS, these reflect total selenium and cannot distinguish oxidation state, including the poorly soluble elemental (Se⁰) form, or distinguish between the free and complexed mineral.

Zinc

Fortunately, zinc is very soluble across a wide range of pH in its ionic and stable divalent (Zn²⁺) form. So there has been no concern about reduction to the less soluble elemental form. Zinc sulfate is a hydrated salt (MW = 287.56 Da) containing 22.7% zinc. Sulfuric acid may be used to adjust product pH to 2.0–4.0. Zinc can form complexes with organic anions, including some amino acids with multifaceted coordination structure. Zinc chloride may form a flocculent precipitate in aqueous solution, owing to cationic hydrolysis to the oxychloride, which is less evident when adjusted to low pH with hydrochloric acid.⁶⁸ Although zinc concentrations can be measured using atomic absorption spectrophotometry or ICP-MS, these reflect total zinc and cannot distinguish the free from the complexed mineral.

Iron

For several reasons, iron salts are not typically added to PN admixtures, nor are they a component of currently available multi–trace element products in the United States. However, iron salt is present as ferric chloride in one product, Addamel N (Fresenius Kabi), that has been imported in times of shortage.⁷⁸ Iron dextran injection, USP is a dark brown, slightly viscous sterile liquid complex of ferric hydroxide and dextran containing 50 mg/ml iron. In its undiluted state, the pH of the solution is between 5.2 and 6.5 with added sodium hydroxide and hydrochloric acid for pH adjustment.⁷⁹ Although iron dextran has been the traditional iron product used in PN, iron sucrose is being used more frequently in clinical practice for iron replacement therapy because of its lower incidence of adverse effects. Iron sucrose injection is a brown, sterile, aqueous complex of polynuclear iron (III)–hydroxide in sucrose (MW = 34,000–60,000 Da). Iron sucrose injection (20 mg/ml) contains ∼30% sucrose wt/vol (300 mg/ml) and has a pH of 10.5–11.1. Undiluted, the injection has an osmolarity of 1250 mOsm/L.⁸⁰

Iron is rarely indicated for inclusion in PN admixtures for short-term use. Typically, patients receiving long-term PN (ie, >3 weeks) will require some form of iron supplementation to meet ongoing needs. Microcytic, hypochromic anemia resulting from iron deficiency can occur in patients receiving iron-free PN in as little as 2 months.⁸¹ When administered parenterally, concern for iron overload limits its usefulness. Unlike enterally administered iron, parenteral iron bypasses the complex homeostatic control of iron bioavailability at the gastrointestinal tract. Routine monitoring of iron status should be considered, as iron overload can increase oxidative stress and the risk of infection.⁸² In addition to the clinical issues associated with parenteral iron (eg, impaired immune function, anaphylaxis risk, support bacterial growth during acute infection), there are significant compatibility concerns.^{81, 83}

Vitamins

Multivitamins

Commercially available adult and pediatric IV multivitamin formulations comprise water-soluble and lipid-soluble vitamins (Tables 5 and 6).^84-87 The M.V.I.-12 formulation for adults, which did not contain vitamin K, is no longer available. Various excipients are added to the multivitamin products to maintain the integrity of the product. To maintain stability, the adult and selected pediatric IV multivitamin formulations are separated into two vials.^85-87 Vial 1 of the Infuvite Adult and Pediatric formulations has a pH of 5.0–6.0, whereas the pH of vial 2 is 5.0–7.0 (Sandrine Banzundama Bazuta Feza, Medical Information Coordinator, Sandoz Canada Inc, written communication, May 20, 2020). After the contents of vials are combined, the mixture is only stable for 4 h under refrigeration.^85-87 Furthermore, once contents of both vials are introduced into a PN admixture, the infusion should be completed within 24 h. A pediatric multivitamin formulation is also available as a lyophilized powder for reconstitution.⁸⁴ The adult multivitamin products should be further diluted in compatible solutions of volumes ≥500 ml,^85-87 whereas pediatric multivitamin products require diluent volumes ≥100 ml.^{84, 86}

Table 5. Pediatric parenteral multivitamin ingredients

Vitamin	Source	Amount^{84, 86}
A	Retinol palmitate	2300 IU (0.7 mg)
B₁	Thiamin hydrochloride	1.2 mg
B₂	Riboflavin 5-phosphate sodium	1.4 mg
B₃	Niacinamide	17 mg
B₅	Dexpanthenol	5 mg
B₆	Pyridoxine	1 mg
C	Ascorbic acid	80 mg
D₃	Cholecalciferol^a	400 IU (10 mcg)
D₂	Ergocalciferol^a	400 IU (10 mcg)
E	dl-α-Tocopheryl acetate	7 IU (7 mg)
K₁	Phytonadione	200 mcg
B₉	Folic acid	140 mcg
Biotin	Biotin	20 mcg
B₁₂	Cyanocobalamin	1 mcg

^a Pediatric Infuvite Multiple Vitamins⁸⁶ contains cholecalciferol, whereas M.V.I. Pediatric⁸⁴ lyophilized powder contains ergocalciferol.

Table 6. Adult parenteral multivitamin ingredients

Vitamin	Source	Amount^{85, 87}
Vial 1 (5 ml) ingredients
A	Retinol palmitate	3300 IU (1 mg)
B₁	Thiamin hydrochloride	6 mg
B₂	Riboflavin 5-phosphate sodium	3.6 mg
B₃	Niacinamide	40 mg
B₅	Dexpanthenol	15 mg
B₆	Pyridoxine	6 mg
C	Ascorbic acid	200 mg
D₃	Cholecalciferol^a	200 IU (5 mcg)
D₂	Ergocalciferol^a	200 IU (5 mcg)
E	dl-α-Tocopheryl acetate	10 IU (10 mg)
K₁	Phytonadione	150 mcg
Vial 2 (5 ml) ingredients
B₉	Folic acid	600 mcg
Biotin	Biotin	60 mcg
B₁₂	Cyanocobalamin	5 mcg

^a Adult Infuvite Multiple Vitamins⁸⁷ contains cholecalciferol, whereas M.V.I. Adult⁸⁵ contains ergocalciferol.

For both adult and pediatric multivitamin products, the lipid-soluble vitamins are solubilized with the aqueous vitamins through the addition of the surfactant polysorbate 80.^84-87 Both M.V.I. Adult and Pediatric products contain an additional surfactant, polysorbate 20.^{84, 85} Infusion of large amounts of polysorbate 80 has been associated with serious adverse events, including fatality in low-birth-weight infants.⁸⁸ In the 1980s, an injectable vitamin E formulation (E-Ferol), containing polysorbate 80 (9%) and polysorbate 20 (1%), administered intravenously led to the development of a multiorgan toxicity referred to as E-Ferol syndrome.⁸⁸ E-Ferol syndrome is a life-threatening condition characterized by vasculopathic hepatotoxicity, ascites, thrombocytopenia, and renal failure in low-birth-weight infants.⁸⁸ Following reports of this lethal reaction, E-Ferol was withdrawn from the market.⁸⁸ Commercially available pediatric multivitamin products contain a small quantity (50 mg) of polysorbate 80 in 5 ml of solution or 1% (wt/vol).^{84, 86}

Propylene glycol facilitates the dissolution of vitamins in the adult multivitamin products.⁸⁹ Because of the risk of hyperosmolality in low-birth-weight infants, propylene glycol is not included in pediatric multivitamin products.⁹⁰ Antioxidants such as gentisic acid ethanolamide, butylated hydroxytoluene, and butylated hydroxyanisole are added to selected adult multivitamin products.^{89, 91} In addition to butylated hydroxytoluene and butylated hydroxyanisole, mannitol serves as an antioxidant in pediatric multivitamin products.⁹² Buffer agents such as sodium hydroxide/hydrochloric acid or citric acid/sodium citrate optimize the pH of the multivitamin products.^84-87 Multivitamin vials of current US products should be refrigerated and protected from light.^84-87 Hypersensitivity reactions to multivitamin products (active ingredients and excipients) have been reported.⁹³ A test dose can be given to the patient to determine the allergenicity.⁹³ If identified, the allergenic component should be avoided. The patient should be closely monitored in an inpatient setting when the test dose or modified PN formulation is initiated.⁹³

Ascorbic acid

Besides being a component of IV multivitamin formulations, ascorbic acid is available as a single-entity product with a concentration of 500 mg/ml.^{94, 95} Excipients include disodium edetate, a chelating agent to protect against oxidation,⁸⁹ and sodium hydroxide and sodium bicarbonate to buffer the solution to a pH range between 5.5 and 7.0.^{94, 95} The solution should be stored under refrigeration and protected from light.^{94, 95} A hypertonic product (5900 mOsm/L), the ascorbic acid solution must be further diluted with a compatible diluent (eg, dextrose 5% in water) to a final concentration of 1–25 mg/ml.⁹⁵ Ascorbic acid is used at lower concentrations when included as an additive to PN admixtures to avoid compatibility and stability concerns.

Thiamin

Thiamin hydrochloride is available as a 100-mg/ml solution as well as an ingredient in IV multivitamin products. The single-entity vial may contain chlorobutanol anhydrous and/or monothioglycerol as preservatives.^{89, 96, 97} Sodium hydroxide is added to maintain a pH range of 2.5–4.5.^{96, 97} The thiamin solution should be stored at room temperature and protected from light.^{96, 97}

Pyridoxine

Pyridoxine is available as a 100-mg/ml solution in addition to being included in IV multivitamin products. The excipients of the single-entity product include chlorobutanol anhydrous, a preservative,⁸⁹ and sodium hydroxide to maintain the pH between 2.0 and 3.8.⁹⁸ The solution should be protected from light and stored at room temperature.⁹⁸

Folic acid

Folic acid is available as a sodium folate 5-mg/ml multiple-dose solution as well as being a component of IV multivitamin solutions. The single-entity folic acid solution is yellow and contains edetate disodium and benzyl alcohol 15 mg as a preservative.^{89, 99} Benzyl alcohol has been associated with hypersensitivity reactions and gasping syndrome in neonates, characterized by respiratory depression, metabolic acidosis, encephalopathy, and intracranial hemorrhage.⁸⁹ The pH of the folic acid solution is maintained between 8.0 and 11.0 with hydrochloric acid and sodium hydroxide, because of the risk for precipitation out of solution at lower pH.⁹⁹ The folic acid vial should be protected from light and stored at room temperature.⁹⁹

Vitamins can be highly reactive molecules, setting up potential compatibility issues with each other or with other components of a PN admixture as well as stability issues on their own or for other components of a PN admixture. Incompatibilities depend on the relative concentrations of ingredients; pH; exposure to temperature, oxygen (air), and light; and the duration of time in reaction. As a result of compatibility and stability concerns, the vitamins are added as the last component and as close to the time of administration as practical.

COMPATIBILITY

Admixture compatibility—Focus on macronutrients

Preparing a PN admixture will influence the properties and safety of the individual macronutrient components. The acidic pH of IV dextrose is well buffered when combined with an amino acid solution. The viscosity of the concentrated dextrose (eg, 70% product) is also tempered in an admixture. One of the first compatibility studies evaluated the combination of amino acids at a final concentration of 4.25% (ie, 4.25 g of amino acids per 100 ml of PN admixture) and dextrose at a final concentration of 25% (AA4.25%-D25%), stored at 4°C, 25°C, or 37°C.¹⁰⁰ The mixture showed a slight decrease in pH over time as well as color formation, especially at higher temperature. Some amino acid degradation occurred (see Stability section). In another study with an AA4.25%-D25% solution without any additives, the pH only changed from 6.5 to 6.4 initially but then remained constant for up to 2 weeks.¹⁰¹ In these glass containers, the average counts of particulate matter in the PN preparation over time were reported, using the microscopic method, as 2.7 per milliliter (particle size 10–25 µm) and 0.13 per milliliter (particle size 25–50 µm), which is well below the USP threshold.¹⁰¹ That threshold criteria (USP method 2) for IV admixtures with a volume >100 ml state there should be ≤12 particles/ml of size ≥10 µm and ≤2 particles/ml of size ≥25 µm.¹⁰²

Owing to the reducing properties of dextrose, a chemical interaction referred to as the Maillard reaction may occur between glucose and free (nonionized) amino acids. As a result, no commercial products with amino acids and dextrose premixed can be manufactured because this glycation reaction would occur during the product's heat sterilization. The Maillard reaction is the same nonenzymatic series of reactions that causes browning in food and that also accounts for HbA_1c formation. Glucosylamines form between glucose carbonyl groups and amino groups of the amino acids, creating unstable Schiff bases that rearrange chemically to form Amadori compounds. The initial steps can be reversible, but subsequent reaction steps are not and lead to loss of amino acid bioavailability. A solution color change may be noted from yellow to a darker amber and brown in the presence of the pigmented compounds.¹⁰³ Unrecovered urinary losses of Maillard products may be accompanied by increased loss of trace elements (eg, zinc).¹⁰⁴ The reaction is more likely to occur with time and at 25–30°C compared with 4°C but is minimized at pH 3.5–4.0.^{103, 105} For example, at AA4.25%-D25%, the average loss of amino acids to the Maillard products was ≤2% between 1 and 30 days when stored at 4°C.¹⁰⁶ This increased to a ∼6% loss at room temperature and will be even greater in the presence of electrolytes and increasing pH values.¹⁰⁶ Amino acid losses may be significant over time, depending on storage conditions.¹⁰⁵ The rates of reaction will vary with the amino acid involved, and lysine may be the most reactive in the Maillard reaction. A nearly 20% loss of lysine may occur by 1 month at 30°C, which tends to be more reactive, especially at higher pH (∼7.5).¹⁰⁷ At lower molar ratios of glucose to amines, the activation energy for the reaction may increase compared with those at higher ratios.¹⁰⁵

The addition of ILE to the admixture of amino acids-dextrose not only will increase the complexity by combining active ingredients and excipients and by shifting pH character but will alter the physical system as the entire PN admixture becomes an oil-in-water emulsion. An early study of a PN admixture (Veinamine AA3.2%-D20%-L4% Intralipid) noted no visually apparent incompatibility, and fatty acid analysis revealed no significant changes in composition at 24 or 48 h at 4°C.²⁰ An evaluation of several adult amino acid products (Travasol, Aminosyn, FreAmine III, FreAmine HBC, HepatAmine, and NephrAmine) in TNAs using a 100% soybean-oil ILE (Soyacal 20%) and dextrose 70% in a 1:1:1 volume ratio plus micronutrients was conducted over 14 days at 4°C.⁴⁹ Although the PN admixture pH and osmolality differed between preparations, they did not change significantly over the study period. An evaluation was performed of a pediatric amino acid product (TrophAmine 10%) in a PN admixture using ILE 20% (Intralipid, Liposyn II, or Nutrilipid) and dextrose 70% with micronutrients at several final macronutrient concentrations for 24 h at 4°C and an additional 24 h at room temperature.¹⁰⁸ No visible precipitates or color changes were noted, and pH remained 5.5–6.6, with the lowest value associated with the highest dextrose concentrations.¹⁰⁸ In addition to the buffering capacity provided by higher amino acid concentration, the higher final dextrose concentration may provide some emulsion protection based on its viscosity.^{109, 110} Stability of the final admixture emulsions will be described later (see Stability section). This section on admixture compatibility reinforces that PN admixtures should be stored at 4°C for reasons above and beyond sterility concerns (ie, the focus of USP Chapter <797>). A broader point is that setting a beyond-use date for PN admixtures will depend on sterility as well as compatibility and stability.

Admixture compatibility—Focus on electrolytes

Adding electrolytes directly to an amino acids solution resulted in visible precipitates in a large proportion of over 80 combinations that differed in proportional ratios of the minerals and in mixing sequences.¹¹¹ The majority of the precipitates evaluated were revealed to contain calcium and phosphate. Monovalent ions can associate with the zwitterionic form of amino acids but are not expected to form precipitate.

When varying the concentration of calcium and magnesium with high phosphorus content (65 mmol/L), higher levels of magnesium protected against precipitation of calcium phosphate.¹¹² This is likely because magnesium forms a more stable and soluble salt with phosphate.¹¹² Methods for predicting calcium phosphate precipitation must be validated by experimentation and consider the entire formulation, preparation technique, and chemical kinetics.

In the United States, we still use inorganic phosphate salts (Na or K), which pose the risk for insoluble complexes with cations—especially calcium. It is important to note that precipitates can occur in PN but are not always visible, will not appear immediately after compounding, and can occur later in the administration set/catheter. These precipitates develop because the three ionic phosphate species exist in an equilibrium in aqueous solution—monobasic (H₂PO₄^–), dibasic (HPO₄^2–), and trivalent (PO₄^3–). Based on the pKa values for each equilibrium, the pH of solution will determine which phosphate species predominates. The risk of insolubility depends on much more than just the combination of calcium and phosphate salt forms and concentrations. This includes final pH, amino acid product, amino acid final concentration, dextrose final concentration, the calcium salt, the phosphate salt, temperature, sequence of calcium and phosphate addition, time that the PN is stored prior to administration, and other factors.^112-114

pH

The final pH of the PN admixture is the most important factor in determining calcium and phosphate compatibility. Inorganic salts (such as calcium chloride, monobasic/dibasic phosphates) are more likely to dissociate into free ions compared with organic salts (eg, calcium gluconate, sodium glycerophosphate). The extent of this dissociation is largely governed by the pH of the final formulation.¹¹² In aqueous solutions, phosphate may exist in three anionic forms—monobasic (H₂PO₄^–), dibasic (HPO₄^2–), and trivalent (PO₄^3–)—and all three forms can produce insoluble calcium phosphate complexes.¹¹⁵ Because extreme alkalinity is generally required to ionize the tribasic phosphate species, it is not normally found in PN admixtures. Although the trivalent phosphate is not expected in PN, the dibasic species readily forms an insoluble complex with calcium that crystallizes and precipitates out of solution. The concentrations of the monobasic and dibasic are pH dependent, and a small change of 0.05 in pH can increase the dibasic phosphate salt concentration by >10% and contribute to the precipitation of dibasic calcium phosphate (CaHPO₄).¹¹⁵ Essentially, as the pH of the final admixture increases, there is more dibasic phosphate present to bind with free calcium. Conversely, the lower the final admixture pH, the greater the percentage of the monobasic (H₂PO₄^–) salt at which it can form more soluble calcium dihydrogen phosphate salt (Ca[H₂PO₄]₂).

Amino acids

Amino acid products from different manufacturers can vary in pH, ranging from 5 to 7. Amino acid products also have an intrinsic buffering system, largely determined by the concentrations of arginine, histidine, and lysine. Higher final concentrations of amino acids have a greater buffering capacity, and therefore, there is less of an increase in pH when phosphate is added to the final formulation. The more moderate pH will support calcium and phosphate solubility. High concentrations of amino acids have also been reported to sequester calcium and decrease its reaction with phosphates. Lysine has been identified as the amino acid most likely to complex with calcium, and the degree of that complexation is also pH dependent. Better determination of safe PN admixtures, especially for neonates, will require additional study using different amino acids products, as well as varying the salt form and concentrations of calcium and phosphate.¹¹⁶

Phosphate salts

Inorganic phosphates injections available in the US market contain a mixture of monobasic and dibasic salts of either sodium or potassium. The sodium phosphates injection contains monobasic sodium phosphate and monohydrate and dibasic sodium phosphate, anhydrous. Similarly, the potassium phosphates injection contains a mixture of monobasic and dibasic potassium phosphate, anhydrous. All sources of potassium or sodium phosphate (ie, amino acids) must be taken into consideration, as high concentrations of phosphate increase the risk of precipitate formation with calcium. Although the use of a single calcium phosphate product has been advocated for use in predicting compatibility, this approach is fraught with error as the results from multiplying the PN calcium concentration (in mEq/L) with PN phosphorus concentration (in mmol/L) vary inconsistently as the calcium concentration decreases and phosphorus concentration increases.¹¹⁷

Sodium glycerophosphate is an organic form of phosphorus available in Europe and has been temporarily available for use in the United States during phosphorus drug shortages. In contrast to inorganic phosphate injections, the phosphate component of sodium glycerophosphate is covalently bound to a glycerol backbone, which hinders the formation of a precipitate with divalent calcium ions. The availability of a monobasic potassium phosphate product in Canada enhanced calcium and phosphate solubility to allow for a greater concentration in a PN admixture, although the aluminum content was not described.¹¹⁸

Calcium salts

Calcium gluconate salts are recommended over calcium chloride salts when compounding PN because of superior compatibility with inorganic phosphates. Because the gluconate salt of calcium dissociates much less extensively in water compared with calcium chloride, it is the preferred salt for use in PN admixtures. Calcium chloride almost completely dissociates in aqueous solutions, so precipitation with phosphate occurs at much lower chloride salt concentrations vs gluconate salt concentrations. One drawback associated with calcium gluconate injections in glass vials is that it is heavily contaminated with aluminum, so the higher aluminum load must be considered, especially when used in neonatal PN admixtures.¹¹² Of note, calcium chloride is the salt included in commercial, standardized, multichamber PN products with electrolytes. Compatibility testing has demonstrated that calcium chloride is physically compatible with sodium glycerophosphate in pediatric PN formulations.¹¹⁹ Thus, compounding with sodium glycerophosphate and calcium chloride provides a mechanism for reducing the aluminum load in PN.

Dextrose concentration

The USP monograph for dextrose injections indicates that these products are generally acidic, with a pH range between 3.2 and 6.5. Thus, higher concentrations of dextrose will lower the final admixture pH, providing a greater percentage of monobasic phosphate (H₂PO₄^–), which can form more soluble calcium dihydrogen phosphate salt (Ca[H₂PO₄]₂).¹¹⁷

Temperature

Temperature may influence the dissociation of organic calcium salts, such as calcium gluconate. As the temperature rises, there is a greater dissociation, rendering more free calcium ions available to complex with phosphate salts. This phenomenon is not associated with inorganic calcium chloride salts, as they are almost fully ionized in aqueous solution. A temperature change from 5°C (41°F) to body temperature at 37°C (98.6°F) has been shown to promote formation of calcium phosphate crystals.¹²⁰ Precipitation from temperature increases may also be related to the shift in the phosphate equilibrium from monobasic to dibasic salts.¹¹²

Sequence of calcium and phosphate additions

The order of mixing calcium and phosphate into PN admixtures can affect the final solubility profile. Inorganic phosphate injections should be added early in the mixing sequence, and calcium gluconate injections should be added nearly last to what should be the most dilute phosphate concentration possible. The official pH range of calcium gluconate injection is 6.0–8.2, and the formulation is more likely to be mildly alkaline than acidic. Thus, it is reasonable to add calcium gluconate at the end of the sequence, so it is most dilute and least likely to influence the overall pH of the formulation. Increasing hypertonicity or osmolarity of the PN formulations, such as those infused into central veins, may also improve calcium and phosphate solubility. High concentrations of amino acids may sequester calcium and decrease its complexation with phosphate salts, whereas higher concentrations of dextrose can reduce the final admixture pH.¹¹⁵

Duration of storage

The duration of time lapsed after compounding has been shown to impact the process of precipitation. Although precipitates may form during the compounding process because of poor mixing procedures, it is more typical for dibasic calcium phosphate crystallization to progress through a time-related induction period. Precipitation of calcium phosphate during PN infusion may not be observed until 12 or more hours after preparation, and a delayed onset of precipitate formation may take up to 24–48 h.^{112, 113} The variable rates in precipitate formation have been attributed to slow crystallization from supersaturated mixtures.¹¹⁷

Admixture compatibility—Focus on trace elements and vitamins

Shortly after the initial American Medical Association (AMA) published recommendations on parenteral trace elements, a study evaluated the compatibility of individual trace elements (chromium chloride, copper sulfate, manganese sulfate, zinc sulfate), as well as the combination of the four trace elements, while accounting for metal contaminants in all components used.¹²¹ The solutions tested included amino acids alone (Travasol 8.5% [pH 5.7]) and a PN formulation (Travasol AA4.25%-D25% [pH 5.75]) in glass with electrolytes and multivitamins. Over the 48-h study period, the initial concentration of each trace element was unchanged, based on atomic absorption spectrophotometry, and infusion through a 0.22-µm filter did not reveal any visible precipitate. Of note, each individual trace element led to some slight color change when admixed with amino acids alone. The incompatibility of copper with some amino acids, resulting in precipitates, was described in an earlier section.¹¹²

In a study of pediatric PN (Vaminolac AA2.75%-D5.64%) that included calcium chloride and glycerophosphate, the addition of multi–trace elements (Pediatrace) led to an unexpected ongoing precipitate within 4 h of admixture, which amounted to ∼14,000 particles/ml, mostly smaller than 1 micron.¹²² The fine subvisual particles were noted using Tyndall light, light obscuration, and turbidimetry and proposed to be the result of incompatibility between copper and cysteine. This was not analyzed further, and additional factors such as pH, contaminant minerals, redox conditions, mixing sequence, temperature, and container type would need to have been taken into account.

Another study evaluated the compatibility of copper in two cysteine-containing pediatric PN admixtures (TrophAmine AA2.4%-D25% plus cysteine 960 mg/L and copper 0.3 mg/L, and AA3.7%-D16% plus cysteine 1480 mg/L and copper 0.45 mg/L) over a 24-h simulated infusion through infusion sets with 0.22-µm inline filters.¹²³ Aside from visual signs of incompatibility, including precipitation, samples were obtained at several time points throughout the infusion and assayed for both cysteine and copper. The first solution exhibited >10% reduction in cysteine and in copper, whereas the second solution had an 8% reduction in cysteine concentration and 11% reduction in copper concentration over the study period. However, no visible particulate matter or discoloration was observed on the filters. Previous work had noted cupric sulfide precipitates that discolored inline filters when using an adult bisulfite-free amino acid product (Novamine AA4.25%) plus cysteine (200 mg/L) and copper (0.88 mg/L).¹²⁴ Switching to a bisulfite-containing amino acid product resulted in no precipitate formation.¹²⁴ A study to examine mechanisms compared a bisulfite-containing but cysteine-free amino acid product (Travasol) with a cysteine solution (400 mg/L), following the addition of copper (1.2 mg/L), and noted significant (98%) reduction of copper concentration only in the cysteine solution.¹²⁵

Early studies suggested that insoluble iron phosphate precipitation can occur in stored PN solutions devoid of multivitamins.⁴² The mechanism is unclear but appears to be inversely related to the amino acid concentration and the duration of storage prior to the addition of the multivitamins. Amino acid brand may also play a role with precipitates seen more with Synthamin (identical to Travasol at the time).⁴² No similar reports have been noted using other amino acid products available in the United States.

Others have reported incompatibility of iron dextran with lower-volume neonatal PN solutions. Incompatibility may be related to the final amino acid concentration, although it is uncertain whether the incompatibility is due to the pH or related to the nature of the amino acid solution used in the PN. In their investigation, Mayhew et al evaluated 20 neonatal PN admixtures that varied in their amino acid concentrations (TrophAmine 0.5%–2.5%) and otherwise contained dextrose 10% as well as micronutrients, heparin and cysteine.¹²⁶ Iron dextran was included at 10 mg/L and control solutions omitted either the iron, trace elements, or multivitamin component and all were stored at 19°C for up to 48 h and observed for precipitate, color change, turbidity, and phase separation. Iron precipitate (rust colored) formed within 12–24 h in those PN admixtures with ≤1.5% amino acids. The authors concluded that lipid-free neonatal PN solutions with a final amino acid concentration of 2% or greater were visually compatible with iron dextran 10 mg/L for 48 h. These findings suggest that adding iron dextran to a peripheral PN solution, which typically has a lower final concentration of amino acids, would be less stable and not recommended.¹²⁶

In a more recent retrospective study, Lee et al evaluated the safety of mixing low-MW (LMW) iron dextran in a cohort of 89 pediatric patients (0–21 years of age) who received PN containing a maintenance dose of LMW iron dextran with a total of 2774 days of exposure (1–196 days per patient).¹²⁷ The mean dose of iron dextran in children decreased with increased weight from <5 kg (0.21 ± 0.05 mg/kg/day) to ≥40 kg (1.9 ± 0.5 mg/day). The authors concluded that PN containing a maintenance dose of LMW iron dextran could be safely administered to hospitalized children with respect to anaphylaxis risk, and further studies would be needed to evaluate the potential to prevent iron deficiency anemia and the need for additional IV iron infusions.¹²⁷ Unfortunately, an evaluation of risks due to potential physicochemical interactions was not performed.

The multivitamin products are compatible in dextrose 5% in water, 0.9% sodium chloride, and PN admixtures according to the manufacturers.^84-87 The compatibility is contingent on the individual vitamins. The vitamins may react with each other, macronutrients, excipients, or the PN bag material.¹¹² For example, vitamin A has long been known to adsorb to plastic materials, making it less available for the patient.¹¹² Sorption to the bag (ie, adsorption) is more likely to occur with retinol acetate than with retinol palmitate because of its greater affinity to plastics.¹²⁸ This has become less of a concern with the use of PVC-free bags and administration sets and the switch from the acetate ester to the less reactive palmitate ester.^{129, 130} As a difficult vitamin to incorporate into parenteral products, vitamin D may bind to plastic containers and infusion sets, but over 65% of the content is expected to remain available.¹³¹ It remains unclear whether there is a significant difference between ergocalciferol and cholecalciferol.

STABILITY

Focus on macronutrients (alone or combined)

Amino acids

Amino acid degradation may be general or specific to an amino acid side chain. General degradation can generate ammonia and carbon dioxide through deamidation/deamination in the presence of elevated temperature. The amino acid l-glutamine is excluded from products because of rapid degradation to toxic by-products upon heat sterilization. The presence of oxygen and/or light will also influence amino acid stability. Higher degradation occurs in plastic than in glass containers, likely associated with greater oxygen permeability.

Nonionizing radiation (eg, ultraviolet or visible light) may cause some minimal decomposition but, in the presence of a chromophore (functional group or substance), may generate oxidizing species (eg, hydrogen peroxide), which can further degrade amino acids. Most susceptible to photo-oxidation are cysteine, histidine, methionine, phenylalanine, tryptophan, and tyrosine. Reactive side chains of cysteine, tryptophan, and tyrosine make them the most susceptible to oxidation. Photo-oxidation of amino acids (eg, tryptophan, tyrosine) is enhanced in the presence of riboflavin (a chromophore). A study of a 1% concentration of pediatric amino acids product in the presence or absence of riboflavin and metabisulfite (an antioxidant preservative) exposed to phototherapy light (425–475 nm wavelength) generated hydrogen peroxide.¹³² Histidine, methionine, and tryptophan undergo photo-oxidation in ambient light in the presence of riboflavin with some protective effect in the presence of adequate ascorbic acid.^{133, 134} Tryptophan is an interesting case because multiple oxidation reactions influence this amino acid with several different degradation products. The indole moiety of tryptophan is prone to ring opening from oxidation with electrophilic substitution. The degradation of tryptophan over time may be associated with yellowing of the amino acid solution.¹³⁵ An AA4.25%-D25% solution stored for 2–4 weeks did not experience significant (>10%) degradation of amino acids, including tryptophan, the most labile component, especially under refrigeration (4°C).^{100, 101}

Lipid emulsions

When the ILE is incorporated into the PN admixture, the entire preparation becomes an oil-in-water emulsion in character. As a result, any evaluation of PN admixture stability needs to consider the physical integrity of the emulsion. In general terms, the final macronutrient concentrations that are expected to provide the best likelihood of a stable emulsion in the PN admixture, for up to 30 h (at 25°C) or 9 days (at 5°C), are amino acids ≥4%, dextrose ≥10%, and ILE ≥2%.⁴ This recommendation was based on data using 100% soybean-oil ILE products and varies with ILE products containing other source oils.

Extemporaneously prepared TNAs (ie, 3-in-1 PN admixture) are expected to be less stable than the ILE product. The compendia PFAT₅ limit of 0.05% might be too strict; hence, the use of PFAT₅ <0.4% (one log higher) has been proposed as the acceptance criterion, even though studies do indicate that it is possible to compound a PN formulation that fulfills the PFAT₅ limit of USP. Driscoll et al demonstrated that when the PFAT₅ level exceeded 0.4%, obvious phase separation or “cracking” occurred, reflecting the irreversible terminal stage of emulsion instability.¹¹⁴

An evaluation was performed of a pediatric amino acid product (TrophAmine 10%) in a TNA using ILE 20% (Intralipid, Liposyn II, or Nutrilipid) and dextrose 70% with micronutrients at several final macronutrient concentrations for 24 h at 4°C and an additional 24 h at room temperature.¹⁰⁸ Although evidence of creaming was observed in all sample PN admixtures over time, no coalescence or phase separation was visible. Particle size analysis revealed that mean diameters remained <0.4 µm; however, greater numbers of large lipid particles (>5 µm) were present with higher concentrations of electrolytes, representing as much as 0.6% of oil droplets.¹⁰⁸ This value is well above the current compendia PFAT₅ limit of 0.05% or the proposed limit of 0.4%, indicating an unstable emulsion.

PN admixture stability is determined by how well lipid droplets remain dispersed in the external (aqueous) phase, as cations and lower pH values decrease the electric charge of the emulsifier tasked with preventing aggregation. With the advent of TNAs, studies emerged that described the influence of trace elements on emulsion stability. One study compared organic (gluconate) with inorganic (chloride) salt forms of trace elements (copper 240 mcg/L, iron 0.5 mg/L, zinc 2 mg/L) with a PN admixture (Azonutril-25 AA2%-D17.5%-L5% Intralipid) stored at 4°C or 25°C for up to 1 week.¹³⁶ Lipid particle measures relied on Coulter counter analysis capable of including diameters from 1 to 25 µm. The gluconate salts did not significantly alter admixture pH, whereas chloride salts decreased pH slightly over the course of the study. Although the numbers and sizes of lipid droplets were not significantly different between them, those in the PN admixture with the chloride salts had more broadly dispersed values over time risking tendency towards emulsion destabilization.

In an early study of emulsion stability of a PN admixture, the influence of two different multimineral solution additives was evaluated.¹³⁷ The PN admixture (Travasol AA3.34%-D12.5%-L3.34% Intralipid) in large ethylene–vinyl acetate (EVA) bags included either a commercial additive (Addamel) or a compounded eight-component additive, but no vitamins, stored for up to 2 weeks at 4°C. Of note, Addamel is not currently approved for use in the United States and is only marketed elsewhere. The lipid droplet size, determined by Coulter counter (able to assess particle diameters between 0.6 and 20.2 µm), did not vary significantly in distribution during storage, nor did the proportion of large (>5 µm) to small (<2 µm), with mean droplet size of ∼0.674–0.684 µm, all of which was no different than the control of the ILE alone.¹³⁷ A study of a 3.1-L PN admixture (Vamin-N AA3.4%-D16%-L1.6% Intralipid), which included electrolytes, multi–trace elements, and multivitamins, was performed (pH of 5.5 with osmolality ∼1472 mOsm/kg) over 14 days at 4°C plus 2 days at 22°C using visual observation, light microscopy, electron microscopy, and Coulter counter techniques.¹³⁸ These revealed an increase in particle size over time, including some exceeding 2 and 5 µm in diameter. The amino acid and dextrose content of a TNA alone is unlikely to be adequate to overcome the influence of high divalent ion concentrations in a PN admixture.¹³⁹

The multivalent cations such as calcium and magnesium, as well as high concentrations of the multivalent copper, iron, and zinc, have the potential to alter the stability of ILE by reducing the negatively charged repulsive forces that keep lipid droplets separated. A recent study evaluated the influence of trace elements (copper 200 mcg, fluorine 570 mcg, iodine 10 mcg, manganese 10 mcg, selenium 60–90 mcg, zinc 3.4–3.5 mg) as well as carnitine (100 mg) on the stability of six different 1-L PN admixtures (TrophAmine AA2.3-3%–D10-14%–L1.9-2.3%; various ILE products used) for pediatric patients that included electrolytes and multivitamins in EVA bags.¹⁴⁰ These admixtures were stored at 4°C or room temperature for up to 96 h with 450 samples analyzed. Lipid droplet diameters were evaluated by laser diffraction. The diameter of droplets did not change significantly throughout the duration of the study and remained between 0.4 and 1 µm in most of the tested admixtures. An interesting exception occurred in those admixtures using a fish-oil component with calcium concentration of 4.5 mmol/L (9 mEq/L) or higher in which 2% of droplets exceeded 5 µm and 12% exceeded 1 micron immediately after compounding.¹⁴⁰ Whether the inclusion of the trace elements increased this risk is not known because no controls without trace elements (or carnitine) were studied.

Compounded PN admixtures for home infusion may rely on a dual-chamber EVA bag in which the ILE portion is stored in one chamber while the remainder of the PN ingredients are in the other for a 7-day period. In this way, any concern for the final emulsion's stability over time is deferred. Just prior to administration, the contents of both chambers are combined with the addition of multivitamins at that time. However, the emulsion stability of the final PN admixture must still be ensured. Varying contents and component concentrations will determine which PN admixtures for use at home remain stable and safe for administration and which do not.^{141, 142}

In the smaller volumes of neonatal PN admixtures, the influence of trace elements and vitamins were evaluated on formulation stability in a practice setting where trace elements and vitamins were routinely not combined in the same bag.^{143, 144} Three PN admixtures were evaluated—one with multivitamins but without multi–trace elements, another with the trace elements but not multivitamins, and the third containing all components.¹⁴³ The formulations were otherwise identical (Primene AA3%-D8%-L3% Lipofundin) in EVA bags with electrolytes that included organic phosphate. Samples were taken at predetermined points for up to 7 days, with storage conditions at 5°C, 25°C, and 40°C, for evaluation of emulsion stability using zeta potential, optical microscopy, dynamic light scattering, and PFAT₅. The osmolality of the PN admixtures was slightly higher than 1000 mOsm/kg. Slight creaming and a darkening color on visual inspection were apparent after 48 h in admixtures stored at 25°C and 40°C.¹⁴³ Zeta potential values and pH did not change significantly throughout the study period for any formulation, nor did lipid droplet diameter.¹⁴³ Most importantly, none of the formulations exceeded a PFAT₅ of 0.05%.¹⁴³

Iron, in either the bound ferric (Fe³⁺) or free ferrous form (Fe²⁺), can generate free oxygen radicals, which has raised concerns regarding its safety with IV administration. This is especially true in the presence of ILE, which serves as an excellent substrate for iron-induced peroxidation because of the large number of double bonds present in the unsaturated fatty acids.¹⁴⁵ Despite its potential benefits, iron, especially the dextran form, should not be added to ILE or TNAs, as it can result in a destabilization of the emulsion.¹⁴⁶ Similar to divalent cations (ie, calcium, magnesium), trivalent cations such as iron may cause a decrease in the surface potential of the lipid droplets, destabilizing the negative surface charge between lipid particles, resulting in aggregation and coalescence of the smaller lipid particles to form larger ones.^{114, 147} These larger lipid globules can be potentially dangerous if infused, as they can result in fat emboli, especially in neonates requiring the iron. In a study involving the necropsy findings of nearly 500 live-born infants (including 30 of 41 PN patients who had received ILE), Puntis and Rushton found intravascular lipids in the small pulmonary capillaries of 15 infants receiving ILE that were not found in the necropsy of enterally fed infants.¹⁴⁸ This group had received significantly more ILE during their PN course (in total amount [grams per kilogram] and duration [number of days]) compared with another group of 15 infants receiving ILE who did not have lipid staining.¹⁴⁸

Focus on micronutrients (alone or combined)

Trace elements

Trace elements may interact with other micronutrients. For example, copper and ascorbic acid interact with each other. Copper can accelerate the degradation of ascorbic acid in solution, especially at pH >4. Conversely, higher concentrations of ascorbic acid can reduce copper to the cuprous form, which is why ascorbic acid content of adult multivitamins was reduced to 100–200 mg from 500 mg.

A study evaluated trace element stability in a 1000-ml PN admixture (Travasol AA3%-D18%) in a DEHP-free but nevertheless semipermeable bag, using an additive of six multi–trace elements and multivitamin with high ascorbic acid (1000 mg) content.¹⁴⁹ Using ICP-MS, trace element concentrations were reported to decline over time (36 h or 30 days) with storage (at 4°C or 20°C protected from light). Of the trace elements formulated into the PN admixture, three (Cu, Mn, Zn) decreased significantly in concentration over the storage time, even when refrigerated.

The influence of copper on ascorbic acid stability was studied in 1-L adult PN admixtures (AA5%-D25%) containing electrolytes and multivitamins (100 mg ascorbic acid), with or without trace elements (1.2 mg copper) in PVC bags at room temperature.¹⁵⁰ Samples were taken at time points from baseline up to 36 h to assay ascorbic acid content (using both spectrophotometry and high-performance liquid chromatography [HPLC] with ultraviolet detection). Based on the preferred HPLC method, ascorbic acid concentrations had diminished to 67% of baseline when no copper was present but decreased to 33% of baseline in the presence of copper by 24 h. The generation of the ascorbic acid degradation product oxalic acid was not quantified in this study.

Both ascorbic acid and copper have the potential ability to reduce selenite to the poorly soluble elemental selenium.¹⁵¹ This may be more likely at ascorbic acid concentrations >100 mg/L in PN admixtures with pH of 5 or lower or at ascorbic acid doses above 500 mg, and it may be less likely in well-buffered admixtures with pH >5.^151-157

After selenium was recognized as a critical trace mineral to include in PN admixtures, several studies evaluated its stability, in particular given the potential risk for reduction to insoluble elemental selenium by ascorbic acid. When only ascorbic acid (100 mg/L) and selenious acid (100 mcg/L) were combined in solution, a 90% loss of selenium occurred by 24 h, with higher concentrations (500 mg/L) of ascorbic acid responsible for near-complete loss of selenium within 1 h using a thin-layer electrophoresis method.¹⁵⁵ The incorporation of selenious acid into a PN admixture (Travasol AA4.25%-D25%) that included electrolytes, a product with four trace elements, and multivitamins with 100 or 500 mg of ascorbic acid revealed minimal selenium loss at the lower amount but over 10% loss with ascorbic acid 500 mg at 24 h. The lower losses in the PN admixture were attributed to the amino acid content of the admixture and its influence on overall pH. Of note, other studies either did not determine the chemical form of selenium or did not evaluate the interaction using a PN admixture.^152-154

When added to a 1.85-L PN admixture (Travasol AA5%-D14.3%-L5.7% Intralipid) with electrolytes, multi–trace elements, and multivitamins in an EVA bag, the inclusion of iron dextran (100 mg) was associated with repeated filter obstruction during infusion accompanied by a yellow-brown fluid layer floating on top of the mix.¹⁴⁶ That supernatant contained much higher lipid concentrations than expected from the homogeneous mix. This was not repeated with an identical PN admixture, which excluded the iron dextran. When much lower concentrations of iron dextran (2 mg/L) were incorporated into simulated 0.5-L PN admixture (Travasol AA0.55-2.2%–D0.4-10%–L2-4% Intralipid or Liposyn II) in glass bottles without added electrolytes, trace elements, or vitamins and stored at 4°C or 25°C protected from light for 48 h, no visually apparent precipitation or phase separation was noted compared with controls.¹⁵⁸ Photon correlation spectroscopy suggested that 90% of lipid droplets were smaller than 1.3 µm, although PFAT₅ was not performed.

In one investigation evaluating the stability of iron sucrose added to lipid-free PN containing pediatric amino acids (TrophAmine AA2.6%-D20% [pH 5.7]; TrophAmine AA3.5%-D20% [pH 5.5]) with electrolytes and trace elements, it was shown that the physical stability of iron sucrose in PN is time and concentration dependent, based on visual and microscopic inspection.¹⁵⁹ Iron sucrose concentrations up to 2.5 mg/L in lipid-free PN compounded with neonatal-range amino acid concentrations and cysteine were physically stable. Iron sucrose concentrations of ≥10 mg/L were physically unstable, with particulate formation within 12 h and evidence of gross physical incompatibility within 24 h.¹⁵⁹ In the PN admixture containing a lower final amino acid concentration, visible particulates manifest by 8 h at the highest iron concentrations and within 24 h for all lower iron concentrations, with crystals determined to be iron complexes including calcium and phosphate.¹⁵⁹

Vitamins

The stability of multivitamin solutions depends on the individual components. The instability of vitamins can be due to oxidation, photodegradation, or exposure to container materials.¹⁶⁰ The main factors affecting the stability of the individual vitamins and the excipients include light exposure, pH, temperature, and PN bag material (eg, plastic or glass).¹¹² Light-induced peroxidation reactions may occur with polysorbate 80 and 20.⁹² This reaction is enhanced at higher temperatures and upon exposure to air.¹⁶¹

When exposed to ultraviolet light, riboflavin has been shown to catalyze oxidation reactions.^{162, 163} Riboflavin can act as a photosensitizer initiating production of reactive oxygen species or directly reacting with other components (eg, amino acids, fatty acids).^164-166 Ultraviolet light activates riboflavin-catalyzed oxidation reactions with some amino acids.¹³³ Exposure to 24 h of ultraviolet light was associated with a statistically significant decrease in the methionine, proline, tryptophan, and tyrosine concentrations in a 2-in-1 PN admixture containing a riboflavin concentration of 1 mg/dl.¹³³ Riboflavin itself can be degraded to luminoflavin, luminochromo, and other compounds in the presence of oxygen and light, including sunlight and artificial light.^167-169 The concentration of riboflavin in a 2-in-1 PN admixture decreased by 47% and 100% upon exposure to 8 h of indirect and direct sunlight, respectively.¹⁷⁰

PN admixtures with pH values above 5.0 reduce the risk for precipitation of folic acid.¹⁷¹ Results from stability studies are conflicting with regards to the effects of storage container, light exposure, and temperature.¹¹² In an early study evaluating the safety of vitamin addition to PN, Chen et al (using a microbiological assay, which was standard at the time for several B vitamins) found that folic acid was stable in a 1-L PN admixture (Travasol AA4.25%-D25%) containing electrolytes, trace elements, and a nine-vitamin formula plus 1 mg folic acid when exposed to fluorescent light, indirect sunlight, or direct sunlight at room temperature for 8 h.¹⁷⁰ When the PN was stored frozen (−40°C) for 5 weeks and then refrigerated (4°C) for another 2 weeks, folic acid maintained stability.¹⁷⁰ This study also revealed that riboflavin, pyridoxine, and thiamin were unstable in sunlight (direct or indirect) over that 8-h period.¹⁷⁰ Precipitation forms at a folic acid concentration of 0.2 mg/L in the low pH, at a high concentration of dextrose 40% at room temperature, and under refrigeration.¹⁷² Direct sunlight was found to degrade pyridoxine by 86% in a 2-in-1 PN admixture after 8 h.¹⁷⁰

Thiamin degradation is dependent on several factors such as the sulfite concentration, pH, temperature, exposure to direct sunlight, and presence of oxygen.¹¹² The stability of thiamin is less of an issue now that bisulfites are rarely included as an antioxidant/preservative in other PN components (eg, amino acid solutions).¹¹² The irreversible degradation of thiamin in solution into a thiazole and pyrimidine by sodium metabisulfite proceeds in a concentration- and pH-dependent manner.¹⁷³ It should be noted that combining the amino acid and dextrose solutions results in a decreased sulfite concentration, minimizing the risk of thiamin degradation.¹⁷⁴

A study conducted by Scheiner et al found that 93% of the initial thiamin amount remained after 24 h in a PVC bag containing a dextrose-electrolyte solution (Normosol M-D5%) and multivitamins, with a pH of ∼5 (pH range, 4.0–6.5; 363 mOsm/L) and storage temperature of 23°C in the presence of fluorescent light.¹⁷³ The labeling of the Normosol M-D5% solution indicated bisulfites as an ingredient, but they were not detected. The authors concluded that the increased oxygen permeability of the PVC bag (as compared with the glass container) accelerated the loss of sulfites, thereby decreasing the potential for destruction of thiamin.¹⁷³ When Normosol M-D5% with multivitamins was stored in a glass container (under the same room temperature, pH, and light conditions), the amount of thiamin decreased to 51% relative to the original amount after 24 h.¹⁷³ Only 3% of the initial thiamin remained in an undiluted FreAmine III 8.5% glass-bottle solution consisting of a high concentration of bisulfites and with a pH of 6.5.¹⁷³ A later study by Smith et al revealed that thiamin degradation was the greatest in 2-in-1 and 3-in-1 PN admixtures (containing electrolytes, trace elements, and multivitamins without vitamin K) with a final bisulfite concentration of 3 mEq/L and pH of 6.4 at a temperature of 25°C compared with other solutions at various pH and bisulfite concentrations.¹⁷⁵ Bowman et al found that no thiamin degradation was detected for 22 h in a 1-L PN admixture (AA4.25%-D25% with sulfite 0.05%) contained in a plastic bag.¹⁷⁴ The PN bags were stored at a temperature of 30–31°C, and the average pH of the PN solution was 5.8.¹⁷⁴

The main determinants of thiamin degradation are the presence of sulfites and direct sunlight. Refrigeration has been shown to slow thiamin loss.¹⁷³ When exposed to direct sunlight, the thiamin concentration in a 2-in-1 PN admixture decreased by 26% after 8 h.¹⁷⁰ But in a sulfite-free 2-in-1 PN admixture remaining refrigerated until use, at least 75% of a large thiamin dose (50 mg) was maintained at 28 days.¹⁷⁶ Patients should be monitored for signs and symptoms of thiamin deficiency if there is a concern for thiamin loss from the PN admixture.

Ascorbic acid is the vitamin most susceptible to degradation, undergoing reversible oxidation to dehydroascorbic acid catalyzed in part by trace elements (ie, copper, manganese, zinc), which can then undergo hydrolysis to diketogulonic acid.^{112, 177, 178} The latter undergoes irreversible oxidation to several metabolites, including oxalic acid.¹¹² Of note, oxalic acid can interact with calcium, forming a poorly soluble complex with risk for precipitation. This decomposition of ascorbic acid is dependent on the content of oxygen available in the PN admixture from bag permeability, dissolved air in solution, and other additives.¹⁷⁷ The degradation is more likely at pH >4 and at higher temperatures (22–35°C).^{179, 180} The oxidation reactions are primarily due to the introduction of oxygen into the PN admixture during the compounding process, through permeation of the PN container, or during infusion.^{112, 162, 181} Smith et al found that ascorbic acid degradation was more pronounced in plastic containers compared with glass containers.¹⁷⁵ The use of multilayered bags to reduce oxygen permeability has been recommended to reduce ascorbic acid degradation, with the caveat that they may trap oxygen transferred into the bag during filling. The monolayered EVA bag is more permeable to oxygen than a multilayered bag, which allows for a greater potential for reactivity with ascorbic acid.¹⁸¹ Inclusion of cysteine, by complexing with copper, may reduce the degradation of ascorbic acid.¹⁸² However, cysteine has a reducing capacity able to irreversibly degrade ascorbic acid/dehydroascorbic acid.¹⁸³

Vitamins may also interact with other micronutrients. A specific evaluation of several B vitamins using nuclear magnetic resonance spectroscopy evaluated the interaction within this group and with other micronutrients.¹⁴⁴ Interactions between the vitamins, especially pyridoxine, reflect proton exchange phenomena as opposed to intermolecular reactions. Pyridoxine and riboflavin are sensitive to the presence of electrolytes—especially pyridoxine in the presence of calcium gluconate and phosphate—and, together with magnesium sulfate, may suggest some self-aggregation of pyridoxine. Divalent cations (calcium, magnesium, and copper and zinc as chloride salts) may also lead to some self-aggregating precipitate of riboflavin, although the presence of nicotinamide seems to inhibit riboflavin aggregation. Notwithstanding these observations, hydrophilic vitamins and minerals in the same PN admixture could be stable for up to 48 h.¹⁴⁴

Studies investigating the effect of riboflavin on ascorbic acid oxidation and in combination with ILE are conflicting. The manufacturers recommend against the direct injection of multivitamins into ILE.^84-87 In the presence of ultraviolet light, riboflavin catalysis of ascorbate to dehydroascorbate, generating hydrogen peroxide and free radicals, has been described in the literature.^{162, 163} The hydrogen peroxide may be the result of lipid and/or ascorbic acid oxidation.¹⁶³ Laborie et al found that ILE containing riboflavin (0.11 mmol/L) and ascorbic acid (13.6 mmol/L) had higher amounts of peroxide production compared with ILE alone or with riboflavin only.¹⁶³ Contrary to the Laborie et al results, Silvers et al found that the hydrogen peroxide concentration decreased by 40% in ILE with multivitamins compared with dextrose 5% in water–0.45% sodium chloride with multivitamins in an ambient light environment.¹⁶² The authors noted that oxidation of ascorbic acid still occurred but that ascorbic acid protected against lipid peroxidation. The major source of light and oxygen exposure occurred during the infusion of the admixture through the catheter. Furthermore, the opaque color of the ILE likely attenuated light-induced oxidation of ascorbic acid.¹⁶² Silvers et al suggested that the findings by Laborie et al were based on unreliable assay methods. Considering the beneficial effect of ascorbic acid on ILE, the precaution against direct injection of multivitamin solutions into ILE may be unnecessary.¹⁶² Conflicting with previous findings, Dupertuis et al found that multivitamins containing riboflavin did not influence the degradation rate of ascorbic acid in 3-in-1 PN admixtures exposed to daylight.¹⁸¹ To prevent the possibility of ascorbic acid loss induced by riboflavin, PN admixtures should be protected from light when practical.

Retinol is the most light-sensitive vitamin, with extensive photodegradation depending on the wavelength and intensity of light.^{169, 184, 185} The presence of ILE in the PN admixture is thought to reduce the degree of vitamin A instability, but this remains controversial.^{128, 175, 185, 186} The use of light-protective covers has been suggested.^{185, 187} Less reactive than retinol, vitamin E undergoes photo-oxidation that is dependent on wavelength and light intensity as well as the oxygen content.¹⁸⁸ Bags that prevent oxygen permeability or light-protective covers may each improve the stability of vitamin E.^{128, 187, 189} The potential degradation of retinol and vitamin E during lipid peroxidation in a PN admixture may be less when MCT oil accompanies soybean oil, compared with ILEs of a soybean-olive oil combination or soybean-oil ILE alone.¹⁹⁰ More has yet to be learned about the stability of different vitamin E isomers. Vitamin K content of the PN admixture is susceptible to photodegradation regardless of the presence of lipids.^{128, 191} Exposure to daylight decreased the phytonadione concentration in PN admixtures (2-in-1 and 3-in-1) by 50% after 3 h.¹²⁸

Given the above, multivitamin solutions are the least stable component in PN admixtures, especially when exposed to oxygen and light.¹¹² For this reason, dispensing a 7-day supply of PN admixtures at a time precludes multivitamin content and requires the patient/caregiver to add the dose each day just prior to administration. Once the multivitamins are added to the compatible diluents, the final admixture should ideally be administered immediately in the absence of photo-oxidative protection and completed within 24 h.^84-87 The vitamin-containing PN admixture should be refrigerated when administration is delayed.^84-87 To prevent degradation due to light exposure during administration, PN bags should be wrapped with an opaque protective covering.¹⁹² Also, orange or yellow infusion tubing may be used to prevent light penetration through the tubing during administration.¹⁹² The practice of photoprotection in the United States remains dependent on product availability.¹⁹³

OTHER

Aside from the common nutrient components of the PN admixture, additional compatibility and stability issues arise with other components. These include medications added to or co-infused with a PN admixture, as well as the materials found in automated compounding device circuits, the PN bag, administration sets, and inline filters.

Medication

As is already apparent, PN admixtures are complex from a compatibility and stability perspective when accounting for all the nutrient components—active ingredients and excipients. The consideration to include a nonnutrient medication in a PN admixture is to be undertaken carefully with serious attention to compatibility and stability concerns. This is true whether preparing a PN completely from individual components or starting with a commercial, standardized, multichamber PN product. Similar attention needs to be given when considering medication administration by Y-site infusion along with a concurrently infusing PN admixture. An evidence-based guidance document discussed the safety of using the PN admixture as a vehicle for nonnutrient medication delivery along with a summary of published data.⁴ Although that can be referred to for further detail, the following provides a concise summary.

Added to the PN admixture

The PN admixture itself will have been determined to be compatible and stable before introducing a medication. There could be an advantage to including a drug within the PN admixture if it can consolidate drug dosing and volume and decrease manipulations of vascular access. In order to be considered, the IV drug required by a patient should be a stable regimen that requires no dose titration and is therapeutically effective by continuous or cyclic infusion. However, this still requires reviewing the available data on compatibility and stability. Unfortunately, there are a limited number of IV drugs that have been evaluated so far, and few of them rely on currently accepted methods of study design and analysis.⁴ Specific criteria for evaluating compatibility and stability studies of medication in PN admixtures should be followed.^{113, 194, 195} Unfortunately, many published studies have relied exclusively on visual compatibility, whereas some newer publications have applied some of the criteria.

Visual physical compatibility of a medication in a PN admixture by itself is not indicative of chemical compatibility or drug stability. At the very least, drug pKa and product pH along with the presumed pH of the PN admixture are important to consider. Once the physical compatibility and chemical stability are both demonstrated, verification of pharmacologic/therapeutic effectiveness in patient care is also necessary. In other words, the drug needs to be soluble, compatible, and stable in the PN admixture as well as therapeutically effective—without altering the compatibility or stability of the initial PN admixture. A close examination of published data is valuable, when available, but extrapolation beyond study parameters (eg, component products, concentrations of each ingredient including the drug) is discouraged. Ideally, a drug would only be considered if supported by pharmaceutical data describing physicochemical compatibility and stability of the added drug and of the final PN admixture under conditions of typical use, plus clinical data confirming expected therapeutic actions of the drug.³

A recent example is a study that purported to show that 1 g of ampicillin (an anion) was stable in two TNAs for 7 days under refrigeration and then for 24 h at room temperature following addition of vitamins and trace elements.¹⁹⁶ Although methods went well beyond visual compatibility, there remained some shortcomings. For example, the drug assay was not shown to indicate stability (but drug concentration was below 90% by 48 h in storage at 4°C), there was no PFAT₅ reported, and no clinical rationale was provided as to how long the ampicillin concentration would be expected to remain above the minimum inhibitory concentration of a given bacteria during PN infusion. There was an increase of admixture pH, no zeta potential lower than −23 mV, for this low ampicillin dose studied in a 250-ml PN, with limited sampling points.¹⁹⁶

Co-infused with the PN admixture

When a medication needs to co-infuse with a PN admixture via Y-site owing to limited patient access, clinicians must consider the drug product excipients and contents of the carrier fluid as well as the medication when evaluating compatibility and stability with PN admixtures. A number of IV drugs have been studied for Y-site compatibility with compounded PN admixtures as well as with commercial, standardized, multichamber PN products.^{22, 197-205}

Studies are often performed in vitro, with small sample volumes, using various dilutions of a PN admixture and a drug to simulate the concentrations combined during an actual Y-site infusion. Given that the contact time during infusion is often <4 h, drug stability is rarely evaluated. Drug compatibility via Y-site administration has most frequently been determined by visual inspection. Within these limitations, 80% (82 of 102) of medication tested with four different 2-in-1 PN admixtures and 78% (83 of 106) of the drugs tested with nine different TNAs were revealed to be visually compatible.^{22, 197} Others revealed 80% (20 of 25) of drugs evaluated were compatible in a PN admixture.¹⁹⁸ Further, 7 of 10 medications used in pediatrics were compatible by Y-site administration with two commercial, standardized, multichamber PN admixtures.²⁰⁰ Also, drug Y-site compatibility with neonatal PN admixtures was found for 16 of 21 medications.²⁰¹ When no restrictions for infusion fluid, concentration, or time were placed, as may occur for neonates in a pragmatic study, only 5 of 131 drugs were considered compatible with two PN admixtures, based on a wide-ranging literature search.²⁰² Y-site administration of medication may rarely need to be considered for patients receiving home PN and should also be evaluated for compatibility under conditions of typical use.²⁰³

The results of compatibility studies with a drug can differ with a change of a single component of the PN admixture, including ILE or the oil source of the ILE.^{204, 205} It is important to appreciate the details (concentrations of each ingredient), assumptions (contact time of all active ingredients and excipients), and limitations of these studies. In the infrequent case that ILE is administered through a separate vascular access device from the 2-in-1 PN admixture, data on Y-site drug compatibility and stability with ILE infusion have been measured.²⁰⁶ As indicated earlier, all details need to be provided, including the pH of each component, dilutions used, and contact time of exposure with performance of PFAT₅. The contact time, including across the filter surface area, can be prolonged during low flow rates, as may be seen with infants.

Other components

Compatibility needs to account for components of the automated compounding device, the PN bag, administration set, and inline filter. Plasticizers incorporated into plastic containers, bags, and infusion sets are not chemically bonded to the plastic and have the potential to leach out. DEHP is the predominant phthalate plasticizer found in PVC containers, bags, and administration sets and is known to specifically leach into lipid-containing admixtures.^{51, 207-210} This poses a risk for toxicity, so the use of EVA or multilayer EVA/ethylene–vinyl alcohol bags is recommended for lipid-containing admixtures to flow through DEHP-free infusion sets. The EVA bags depend on vinyl acetate content to make the plastic supple, and polyolefins do not require a plasticizer for their flexibility. Beyond the direct clinical risks from the toxicant DEHP leaching into lipid-containing admixtures, this compound may also disrupt emulsion stability, likely through the release of phthalic acid (pKa = 2.89).¹¹⁴

Plasticizers are not the only component of concern. For example, flexible plastic containers, bags, and administration sets (PVC, EVA, polypropylene) contain several metals (eg, Ba, Cd, Pb, Sn, Zn) to stabilize the plastic polymers.²¹¹ These metals also have the potential to leach into the infusate. For example, a study of trace element stability in PN noted that cobalt may have leached from the administration set.¹⁴⁹ When tested using a dextrose (10%), sodium chloride (0.9%), and Tween-80 (5%) solution, in the three bag composition types, the primary metal appearing in the solution was zinc (at concentrations in the micrograms per liter range) from both PVC and EVA.²¹¹ The findings were similar when commercial sodium chloride or dextrose bags were infused through PVC administration sets, with zinc content increasing significantly (remaining at the micrograms per liter range) despite brief infusion. Although not evaluated with an amino acid solution or a complete PN admixture, given the propensity of amino acids to complex with cadmium and lead, there may be a risk for those metals leaching into the PN admixture as well.⁴⁰

SUMMARY

Maintaining the safety of patients requiring PN therapy requires a recognition of the available data and the application of pharmaceutical science to ensure the compatibility and stability of the ordered PN prescription. This pursuit requires continued vigilance by the pharmacist reviewing PN orders within the use process. The pharmacist with the requisite knowledge, skills, and experiences ensures the safety of PN regimens as they fulfill their professional responsibility. The effort occurs in collaboration with those recommending, prescribing, and preparing PN admixtures, as well as those involved in procuring all the relevant components.

CONFLICT OF INTEREST

Joseph I. Boullata is a consultant and speaker for American Regent, a consultant and speaker for B. Braun, and a speaker for Fresenius Kabi. Jay M. Mirtallo is a consultant and speaker for Fresenius Kabi. Gordon S. Sacks is an employee of Fresenius Kabi. Genene Salman has no conflict of interest. Kathleen Gura is a consultant for Fresenius Kabi; on the advisory board for B. Braun, Baxter, Fresenius Kabi, and Pfizer; and receives royalties from Fresenius Kabi. Genene Salman, Todd Canada, and Angela Maguire have no conflict of interest.

FINANCIAL DISCLOSURE

None declared.

Supporting Information

REFERENCES

1Flack HL, Gans JA, Serlick SE, Dudrick SJ. The current status of parenteral hyperalimentation. Am J Hosp Pharm. 1971; 28(5): 326-335.
CASPubMedGoogle Scholar
2Mirtallo J, Canada T, Johnson D, et al. Safe practices for parenteral nutrition. JPEN J Parenter Enteral Nutr. 2004; 28(6_suppl): S39-S70. [Erratum published in JPEN J Parenter Enteral Nutr. 2006;30(2):177].
10.1177/01486071040280S601
PubMedWeb of Science®Google Scholar
3Ayers P, Adams S, Boullata J, et al. American Society for Parenteral and Enteral Nutrition. A.S.P.E.N. parenteral nutrition safety consensus recommendations. JPEN J Parenter Enteral Nutr. 2014; 38(3): 296-333.
10.1177/0148607113511992
PubMedWeb of Science®Google Scholar
4Boullata JI, Gilbert K, Sacks G, et al. American Society for Parenteral and Enteral Nutrition. A.S.P.E.N. Clinical guidelines: parenteral nutrition ordering, order review, compounding, labeling, and dispensing. JPEN J Parenter Enteral Nutr. 2014; 38(3): 334-377.
10.1177/0148607114521833
PubMedWeb of Science®Google Scholar
5Worthington P, Balint J, Bechtold M, et al. When is parenteral nutrition appropriate? JPEN J Parenter Enteral Nutr. 2017; 41(3): 324-377.
10.1177/0148607117695251
CASPubMedWeb of Science®Google Scholar
6Boullata JI, Holcombe B, Sacks G, et al. Standardized competencies for parenteral nutrition order review and parenteral nutrition preparation, including compounding: the ASPEN model. Nutr Clin Pract. 2016; 31(4): 548-555.
10.1177/0884533616653833
PubMedWeb of Science®Google Scholar
7Ayers P, Boullata J, Sacks G. Parenteral nutrition safety: the story continues. Nutr Clin Pract. 2018; 33(1): 46-52.
10.1002/ncp.10023
PubMedWeb of Science®Google Scholar
8Allen LV, ed. Remington: The Science and Practice of Pharmacy. 22nd ed. Pharmaceutical Press; 2013.
Google Scholar
9Barnett MI, Cosslett AG, Duffield JR, et al. Parenteral nutrition: pharmaceutical problems of compatibility and stability. Drug Saf. 1990; 5(suppl 1): 101-106.
10.2165/00002018-199000051-00016
CASPubMedGoogle Scholar
10Trissel LA. Compounding our problems. Am J Hosp Pharm. 1994; 51(12): 1534.
Google Scholar
11Allwood MC. Reflections on current issues concerning the stability of parenteral nutrition mixtures. Nutrition. 2002; 18(7-8): 691-692.
10.1016/S0899-9007(02)00788-8
PubMedWeb of Science®Google Scholar
12Salman G, Boullata JI, Sacks GS, Tellez-Corrales E. Call to Action: the need for enhanced nutrition support education and training in pharmacy schools and postgraduate programs. J Am Coll Clin Pharm. 2021; 4(5): 612-615.
10.1002/jac5.1401
Google Scholar
13Allwood MC. Compatibility and stability of TPN mixtures in big bags. J Clin Hosp Pharm. 1984; 9(3): 181-198.
10.1111/j.1365-2710.1984.tb01076.x
CASPubMedGoogle Scholar
14 ME Aulton, KMG Taylor, eds. Aulton's Pharmaceutics. 4th ed. Churchill Livingstone-Elsevier; 2013.
Google Scholar
15Florence AT, Attwood D. Physicochemical Principles of Pharmacy. 5th ed. Pharmaceutical Press; 2011.
Google Scholar
16Mirtallo JM. Parenteral formulas. In: JL Rombeau, RH Rolandelli, eds. Clinical Nutrition: Parenteral Nutrition. 3rd ed. WB Saunders; 2001: 118-139.
Google Scholar
17Tovsen ML, Smistad G, Bjerke TM, Tønnesen HH, Kristensen S. Physicochemical stability of emulsions and admixtures for parenteral nutrition during irradiation by glass-filtered daylight at standardized conditions. PDA J Pharm Sci Technol. 2015; 69(3): 346-354.
10.5731/pdajpst.2015.01036
PubMedGoogle Scholar
18Murphy S, Craig DQ, Murphy A. An investigation into the physical stability of a neonatal parenteral nutrition formulation. Acta Paediatr. 1996; 85(12): 1483-1486.
10.1111/j.1651-2227.1996.tb13956.x
CASPubMedWeb of Science®Google Scholar
19Jeppsson RI, Sjöberg B. Compatibility of parenteral nutrition solutions when mixed in a plastic bag. Clin Nutr. 1984; 2(3-4): 149-158.
10.1016/0261-5614(84)90018-9
CASPubMedGoogle Scholar
20Ang SD, Canham JE, Daly JM. Parenteral infusion with an admixture of amino acids, dextrose, and fat emulsion solution: compatibility and clinical safety. JPEN J Parenter Enteral Nutr. 1987; 11(1): 23-27. [Erratum published in JPEN 1987;11(3):274].
10.1177/014860718701100123
CASPubMedWeb of Science®Google Scholar
21Hardy G, Puzovic M. Formulation, stability, and administration of parenteral nutrition with new lipid emulsions. Nutr Clin Pract. 2009; 24(5): 616-625.
10.1177/0884533609342445
CASPubMedWeb of Science®Google Scholar
22Trissel LA, Gilbert DL, Martinez JF, Baker MB, Walter WV, Mirtallo JM. Compatibility of medications with 3-in-1 parenteral nutrition admixtures. JPEN J Parenter Enteral Nutr. 1999; 23(2): 67-74.
10.1177/014860719902300267
CASPubMedWeb of Science®Google Scholar
23Staven V, Wang S, Grønlie I, Tho I. Development and evaluation of a test program for Y-site compatibility testing of total parenteral nutrition and IV drugs. Nutr J. 2016; 15(1): 29.
10.1186/s12937-016-0149-x
PubMedWeb of Science®Google Scholar
24Wormleighton CV, Catling TB. Stability issues in home parenteral nutrition. Clin Nutr. 1998; 17(5): 199-203.
10.1016/S0261-5614(98)80059-9
CASPubMedWeb of Science®Google Scholar
25Lumpkin MM. Safety alert: hazards of precipitation associated with parenteral nutrition. Am J Health-Syst Pharm. 1994; 51(11): 1427-1428.
10.1093/ajhp/51.11.1427
CASGoogle Scholar
26Shay DK, Fann LM, Jarvis WR. Respiratory distress and sudden death associated with receipt of a peripheral parenteral nutrition admixture. Infect Control Hosp Epidemiol. 1997; 18(12): 814-817.
10.1086/647550
CASPubMedWeb of Science®Google Scholar
27 Aminosyn-PF 7% and Aminosyn-PF 10% amino acid injection. Package insert. ICU Medical; 2018.
Google Scholar
28 Aminosyn II 10% and Aminosyn II 15% amino acid injection. Package insert. Hospira; 2010.
Google Scholar
29 Clinisol 15% amino acid injection. Package insert. Baxter Healthcare; 2018.
Google Scholar
30 FreAmine III 10% amino acid injection. Package insert. B. Braun Medical; 2020.
Google Scholar
31 HepatAmine 8% amino acid injection. Package insert. B. Braun Medical; 2020.
Google Scholar
32 Plenamine 15% amino acids injection Package insert. B. Braun Medical; 2020.
Google Scholar
33 Premasol 6% and Premasol 10% amino acid injection. Package insert. Baxter Healthcare; 2014.
Google Scholar
34 ProSol 20% amino acid injection. Package insert. Baxter Healthcare; 2014.
Google Scholar
35 Travasol amino acid injection. Package insert. Baxter Healthcare; 2020.
Google Scholar
36 TrophAmine 6% and TrophAmine 10% amino acid injection. Package insert. B. Braun Medical; 2020.
Google Scholar
37Hardy G, Ball P, McElroy B. Basic principles for compounding all-in-one parenteral nutrition admixtures. Curr Opin Clin Nutr Metab Care. 1998; 1(3): 291-296.
10.1097/00075197-199805000-00010
CASPubMedGoogle Scholar
38do Nascimento PC, da Marques MS, Bohrer D, de Carvalho LM, Carvalho CW. Chromatography and atomic absorption spectrometry for the assessment of heavy metal distribution among amino acids used in parenteral nutrition formulations: studies with cadmium and lead. Biol Trace Elem Res. 2011; 141(1-3): 76-85.
10.1007/s12011-010-8720-3
PubMedWeb of Science®Google Scholar
39Olson LM, Wieruszewski PM, Jannetto PJ, et al. Quantitative assessment of trace-element contamination in parenteral nutrition components. JPEN J Parent Enteral Nutr. 2019; 43(8): 970-976.
10.1002/jpen.1668
CASPubMedWeb of Science®Google Scholar
40Unger N, Holzgrabe U. Stability and assessment of amino acids in parenteral nutrition solutions. J Pharm Biomed Anal. 2018; 147: 125-139.
10.1016/j.jpba.2017.07.064
CASPubMedWeb of Science®Google Scholar
41Harraki B, Guiraud P, Rochat M, Alary J, Favier A. Interactions related to trace elements in parenteral nutrition. Pharm Acta Helv. 1995; 70(4): 269-278.
10.1016/0031-6865(95)00034-8
CASPubMedGoogle Scholar
42Allwood MC, Martin H, Greenwood M, Maunder M. Precipitation of trace elements in parenteral nutrition mixtures. Clin Nutr. 1998; 17(5): 223-226.
10.1016/S0261-5614(98)80063-0
CASPubMedWeb of Science®Google Scholar
43Thibault M. Possible incompatibility between amino acids and copper in solutions for pediatric parenteral nutrition. Can J Hosp Pharm. 2014; 67(2): 160-164.
PubMedGoogle Scholar
44Foinard A, Perez M, Barthélémy C, et al. In vitro assessment of interaction between amino acids and copper in neonatal parenteral nutrition. JPEN J Parenter Enteral Nutr. 2016; 40(6): 827-834.
10.1177/0148607115571967
CASPubMedWeb of Science®Google Scholar
45Sweetman SC. Martindale: The Complete Drug Reference. 34th ed. Pharmaceutical Press; 2005.
Google Scholar
46Gennaro AR. Remington: The Science and Practice of Pharmacy. 20th ed. Lippincott Williams & Wilkins; 2000.
Google Scholar
47 Dextrose injection. Package insert. Baxter Healthcare; 2014.
Google Scholar
48 Dextrose–dextrose monohydrate injection, solution. Package insert. Fresenius Kabi; 2018.
Google Scholar
49Barat AC, Harrie K, Jacob M, Diamantidis TG, McIntosh NL. Effect of amino acid solutions on total nutrient admixture stability. JPEN J Parenter Enteral Nutr. 1987; 11(4): 384-388.
10.1177/0148607187011004384
CASPubMedWeb of Science®Google Scholar
50Didier ME, Fischer S, Maki DG. Total nutrient admixtures appear safer than lipid emulsion alone as regards microbial contamination: growth properties of microbial pathogens at room temperature. JPEN J Parenter Enteral Nutr. 1998; 22(5): 291-296.
10.1177/0148607198022005291
CASPubMedWeb of Science®Google Scholar
51 National Toxicology Program, Center for the Evaluation of Risks to Human Reproduction. NTP-CERHR Monograph on the Potential Human Reproductive and Developmental Effects of Di(2-Ethylhexyl) Phthalate (DEHP). National Institutes of Health, US Department of Health and Human Services; 2006. Accessed January 29, 2020. NIH Publication No. 06–4476. https://ntp.niehs.nih.gov/ntp/ohat/phthalates/dehp/dehp-monograph.pdf
Google Scholar
52Mirtallo JM, Ayers P, Boullata JI, et al. ASPEN lipid injectable emulsion safety recommendations, Part 1: background and adult considerations. Nutr Clin Pract. 2020; 35(5): 769-782.
10.1002/ncp.10496
CASPubMedWeb of Science®Google Scholar
53 Clinolipid. Package insert. Baxter Healthcare; 2020.
Google Scholar
54 Intralipid. Package insert. Baxter Healthcare; 2020.
Google Scholar
55 Nutrilipid injection. Package insert. B. Braun Medical; 2020.
Google Scholar
56 Omegaven injection. Package insert. Fresenius Kabi; 2020.
Google Scholar
57 SMOFLipid. Package insert. Fresenius Kabi; 2020.
Google Scholar
58Llop Talaverón JM, Novak A, Suñé Negre JM, Tahull MB, Badosa EL, Ticó Grau JR. Phytosterol determination in lipid emulsions for parenteral nutrition. Farm Hosp. 2018; 42(3): 116-119.
PubMedWeb of Science®Google Scholar
59Anez-Bustillos L, Dao DT, Baker MA, Fell GL, Puder M, Gura KM. Intravenous fat emulsion formulations for the adult and pediatric patient: understanding the differences. Nutr Clin Pract. 2016; 31(5): 596-609.
10.1177/0884533616662996
CASPubMedWeb of Science®Google Scholar
60Driscoll DF. Physicochemical assessment of total nutrient admixture stability and safety: quantifying the risk. Nutrition. 1997; 13(2): 166-167.
10.1016/S0899-9007(96)00394-2
CASPubMedWeb of Science®Google Scholar
61Klang MG. PFAT5 and the evolution of lipid admixture stability. JPEN J Parenter Enteral Nutr. 2015; 39(1_suppl): 67S-71S.
10.1177/0148607115595976
PubMedWeb of Science®Google Scholar
62Brown R, Quercia RA, Sigman R. Total nutrient admixture: a review. JPEN J Parenter Enteral Nutr. 1986; 10(6): 650-658.
10.1177/0148607186010006650
CASPubMedWeb of Science®Google Scholar
63Sayeed FA, Tripp MG, Sukumaran KB, Mikrut BA, Stelmach HA, Raihle JA. Stability of total nutrient admixtures using various intravenous fat emulsions. Am J Hosp Pharm. 1987; 44(10): 2271-2280.
CASPubMedWeb of Science®Google Scholar
64Watrobska-Swietlikowska D. Stability of commercial parenteral lipid emulsions repacking to polypropylene syringes. PLoS One. 2019; 14(4):e0214451.
10.1371/journal.pone.0214451
CASPubMedWeb of Science®Google Scholar
65Driscoll DF, Parikh M, Silvestri AP, Klütsch K, Bistrian BR, Nehne J. Establishing a stability window for medium- and long-chain-triglyceride lipid-based total nutrient admixtures using USP standards. Am J Health-Syst Pharm. 2006; 63(21): 2135-2143.
10.2146/ajhp050520
CASPubMedWeb of Science®Google Scholar
66Floyd AG. Top ten considerations in the development of parenteral emulsions. Pharm Sci Technlo Today. 1999; 4(4): 134-143.
10.1016/S1461-5347(99)00141-8
CASPubMedGoogle Scholar
67Guidelines for essential trace element preparations for parenteral use: a statement by an expert panel. AMA Department of Foods and Nutrition. JAMA. 1979; 241(19): 2051-2054.
10.1001/jama.1979.03290450049024
PubMedWeb of Science®Google Scholar
68Weinstein MM, Siegel FP, Blake MI. Formulation of a stable trace element solution for TPN. Am J Hosp Pharm. 1980; 37(12): 1620.
CASPubMedGoogle Scholar
69Vanek VW, Borum P, Buchman A, et al. Novel Nutrient Task Force, Parenteral Multi-Vitamin and Multi–Trace Element Working Group; American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) Board of Directors. A.S.P.E.N. position paper: recommendations for change in commercially available parenteral multivitamin and multi-trace element products. Nutr Clin Pract. 2012; 27(4): 440-491.
10.1177/0884533612446706
PubMedWeb of Science®Google Scholar
70Vanek VW, Borum P, Buchman A, et al. A call to action to bring safer parenteral micronutrient products to the U.S. market. Nutr Clin Pract. 2015; 30(4): 559-569.
10.1177/0884533615589992
PubMedWeb of Science®Google Scholar
71Stehle P, Stoffel-Wagner B, Kuhn KS. Parenteral trace element provision: recent clinical research and practical conclusions. Eur J Clin Nutr. 2016; 70(8): 886-893.
10.1038/ejcn.2016.53
CASPubMedWeb of Science®Google Scholar
72Jin J, Mulesa L, Rouillet MC. Trace elements in parenteral nutrition: considerations for the prescribing clinician. Nutrients. 2017; 9(5): 440.
10.3390/nu9050440
PubMedWeb of Science®Google Scholar
73 MultiTrace-4 Pediatric and MultiTrace-4 Neonatal. Package insert. American Regent; 2018.
Google Scholar
74 Tralement trace elements 4 injection, USP. Package insert. American Regent; 2020.
Google Scholar
75 Selenious acid injection. Package insert. American Regent; 2020.
Google Scholar
76 Zinc sulfate injection. Package insert. American Regent; 2020.
Google Scholar
77Mitrano FP, Baptista RJ. Stability of cimetidine HCl and copper sulfate in a TPN solutions. DICP. 1989; 23(5): 429-430.
10.1177/106002808902300515
Google Scholar
78 Power-Burns M. Adult multi-trace element availability. May 2013. https://www.fda.gov/media/86363/download. Accessed October 21, 2020
Google Scholar
79 INFED. Package insert. Allergan Sales; 2020.
Google Scholar
80 Venofer. Package insert. American Regent; 2020.
Google Scholar
81Khaodhiar L, Keane-Ellison M, Tawa NE, et al. Iron deficiency anemia in patients receiving home total parenteral nutrition. JPEN J Parenter Enteral Nutr. 2002; 26(2): 114-119.
10.1177/0148607102026002114
CASPubMedWeb of Science®Google Scholar
82Domellöf M, Szitanyi P, Simchowitz V, Franz A, Mimouni F; ESPGHAN/ESPEN/ESPR/CSPEN Working Group on Pediatric Parenteral Nutrition. ESPGHAN/ESPEN/ESPR/CSPEN guidelines on pediatric parenteral nutrition: iron and trace minerals. Clin Nutr. 2018; 37(6): 2354-2359.
10.1016/j.clnu.2018.06.949
CASPubMedWeb of Science®Google Scholar
83Barry DM, Reeve AW. Increased incidence of gram-negative neonatal sepsis with intramuscular iron administration. Pediatrics. 1977; 60(6): 908-912.
10.1542/peds.60.6.908
CASPubMedWeb of Science®Google Scholar
84 M.V.I. pediatric. Package insert. Hospira; 2019.
Google Scholar
85 M.V.I. adult. Package insert. Hospira; 2020.
Google Scholar
86 Pediatric Infuvite multiple vitamins. Package insert. Sandoz Canada; 2016.
Google Scholar
87 Adult Infuvite multiple vitamins. Package insert. Sandoz Canada; 2012.
Google Scholar
88Kriegel C, Festag M, Kishore RSK, Roethlisberger D, Schmitt G. Pediatric safety of polysorbates in drug formulations. Children. 2020; 7(1):1. https://doi.org/10.3390/children7010001
10.3390/children7010001
Google Scholar
89Brayfield A. Martindale: The Complete Drug Reference. 39th ed. Pharmaceutical Press; 2019.
Google Scholar
90MacDonald MG, Fletcher AB, Johnson EL, Boeckx RL, Getson PR, Miller MK. The potential toxicity to neonates of multivitamin preparations used in parenteral nutrition. JPEN J Parenter Enteral Nutr. 1987; 11(2): 169-171.
10.1177/0148607187011002169
CASPubMedWeb of Science®Google Scholar
91Abedi F, Razavi BM, Hosseinzadeh H. A review on gentisic acid as a plant derived phenolic acid and metabolite of aspirin: comprehensive pharmacology, toxicology, and some pharmaceutical aspects. Phytother Res. 2020; 34(4): 729-741.
10.1002/ptr.6573
CASPubMedWeb of Science®Google Scholar
92Lavoie JC, Bélanger S, Spalinger M, Chessex P. Admixture of a multivitamin preparation to parenteral nutrition: the major contributor to in vitro generation of peroxides. Pediatrics. 1997; 99(3): E6.
10.1542/peds.99.3.e6
CASPubMedWeb of Science®Google Scholar
93Christian VJ, Tallar M, Walia CLS, Sieracki R, Goday PS. Systematic review of hypersensitivity to parenteral nutrition. JPEN J Parenter Enteral Nutr. 2018; 42(8): 1222-1229.
10.1002/jpen.1169
PubMedWeb of Science®Google Scholar
94 ASCOR. Package insert. McGuff Pharmaceuticals; 2019.
Google Scholar
95 Ascorbic acid injection. Package insert. Mylan Institutional; 2016.
Google Scholar
96 Thiamine hydrochloride injection. Package insert. Physicians Total Care; 2010.
Google Scholar
97 Thiamine hydrochloride injection. Package insert. Mylan; 2017.
Google Scholar
98 Pyroxidine hydrochloride injection. Package insert. Fresenius Kabi USA; 2019.
Google Scholar
99 Folic acid injection. Package insert. Fresenius Kabi USA; 2019.
Google Scholar
100Laegeler WL, Tio JM, Blake MI. Stability of certain amino acids in a parenteral nutrition solution. Am J Hosp Pharm. 1974; 31(8): 776-779.
CASPubMedGoogle Scholar
101Jurgens RW, Henry RS, Welco A. Amino acid stability in a mixed parenteral nutrition solution. Am J Hosp Pharm. 1981; 38(9): 1358-1359.
CASPubMedGoogle Scholar
102 USP Chapter <788> Particulate matter injections. USP 42–NF 37. The United States Pharmacopeial Convention; 2019.
Google Scholar
103Newton DW. Maillard reactions in pharmaceutical formulations and human health. Int J Pharm Compound. 2011; 15(1): 32-40.
PubMedGoogle Scholar
104Stegink LD, Freeman JB, Den Besten L, Filer LJ. Maillard reaction products in parenteral nutrition. Prog Food Nutr Sci. 1981; 5(1-6): 265-278.
CASPubMedGoogle Scholar
105Labuza TP, Massaro SA. Browning and amino acid loss in model total parenteral nutrition solutions. J Food Sci. 1990; 55(3): 821-826.
10.1111/j.1365-2621.1990.tb05239.x
CASWeb of Science®Google Scholar
106Fry LK, Stegink LD. Formation of Maillard reaction products in parenteral alimentation solutions. J Nutr. 1982; 112(8): 1631-1637.
10.1093/jn/112.8.1631
CASPubMedWeb of Science®Google Scholar
107Novaes MR, Lima LA, Sousa MV. Maillard's reaction in parenteral solutions supplemented with arginine. Arch Latinoam Nutr. 2001; 51(3): 265-268.
CASPubMedWeb of Science®Google Scholar
108Bullock L, Fitzgerald JF, Walter WV. Emulsion stability in total nutrient admixtures containing a pediatric amino acid formulation. JPEN J Parenter Enteral Nutr. 1992; 16(1): 64-68.
10.1177/014860719201600164
CASPubMedWeb of Science®Google Scholar
109Washington C, Athersuch A, Kynoch DJ. The electrokinetic properties of phospholipid-stabilized fat emulsions. IV: the effect of glucose and pH. Int J Pharm. 1990; 64(2-3): 217-222.
10.1016/0378-5173(90)90271-5
CASWeb of Science®Google Scholar
110Washington C, Connoly MA, Manning R. The electrokinetic properties of phospholipid-stabilized fat emulsions. V: the effect of amino acids on stability. Int J Pharm. 1991; 65(1): 57-63.
10.1016/0378-5173(91)90301-4
Google Scholar
111Kaminski MV, Harris DF, Collin CF, Sommers GA. Electrolyte compatibility in a synthetic amino acid hyperalimentation solution. Am J Hosp Pharm. 1974; 31(3): 244-246.
CASPubMedGoogle Scholar
112Allwood MC, Kearney MCJ. Compatibility and stability of additives in parenteral nutrition admixtures. Nutrition. 1998; 14(9): 697-706.
10.1016/S0899-9007(98)00063-X
CASPubMedWeb of Science®Google Scholar
113Niemiec PW, Vanderveen TW. Compatibility considerations in parenteral nutrient solutions. Am J Hosp Pharm. 1984; 41(5): 893-911.
CASPubMedGoogle Scholar
114Driscoll DF, Bhargava HN, Zaim RH, Babayan VK, Bistrian BR. Physicochemical stability of total nutrient admixtures. Am J Health- Syst Pharm. 1995; 52(6): 623-634.
10.1093/ajhp/52.6.623
CASPubMedWeb of Science®Google Scholar
115Newton DW, Driscoll DF. Chemistry and safety of phosphates injections. Am J Health-Syst Pharm. 2008; 65(18): 1761-1766.
10.2146/ajhp080015
CASPubMedWeb of Science®Google Scholar
116Watrobska-Swietlikowska D. Compatibility of maximum inorganic and organic calcium and phosphate content in neonatal parenteral solutions. Sci Report. 2019; 9(1):10525.
10.1038/s41598-019-46987-y
PubMedGoogle Scholar
117Newton DW, Driscoll DF. Calcium and phosphate compatibility: revisited again. Am J Health-Syst Pharm. 2008; 65(1): 73-80.
10.2146/ajhp070138
CASPubMedWeb of Science®Google Scholar
118Wong JC, McDougal AR, Tofan M, et al. Doubling calcium and phosphate concentrations in neonatal parenteral nutrition solutions using monobasic potassium phosphate. J Am Coll Nutr. 2006; 25(1): 70-77.
10.1080/07315724.2006.10719517
CASPubMedWeb of Science®Google Scholar
119Anderson C, MacKay M. Physical compatibility of calcium chloride and sodium glycerophosphate in pediatric parenteral nutrition solutions. JPEN J Parenter Enteral Nutr. 2016; 40(8): 1166-1169.
10.1177/0148607115592673
CASPubMedWeb of Science®Google Scholar
120Robinson LA, Wright BT. Central venous catheter occlusion caused by body heat mediated calcium phosphate precipitation. Am J Hosp Pharm. 1982; 39(1): 120-121.
CASPubMedGoogle Scholar
121Boddapati S, Yang K, Murty R. Intravenous solution compatibility and filter-retention characteristics of trace-element preparations. Am J Hosp Pharm. 1981; 38(11): 1731-1736.
Google Scholar
122Staven V, Wang S, Grønlie I, Tho I. Physical stability of an all-in-one parenteral nutrition admixture for preterm infants upon mixing with micronutrients and drugs. Eur J Hosp Pharm. 2020; 27(1): 36-42.
10.1136/ejhpharm-2018-001562
PubMedWeb of Science®Google Scholar
123Cochran EB, Boehm KA. Prefilter and postfilter cysteine/cysteine and copper concentrations in pediatric parenteral nutrition solutions. JPEN J Parenter Enteral Nutr. 1992; 16(5): 460-463.
10.1177/0148607192016005460
CASPubMedWeb of Science®Google Scholar
124Bates CG, Greiner G. Gegenheimer A. Precipitate in admixtures of new amino acid injection. Am J Hosp Pharm. 1984; 41(7): 1312.
Google Scholar
125Chatterji DC, Kapoor J. Precipitate in admixtures of new amino acid injection (reply). Am J Hosp Pharm. 1984; 41(7):1312,1316.
PubMedGoogle Scholar
126Mayhew SL, Quick MW. Compatibility of iron dextran with neonatal parenteral nutrient solutions. Am J Health-Syst Pharm. 1997; 54(5): 570-571.
10.1093/ajhp/54.5.570
CASPubMedWeb of Science®Google Scholar
127Lee D, Barsky D, Hughes R, Driggers J, Poupard N, Mascarenhas M. Evaluation of the safety of iron dextran with parenteral nutrition in the paediatric inpatient setting. Nutr Diet. 2017; 74(5): 471-475.
10.1111/1747-0080.12355
PubMedWeb of Science®Google Scholar
128Billion-Rey F, Guillaumont M, Frederich A, Aulagner G. Stability of fat-soluble vitamins A (retinol palmitate), E (tocopherol acetate), and K1 (phylloquinone) in total parenteral nutrition at home. JPEN J Parenter Enteral Nutr. 1993; 17(1): 56-60.
10.1177/014860719301700156
CASPubMedWeb of Science®Google Scholar
129Henton DH, Merritt RJ. Vitamin A sorption to polyvinyl and polyolefin intravenous tubing. JPEN J Parenter Enteral Nutr. 1990; 14(1): 79-81.
10.1177/014860719001400179
CASPubMedWeb of Science®Google Scholar
130Koletzko B, Goulet O, Hunt J, Krohn K, Shamir R, Parenteral Nutrition Guidelines Working Group; European Society for Clinical Nutrition and Metabolism; European Society of Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN); European Society of Paediatric Research (ESPR). Guidelines on pediatric parenteral nutrition of the European Society of Pediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) and the European Society for Clinical Nutrition and Metabolism (ESPEN), supported by the European Society of Pediatric Research (ESPR). J Pediatr Gastroenterol Nutr. 2005; 41Suppl2: S1-S87.
Google Scholar
131Gillis J, Jones G, Pencharz P. Delivery of vitamins A, D, and E in total parenteral nutrition solutions. JPEN J Parenter Enteral Nutr. 1983; 7(1): 11-14.
10.1177/014860718300700111
CASPubMedWeb of Science®Google Scholar
132Brawley V, Bhatia J, Karp WB. Effect of sodium metabisulfite on hydrogen peroxide production in light exposed pediatric parenteral amino acid solutions. Am J Health-Syst Pharm. 1998; 55(12): 1288-1292.
10.1093/ajhp/55.12.1288
CASPubMedWeb of Science®Google Scholar
133Bhatia J, Stegink LD, Ziegler EE. Riboflavin enhances photo-oxidation of amino acids under simulated clinical conditions. JPEN J Parenter Enteral Nutr. 1983; 7(3): 277-279.
10.1177/0148607183007003277
CASPubMedWeb of Science®Google Scholar
134García J, Silva E. Flavin-sensitized photooxidation of amino acids present in a parenteral nutrition infusate: protection by ascorbic acid. J Nutr Biochem. 1997; 8(6): 341-345.
10.1016/S0955-2863(97)00024-7
Web of Science®Google Scholar
135Unger N, Ferraro A, Holzgrabe U. Investigation of tryptophan-related yellowing in parenteral amino acid solution: development of a stability-indicating method and assessment of degradation products in pharmaceutical formulations. J Pharm Biomed Anal. 2020; 177:112839.
10.1016/j.jpba.2019.112839
CASPubMedWeb of Science®Google Scholar
136Harraki B, Guiraud P, Rochat MH, Fussellier M, Favier A. Influence of copper, iron, and zinc on the physicochemical properties of parenteral admixture. J Parenter Sci Technol. 1993; 47(5): 199-204.
CASPubMedWeb of Science®Google Scholar
137Iliano L, Delanghe M, van Den Baviere H, Lauwers A. Effect of electrolytes in the presence of some trace elements on the stability of all-in-one emulsion mixtures for total parenteral nutrition. J Clin Hosp Pharm. 1984; 9(2): 87-93.
Google Scholar
138Deitel M, Smith LC, Friedman KL, et al. Physical stability of a total nutrient admixture for total parenteral nutrition. Can J Surg. 1989; 32(4): 240-243.
CASPubMedWeb of Science®Google Scholar
139Télessy IG, Balogh J, Turmezei J, Dredán J, Zelkó R. Stability assessment of o/w parenteral nutrition emulsions in the presence of high glucose and calcium concentrations. J Pharm Biomed Anal. 2011; 56(2): 159-164.
10.1016/j.jpba.2011.05.002
CASPubMedWeb of Science®Google Scholar
140Forchielli ML, Bonoli A, Stancari A, et al. Do carnitine and extra trace elements change stability of paediatric parenteral nutrition admixtures?. Clin Nutr. 2019; 38(5): 2369-2374.
10.1016/j.clnu.2018.10.016
CASPubMedWeb of Science®Google Scholar
141Tripp MG, Menon SK, Mikrut BA. Stability of total nutrient admixtures in a dual-chamber flexible container. Am J Hosp Pharm. 1990; 47(11): 2496-2503.
CASPubMedGoogle Scholar
142Watrobska-Swietlikowska D, Szlagatys-Sidorkiewicz A, Katarzyna Ł. Evaluation of physical stability of all in one parenteral admixtures for pediatric home care with high electrolytes concentrations. Nutr Hosp. 2015; 31(1): 236-243.
Web of Science®Google Scholar
143Lobo BW, da Veiga VF, Cabral LM, et al. Influence of the relative composition of trace elements and vitamins in physicochemical stability of total parenteral nutrition formulations for neonatal use. Nutr J. 2012; 11(1): 26.
10.1186/1475-2891-11-26
CASPubMedGoogle Scholar
144Uccello-Barretta G, Balzano F, Aiello F, Falugiani N, Desideri I. Stability of hydrophilic vitamins mixtures in the presence of electrolytes and trace elements for parenteral nutrition: a nuclear magnetic resonance spectroscopy investigation. J Pharm Biomed Anal. 2015; 107: 7-10.
10.1016/j.jpba.2014.12.008
CASPubMedWeb of Science®Google Scholar
145Lavoie JC, Chessex P. Bound iron admixture prevents the spontaneous generation of peroxides in total parenteral nutrition solutions. J Pediatr Gastroenterol Nutr. 1997; 25(3): 307-311.
10.1002/j.1536-4801.1997.tb01753.x
CASPubMedWeb of Science®Google Scholar
146Vaughan LM, Small C, Plunkett V. Incompatibility of iron dextran and a total nutrient admixture. Am J Hosp Pharm. 1990; 47(8): 1745-1746.
CASPubMedWeb of Science®Google Scholar
147Kumpf VJ. Update on parenteral iron therapy. Nutr Clin Pract. 2003; 18(4): 318-326.
10.1177/0115426503018004318
PubMedGoogle Scholar
148Puntis JW, Rushton DI. Pulmonary intravascular lipid in neonatal necropsy specimens. Arch Dis Child. 1991; 66(1 Spec No): 26-28.
10.1136/adc.66.1_Spec_No.26
CASPubMedWeb of Science®Google Scholar
149Pluhator-Murton MM, Fedorak RN, Audette RJ, et al. Trace element contamination of total parenteral nutrition. 2: effect of storage duration and temperature. JPEN J Parenter Enteral Nutr. 1999; 23(4): 228-232.
10.1177/0148607199023004228
CASPubMedWeb of Science®Google Scholar
150Burge JC, Flancbaum L, Holcombe B. Copper decreases ascorbic acid stability in total parenteral nutrition solutions. J Am Diet Assoc. 1994; 94(7): 777.
10.1016/0002-8223(94)91949-6
CASPubMedGoogle Scholar
151Allwood MC, Greenwood M. Assessment of trace element compatibility in total parenteral nutrition infusions. Pharm Weekbl (Sci). 1992; 14(5): 321-324.
10.1007/BF01977621
CASPubMedGoogle Scholar
152Shils ME, Levander OA. Selenium stability in TPN solutions. Am J Clin Nutr. 1982; 35(4): 829.
Web of Science®Google Scholar
153Levander OA. Considerations in the design of selenium bioavailability studies. Fed Proc. 1983; 42(6): 1721.
CASPubMedGoogle Scholar
154McGee CD, Mascarenhas MG, Ostro MJ, Tsallas G, Jeejeebhoy K. Selenium and vitamin E stability in parenteral solutions. JPEN J Parenter Enteral Nutr. 1985; 5(5): 568-570.
10.1177/0148607185009005568
Google Scholar
155Ganther HE, Kraus RJ. Chemical stability of selenious acid in total parenteral nutrition solutions containing ascorbic acid. JPEN J Parenter Enteral Nutr. 1989; 13(2): 185-188.
10.1177/0148607189013002185
CASPubMedWeb of Science®Google Scholar
156Sill-Steffens R, Hardy G. Sodium selenite stability in intravenous solutions. Clin Nutr. 2004; 23(4): 827-828.
Google Scholar
157Hardy G, Hardy I, Martin H. Selenium stability in vitamin and glutamine mixtures. Clin Nutr. 2005; 24(4): 596.
Google Scholar
158Tu Y-H, Knox NL, Bringer JM, et al. Compatibility of iron dextran with total nutrient admixtures. Am J Hosp Pharm. 1992; 49(9): 2233-2235.
CASPubMedGoogle Scholar
159MacKay M, Rusho W, Jackson D, McMillin G, Winther B. Physical and chemical stability of iron sucrose in parenteral nutrition. Nutr Clin Pract. 2009; 24(6): 733-737.
10.1177/0884533609351528
PubMedWeb of Science®Google Scholar
160Ferguson TI, Emery S, Price-Davies R, Cosslett AG. A review of stability issues associated with vitamins in parenteral nutrition. eSPEN J. 2014; 9(2): e49-e53.
Google Scholar
161Ha E, Wang W, Wang YJ. Peroxide formation in polysorbate 80 and protein stability. J Pharm Sci. 2002; 91(10): 2252-2264.
10.1002/jps.10216
CASPubMedWeb of Science®Google Scholar
162Silvers KM, Darlow BA, Winterbourn CC. Lipid peroxide and hydrogen peroxide formation in parenteral nutrition solutions containing multivitamins. JPEN J Parenter Enteral Nutr. 2001; 25(1): 14-17.
10.1177/014860710102500114
CASPubMedWeb of Science®Google Scholar
163Laborie S, Lavoie JC, Chessex P. Paradoxical role of ascorbic acid and riboflavin in solutions of total parenteral nutrition: implication in photoinduced peroxide generation. Pediatr Res. 1998; 43(5): 601-606.
10.1203/00006450-199805000-00007
CASPubMedWeb of Science®Google Scholar
164Silva E, Ugarte R, Andrade A, Edwards AM. Riboflavin-sensitized photoprocesses of tryptophan. J Photochem Photobiol B. 1994; 23(1): 43-48.
10.1016/1011-1344(93)06984-B
CASPubMedWeb of Science®Google Scholar
165Huvaere K, Cardoso DR, Homem-de-Mello P, Westermann S, Skibsted LH. Light-induced oxidation of unsaturated lipids as sensitized by flavins. J Phys Chem B. 2010; 114(16): 5583-5593.
10.1021/jp9121744
CASPubMedWeb of Science®Google Scholar
166Natera J, Massad WA, García NA. Vitamin B1 as a scavenger of reactive oxygen species photogenerated by vitamin B2. Photochem Photobiol. 2011; 87(2): 317-323.
10.1111/j.1751-1097.2010.00867.x
CASPubMedWeb of Science®Google Scholar
167Ribeiro DO, Pinto DC, Lima LMTR, Volpato NM, Cabral LM, de Sousa VP. Chemical stability study of vitamins thiamine, riboflavin, pyridoxine and ascorbic acid in parenteral nutrition for neonatal use. Nutr J. 2011; 10(1): 47-56.
10.1186/1475-2891-10-47
CASPubMedWeb of Science®Google Scholar
168Mirkovic DC, Basic Z, Roganovic B. PP185-Sun influence of type of bag, sort of admixtures and exposure to daylight upon stability of vitamins C and B2 in parenteral nutrition admixtures. Clin Nutr Suppl. 2011; 6(1): 93-94.
10.1016/S1744-1161(11)70238-1
Google Scholar
169Ferguson TI, Price-Davies R, Cosslett A. Vitamins—an unknown quantity (PP259). Clin Nutr. 2013; 32: S219.
10.1016/S0261-5614(13)60569-5
Google Scholar
170Chen MF, Boyce HW Jr, Triplett L. Stability of the B vitamins in mixed parenteral nutrition solution. JPEN J Parenter Enteral Nutr. 1983; 7(5): 462-464.
10.1177/0148607183007005462
CASPubMedWeb of Science®Google Scholar
171Barker A, Hebron BS, Beck PR, Ellis B. Folic acid and total parenteral nutrition. JPEN J Parenter Enteral Nutr. 1984; 8(1): 3-8.
10.1177/014860718400800103
CASPubMedWeb of Science®Google Scholar
172McEvoy GK. Handbook on Injectable Drugs. 20th ed. American Society of Health-System Pharmacists; 2018.
Google Scholar
173Scheiner JM, Araujo MM, DeRitter E. Thiamin destruction by sodium bisulfite in infusion solutions. Am J Hosp Pharm. 1981; 38(12): 1911-1913.
CASPubMedGoogle Scholar
174Bowman BB, Nguyen P. Stability of thiamin in parenteral nutrition solutions. JPEN J Parenter Enteral Nutr. 1983; 7(6): 567-568.
10.1177/0148607183007006567
CASPubMedWeb of Science®Google Scholar
175Smith JL, Canham JE, Kirkland WD, Wells PA. Effect of intralipid, amino acids, container, temperature, and duration of storage on vitamin stability in total parenteral nutrition admixtures. JPEN J Parenter Enteral Nutr. 1988; 12(5): 478-483. [Erratum published in JPEN J Parenter Enteral Nutr 1989;13(2):167.]
10.1177/0148607188012005478
CASPubMedWeb of Science®Google Scholar
176Allwood MC. Stability of vitamins in total parenteral nutrition solutions stored in three litter bags. Brit J Int Ther. 1982; 1(1): 22.
Google Scholar
177Rockwell GF, Campfield T, Nelson BC, Uden PC. Oxalogenesis in parenteral nutrition solution components. Nutrition. 1998; 14(11-12): 836-839.
10.1016/S0899-9007(98)00104-X
CASPubMedWeb of Science®Google Scholar
178Ferreyra ME, Ocaña MC, Vega R. Trace elements influence on ascorbic acid availability (PP260). Clin Nutr. 2013; 32: S121.
10.1016/S0261-5614(13)60305-2
Google Scholar
179Proot P, De Pourcq L, Raymakers AA. Stability of ascorbic acid in a standard total parenteral nutrition mixture. Clin Nutr. 1994; 13(5): 273-279.
10.1016/0261-5614(94)90049-3
CASPubMedWeb of Science®Google Scholar
180Gibbons E, Allwood MC, Neal T, Hardy G. Degradation of dehydroascorbic acid in parenteral nutrition mixtures. J Pharm Biomed Anal. 2001; 25(3-4): 605-611.
10.1016/S0731-7085(00)00589-6
CASPubMedWeb of Science®Google Scholar
181Dupertuis YM, Morch A, Fathi M. Physical characteristics of total parenteral nutrition bags significantly affect the stability of vitamins C and B1: a controlled prospective study. JPEN J Parenter Enteral Nutr. 2002; 26(5): 310-316.
10.1177/0148607102026005310
CASPubMedWeb of Science®Google Scholar
182Allwood MC. Factors influencing the stability of ascorbic acid in total parenteral nutrition infusions. J Clin Hosp Pharm. 1984; 9(2): 75-85.
10.1111/j.1365-2710.1984.tb01063.x
CASPubMedGoogle Scholar
183Kearney MC, Allwood MC, Martin H, Neal T, Hardy G. The influence of amino acid source on the stability of ascorbic acid in TPN admixtures. Nutrition. 1998; 14(2): 173-178.
10.1016/S0899-9007(97)00430-9
CASPubMedWeb of Science®Google Scholar
184Allwood MC, Plane JH. The wavelength-dependent degradation of vitamin A exposed to ultraviolet light. Int J Pharmaceutics. 1986; 31(1-2): 1-7.
10.1016/0378-5173(86)90206-1
CASWeb of Science®Google Scholar
185Allwood MC, Martin HJ. The photodegradation of vitamins A and E in parenteral nutrition mixtures during infusion. Clin Nutr. 2000; 19(5): 339-342.
10.1054/clnu.2000.0109
CASPubMedWeb of Science®Google Scholar
186Haas C, Genzel-Boroviczény O, Koletzko B. Losses of vitamin A and E in parenteral nutrition suitable for premature infants. Eur J Clin Nutr. 2002; 56(9): 906-912.
10.1038/sj.ejcn.1601417
CASPubMedWeb of Science®Google Scholar
187Allwood MC, Ball PA, Driscoll DF, Sizer T. Light protection during parenteral nutrition infusion: is it really necessary?. Nutrition. 2000; 16(3): 234-235.
10.1016/S0899-9007(99)00274-9
CASPubMedWeb of Science®Google Scholar
188Drott P, Meurling S, Meurling L. Clinical adsorption and photodegradation of the fat soluble vitamins A and E. Clin Nutr. 1991; 10(6): 348-351.
10.1016/0261-5614(91)90065-K
CASPubMedWeb of Science®Google Scholar
189Lavoie J, Chessex P, Rouleau T, Tsopmo A, Friel J. Shielding parenteral multivitamins from light increases vitamin A and E concentration in lung of newborn guinea pigs. Clin Nutr. 2007; 26(3): 341-347.
10.1016/j.clnu.2006.12.006
CASPubMedWeb of Science®Google Scholar
190Guidetti M, Sforzini A, Bersani G, et al. Vitamin A and vitamin E isoforms stability and peroxidation potential of all-in-one admixtures for parenteral nutrition. Int J Vitam Nutr Res. 2008; 78(3): 156-166.
10.1024/0300-9831.78.3.156
CASPubMedWeb of Science®Google Scholar
191Schmutz CW, Martinelli E, Mühlebach S. Stability of vitamin K1 assessed by HPLC in total parenteral nutrition (TPN) admixtures. Clin Nutr. 1992; 12(169): 110-111.
10.1016/0261-5614(92)90362-T
Google Scholar
192Laborie S, Lavoie JC, Pineault M, Chessex P. Protecting solutions of parenteral nutrition from peroxidation. JPEN J Parenter Enteral Nutr. 1999; 23(2): 104-108.
10.1177/0148607199023002104
CASPubMedWeb of Science®Google Scholar
193Robinson DT, Ayers P, Fleming B, et al. Recommendations for photoprotection of parenteral nutrition for neonates and premature infants: an ASPEN position paper. Nutr Clin Pract. 2021; 36(5): 927-941.
10.1002/ncp.10747
CASPubMedWeb of Science®Google Scholar
194Driscoll DF, Baptista RJ, Mitrano FP, et al. Parenteral nutrient admixtures as drug vehicles: theory and practice in the critical care setting. DICP. 1991; 25(3): 276-283.
10.1177/106002809102500312
CASPubMedGoogle Scholar
195Newton DW. Y-site compatibility of intravenous drugs with parenteral nutrition. JPEN J Parenter Enteral Nutr. 2013; 37(3): 297-299.
10.1177/0148607112465653
PubMedWeb of Science®Google Scholar
196Stawny M, Gostyńska A, Dettlaff, et al. Effect of lipid emulsion on stability of ampicillin in total parenteral nutrition. Nutrients. 2019; 11(3): 559.
10.3390/nu11030559
CASPubMedWeb of Science®Google Scholar
197Trissel LA, Gilbert DL, Martinez JF, Baker MB, Walter WV, Mirtallo JM. Compatibility of parenteral nutrient solutions with selected drugs during simulated Y-site administration. Am J Health-Syst Pharm. 1997; 54(11): 1295-1300.
10.1093/ajhp/54.11.1295
CASPubMedWeb of Science®Google Scholar
198Bouchoud L, Fonzo-Christe C, Klingmüller M, Bonnabry P. Compatibility of intravenous medications with parenteral nutrition: in vitro evaluation. JPEN J Parenter Enteral Nutr. 2013; 37(3): 416-424.
10.1177/0148607112464239
CASPubMedWeb of Science®Google Scholar
199Nezvalova-Henriksen K, Nilsson N, Østerberg CT, Berge VS, Tho I. Y-site physical compatibility of Numeta G13E with drugs frequently used at neonatal intensive care. Pharmaceutics. 2020; 12(7): 677.
10.3390/pharmaceutics12070677
CASPubMedWeb of Science®Google Scholar
200Staven V, Iqbal H, Wang S, Grønlie I, Tho I. Physical compatibility of total parenteral nutrition and drugs in Y-site administration to children from neonates to adolescents. J Pharm Pharmacol. 2017; 69(4): 448-462.
10.1111/jphp.12647
CASPubMedWeb of Science®Google Scholar
201Veltri M, Lee CKK. Compatibility of neonatal parenteral nutrient solutions with selected intravenous drugs. Am J Health-Syst Pharm. 1996; 53(21): 2611-2613.
10.1093/ajhp/53.21.2611
CASPubMedWeb of Science®Google Scholar
202Kalikstad B, Skjerdal A, Hansen TW. Compatibility of drug infusions in the NICU. Arch Dis Child. 2010; 95(9): 745-748.
10.1136/adc.2009.174268
PubMedWeb of Science®Google Scholar
203Watrobska-Swietlikowska D, Pietka M, Klek S. Evaluation of Y-site compatibility of home total parenteral nutrition and intravenous loop diuretics. Medicine. 2019; 98(21):(e15747).
10.1097/MD.0000000000015747
CASPubMedWeb of Science®Google Scholar
204Stawny M, Gostyńska A, Nadolna M, Jelińska A. Safe practice of Y-site drug administration: the case of colistin and parenteral nutrition. Pharmaceutics. 2020; 12(3): 292.
10.3390/pharmaceutics12030292
CASPubMedWeb of Science®Google Scholar
205Stawny M, Nadolna M, Jelinska A. In vitro compatibility studies of vancomycin with ready-to-use parenteral nutrition admixtures for safer clinical practice. Clin Nutr. 2020; 39(8): 2539-2546.
10.1016/j.clnu.2019.11.014
CASPubMedWeb of Science®Google Scholar
206Ross EL, Salinas A, Petty K, Her C, Carpenter JF. Compatibility of medications with intravenous lipid emulsions: effects of simulated Y-site mixing. Am J Health-Syst Pharm. 2020; 77(23): 1980-1985.
10.1093/ajhp/zxaa299
PubMedWeb of Science®Google Scholar
207 Agency for Toxic Substances and Disease Registry. Toxicological Profile for di(2-ethylhexyl)phthalate (DEHP). Agency for Toxic Substances and Disease Registry; 2002.
Google Scholar
208Loff S, Kabs F, Witt K, Sartoris J, Mandl B, Niessen KH, et al. Polyvinylchloride infusion lines expose infants to large amounts of toxic plasticizers. J Pediatr Surg. 2000; 35(12): 1775-1781.
10.1053/jpsu.2000.19249
CASPubMedWeb of Science®Google Scholar
209Green R, Hauser R, Calafat AM, et al. Use of di(2-ethylhexyl) phthalate-containing medical products and urinary levels of mono(2-ethylhexyl) phthalate in neonatal intensive care unit infants. Environ Health Perspect. 2005; 113(9): 1222-1225.
10.1289/ehp.7932
CASPubMedWeb of Science®Google Scholar
210Weuve J, Sanchez BN, Calafat AM, et al. Exposure to phthalates in neonatal intensive care unit infants: urinary concentrations of monoesters and oxidative metabolites. Environ Health Perspect. 2006; 114(9): 1424-1431.
10.1289/ehp.8926
CASPubMedWeb of Science®Google Scholar
211Mattiazzi P, Bohrer D, Viana C, et al. Extraction/leaching of metal-containing additives from polyvinyl chloride, ethyl vinyl acetate, and polypropylene bags and infusion sets into infusion solutions. PDA J Pharm Sci and Tech. 2019; 73(1): 60-69.
10.5731/pdajpst.2018.009019
CASPubMedGoogle Scholar

Citing Literature

Volume46, Issue2

February 2022

Pages 273-299

Parenteral nutrition compatibility and stability: A comprehensive review

Abstract

INTRODUCTION

Pharmacy science

Compatibility

Stability

Application to practice

PN COMPONENTS AND RELEVANT PROPERTIES

Macronutrients

Amino acids

Dextrose

Lipid emulsions

Micronutrients

Electrolytes

Multielectrolyte additives

Sodium and potassium chloride or acetate

Magnesium sulfate or chloride

Calcium gluconate or chloride

Sodium or potassium phosphate

Trace elements

Multielement additives

Chromium

Copper

Manganese

Selenium

Zinc

Iron

Vitamins

Multivitamins

Ascorbic acid

Thiamin

Pyridoxine

Folic acid

COMPATIBILITY

Admixture compatibility—Focus on macronutrients

Admixture compatibility—Focus on electrolytes

pH

Amino acids

Phosphate salts

Calcium salts

Dextrose concentration

Temperature

Sequence of calcium and phosphate additions

Duration of storage

Admixture compatibility—Focus on trace elements and vitamins

STABILITY

Focus on macronutrients (alone or combined)

Amino acids

Lipid emulsions

Focus on micronutrients (alone or combined)

Trace elements

Vitamins

OTHER

Medication

Added to the PN admixture

Co-infused with the PN admixture

Other components

SUMMARY

CONFLICT OF INTEREST

FINANCIAL DISCLOSURE

Supporting Information

REFERENCES

Citing Literature

References

Related

Information