Parenteral nutrition in clinical practice: International challenges and strategies

Abstract

Purpose

Parenteral nutrition (PN) is an established therapy when oral/enteral feeding is not sufficient or is contraindicated, but nevertheless PN remains a complex, high-alert medication that is susceptible to errors that may affect patient safety. Over time, considerable progress has been made to make PN practices safer. The purpose of this article is to address ongoing challenges to improve the PN use process from prescription to administration and monitoring, and to outline practical aspects fostering the safety, quality, and cost-effectiveness of PN, as discussed at the International Safety and Quality of PN Summit.

Summary

Opportunities to improve the PN use process in clinical practice include the promotion of inter-disciplinary communication, vigilant surveillance for complications, staff education to increase competency, and more consistent use of advanced technologies that allow automated safety checks throughout the PN process. Topics covered include considerations on PN formulations, including the value of intravenous lipid emulsions (ILEs), trends in compounding PN, the current and future role of market-authorized multi-chamber PN bags containing all 3 macronutrients (amino acids, glucose/dextrose, and ILE) in the United States and in Europe, and strategies to cope with the increasing global problem of PN product shortages.

Conclusion

This review outlines potential strategies to use in clinical practice to overcome ongoing challenges throughout the PN use process, and ultimately promote PN patient safety.

Key Points.

• Although parenteral nutrition (PN) is a well-established therapy, the complexity of PN makes it susceptible to errors that can affect patient safety.

• There are many opportunities within clinical practice to further improve the quality and safety of PN throughout the whole PN use process.

• Robust strategies should be adopted to handle the international problem of PN product shortages.

Parenteral nutrition (PN), used in a variety of settings in order to provide nutritional needs when oral/enteral feeding is not sufficient or is contraindicated, is a complex, high-alert medication requiring training and ongoing competency assessment to ensure patient safety.¹ The PN use process (Figure 1) includes a number of key steps: patient assessment, prescribing and ordering, order review, preparation and administration, routine patient monitoring, and patient reassessment.^2-7 This process involves practitioners working together from multiple disciplines, different departments, and even different locations. It is, therefore, evident that excellent communication between experts with a high level of competency, as well as standardization within this process, are key for safe and effective PN delivery.¹

Figure 1.

The parenteral nutrition (PN) use process: potential sources and reported incidence of errors. Compiled using data from references 2 through 7.

This article is based upon presentations given by an international group of experts who convened at the International Safety and Quality of Parenteral Nutrition summit held from November 8 to 10, 2021, at two locations (Charleston, SC, and Bad Homburg, Germany). During discussions at the summit, the experts identified challenges and strategies to ensure the quality, safety, and cost-effectiveness of the PN use process in different organizational and geographical settings. This involved understanding opportunities for standardization, regardless of whether compounded bags or industrially prepared, market-authorized multi-chamber bags (MCBs) are used for PN. MCBs are available in two formats: either as a 2-chamber bag (2CB) containing glucose/dextrose and amino acids, or the more modern 3-chamber bag (3CB) containing glucose/dextrose, amino acids, and intravenous lipid emulsions (ILEs).⁸ The quality and safety of PN formulations in terms of composition, stability, and management during PN shortages were also discussed. A set of consensus statements were formulated concerning PN quality and safety, as detailed in the expert consensus statements and summary of proceedings publication.⁹ The other chapters of this supplement complement each other, addressing specific challenges associated with short- or long-term PN and identifying areas for improvement as well as research gaps. It is important to understand that this article does not constitute any recommendations—these are to be found in the expert consensus statement publication⁹—but does present and summarize aspects from the international summit as a learning experience.

Safety and quality of the PN process

The first step in the PN process is an in-depth assessment of nutritional status followed by the development of a nutrition care plan for the patient. Depending on the local situation, this may be the responsibility of a dietitian, physician, pharmacist, physician assistant, nurse, nurse practitioner, or another qualified member of the nutrition support team. A critical step is the review of the PN order (generally done by the pharmacist or the staff of a compounding center) to check for dosing appropriateness, stability, and compatibility. This step is of great importance, since PN represents a highly complex process, involving 40 to 50 different components, and is prone to errors that can lead to serious complications.^2,6,10,11 The use of standardized PN formulations for selected patient populations (and/or standard order forms when individualized PN is required) has been advocated by several organizations. These include the United States–based Institute for Safe Medication Practices¹² and the American Society for Enteral and Parenteral Nutrition (ASPEN),¹¹ whose stated objectives include reducing prescribing errors, reducing incompatibility and stability issues, and increasing efficiency. An example template for a PN therapy order, proposed by ASPEN, can be obtained from its PN safety consensus recommendations.¹⁰ The preparation of a PN prescription involves compounding, labeling, and dispensing and takes place after the order has been reviewed and verified. Compounding, carried out by the pharmacist, is another step in the PN process that is associated with a high risk of errors, owing to the need for a sterile environment and for numerous additions.¹⁰

Administration of PN in the hospital environment is generally the responsibility of the nurse, while in the home setting, the patient or caregiver is in charge. This step involves a further verification of the order, the proper choice of venous access, and infection control. The patient must be monitored regularly, taking into account not only nutritional outcomes but also any complications arising from mechanical or metabolic problems. Standardization throughout the entire PN process is important to minimize the risk of errors¹³ and is also emphasized in the consensus statements arising from this summit (see statements 5-7 in the summary article).⁹

Risk of errors throughout the PN use process

As mentioned previously, the PN use process is susceptible to errors because of its complexity. In Figure 1, the frequency of medication errors and their potential sources related to the PN process are shown, compiled from surveys conducted in different settings, including adult and pediatric inpatient and outpatients.^2-7 Overall, the reported incidence of errors varies between 3 and 16 errors per 1,000 prescriptions across various settings (Figure 1). However, PN may not be perceived as a medication in some settings, which may lead to underreporting of errors.¹ A US survey of nutrition support providers for the month of August 2011 revealed that, at that time, 44% of organizations did not track or were unaware of PN-related medication errors.¹⁴ Today, error prevention programs are integral part of a modern PN process (see statement 10 in the summary article).⁹

Several studies carefully assessed the safety of the PN process by reviewing errors reported to US national databases.^6,15,16 In a database analysis between May 2009 and April 2011, only 19 of 1,311 PN-related errors were classified as potentially or actually harmful to patients.¹⁶ These errors occurred most frequently during administration (n = 10) and ordering (n = 8). A 2006-2016 analysis of PN error data by the Institute for Safe Medication Practices found that most PN-related errors occurred in the compounding/dispensing and administration steps.⁶ Among the PN ingredients, ILEs, followed by electrolytes, were most frequently associated with errors,¹⁶ and PN shortages have also been identified as sources of PN errors (both of these topics are covered in greater detail later in this article). Overall, these errors highlight that further opportunities remain for improvements to PN process safety (see statement 10 in the summary article).⁹

US and European considerations regarding PN safety and quality

All-in-one (AIO) admixtures for PN contain 40 to 50 ingredients, including amino acids (15%-20% of energy supplied), glucose/dextrose (~50% of energy supplied), lLEs (30%-35% of energy supplied), water-soluble and fat-soluble vitamins, trace elements, fluids, and electrolytes (according to requirements). Guidance for the nutritional requirements of stable patients requiring PN is given in guidelines such as those concerning chronic intestinal failure in adults by the European Society for Clinical Nutrition and Metabolism (ESPEN).¹⁷ With regard to energy requirements, it is agreed that these should be calculated individually based on patient needs and regularly reevaluated. As a general rule, 20 to 35 kcal per kilogram of body weight (BW) per day and 0.8 to 1.4 g of protein per kilogram of BW per day day are sufficient for the majority of stable patients, including noncritically ill hospitalized and long-term PN patients. Requirements differ from these estimates in other patient settings, such as in the intensive care unit (ICU), during situations associated with volume limitations, in patients with chronic kidney disease, in pediatric and neonatal patients, or with those receiving comedications providing substantial amounts of additional energy. A wide variety of guidelines on the composition of PN formulations for different indications, patient settings, and age groups are available from international nutrition societies such as ESPEN in Europe and ASPEN in the US.

AIO PN admixtures can include a large number of components, mostly representing reactive species, and thus they must be evaluated thoroughly for stability and compatibility to ensure PN safety and quality.¹⁸ Potential physicochemical interactions affecting PN admixtures include emulsion instabilities, solubility issues (precipitations), photo-induced reactions, thermic reactions, material interactions (sorption, permeation), and chemical reactions (eg, oxidation, reduction, hydrolysis, polymerization, decarboxylation, complexation, lipid peroxidation, or Maillard reactions).¹⁸ The physicochemical stability of a compounded PN formulation depends on numerous factors, including whether all macronutrients are admixed, micronutrients are added, or light-proofed bags are used.¹⁹ Thus, it is important that AIO PN admixture stability and compatibility should be guaranteed up to administration by the declared expiry date, under defined conditions.¹⁸

Improving PN formulations and the PN process in clinical practice

ASPEN has developed several documents dedicated to providing evidence-based guidance for safe PN practices. For instance, guidelines for PN prescribing, order review, and preparation were published in 2014,¹¹ and a standardized ASPEN model for PN administration competency was published in 2018.¹³ At the summit, 6 major aspects were identified regarding the safety and quality of PN formulations and the PN process in clinical practice (Figure 2).

Figure 2.

Major aspects of the improvement of parenteral nutrition (PN) formulations and the PN process in clinical practice. MCB indicates multichamber bag; TE, trace element; US, United States.

Increased use of innovative technologies for prescription and preparation of PN.

To decrease the risk of PN-related errors, various advanced technologies can guide the prescribing and preparation of PN and perform automated safety checks. These include computerized provider order entry (CPOE), electronic health records (EHRs), and clinical decision support. Enhancements to EHR systems to improve safety and efficacy in the PN process have been proposed by a collaborative working group from ASPEN, the Academy of Nutrition and Dietetics, and the American Society of Health-System Pharmacists (ASHP).¹ This working group involved experts in PN, EHR functionality, and health information technology standards.¹ Recommendations are summarized in Box 1.^1,20

Box 1. Recommendations for Electronic Health Record Functionality and Health Information Technology Standards^1,20.

When introducing/optimizing an electronic health record system the following factors should be considered: • Analyze current PN workflow from prescription to administration, evaluating options for process optimization. • Use standardized and validated PN labelling templates, such as those developed by ASPEN. • Incorporate clinical decision support to guide prescribers regarding adult, pediatric, and neonatal patients’ requirements for macronutrients and micronutrients. • Review documentation processes, aiming for no/minimal manual transcription of prescriptions and preventing double documentation (similar information included in different parts of a patient’s electronic health record). • Include functionalities that can address special PN regimens or situations such as:  ◦ Particular requirements for cyclic PN (eg, giving PN during the night and fasting during the day, allowing some hours for metabolic rest)  ◦ Transitions from hospital PN orders to home PN orders, and vice versa

Abbreviations: ASPEN, American Society for Parenteral and Enteral Nutrition; PN, parenteral nutrition.

It is important to acknowledge that clinical decision support should be used in both automated compounding devices (ACDs) and EHR systems. For example, EHRs and ACDs should allow the setting of “soft-stop” and “hard-stop” clinical decision support alerts, limit alerts, and “best practice” alerts, which are configurable and customizable.²⁰ Typically, these should include dosing alerts (both upper and lower limits for clinical effectiveness and stability of PN) and precipitation warnings for calcium and phosphorus, based on calcium/phosphate solubility curves for the PN components ordered.²⁰ EHRs should provide real-time clinical decision support alerts to help avoid issues such as PN toxicity or instability: for prescribers at the time of PN order, pharmacists at the time of PN order verification and compounding, and medical staff at the time of PN administration.²⁰ However, clinical decision support alerts should be designed to minimize false positives, false negatives, and unclear alerts, which can all lead to staff distrust of the system, excessive overrides, and staff alert fatigue that cumulatively can result in adverse patient outcomes.²⁰

It takes time for the implementation of technical aids shown in Box 1 to be integrated into the PN process. Surveys conducted by ASPEN in 2011 and 2014 showed a considerable increase in the use of CPOE systems over this period but only marginal advancement in terms of auto-populated fields in computerized systems or ACD use (Table 1).^14,21 Likewise, there was only a small increase in the use of electronic interfaces between ACD and EHR computer systems or with regard to dosing guidelines and clinical decision support built into the systems. Regrettably, more recent data are not available. It is also worth noting that, as with many automated systems or devices, the potential benefits of ACDs are likely to be fully realized when the technology is used appropriately, including integration/interconnection of all systems and processes involved. Thus, for example, ACDs may not decrease the risk of errors when manual transcription remains as a component of the PN medication-use process.

Table 1.

Availability of Technological Aids for Parenteral Nutrition (PN) Preparation According to Surveys Conducted in 2011 and 2014 by The American Society for Enteral and Parenteral Nutrition (ASPEN).^14,21

PN characteristic	Survey of PN Safe Practices 2011 (895 total survey responders)a	Survey of PN Safe Practices 2014 (1125 total survey responders)a
Computer prescriber order entry for PN	33% (287/876)	63% (436/689)
Auto-populate fields	33% (38/114)	42% (204/491)
Automated compounding device interface with electronic health records	19% (73/396)	28% (115/416)
Dosing guidelines and clinical decision support built into system	54% (62/114)	50% (246/491)
Automated compounding device for PN compounding	67% (410/608)	71% (330/465)

^aNot all respondents answered all survey questions. Each value is presented as a percentage based on the number of respondents to each question.

Need for a larger selection of MCBs.

AIO systems are now generally preferred, whether in the form of compounded bags (from hospital pharmacies or outsourced compounding facilities) or commercially available MCBs. MCBs are a valid option to improve the safety of PN.²² In the US, 2CBs (containing glucose/dextrose and amino acids) predominate, whereas the typical European MCB is a 3CB composed of all 3 macronutrients (glucose/dextrose, amino acids, and an ILE). Compared with compounded bags, MCBs require less manipulation to prepare.^23-25 A limited number of additives (eg, micronutrients) are added to the bag at activation.^24,25 A systematic literature review of 18 studies conducted in Europe, the US, Canada, Latin America, and Turkey identified evidence of potential clinical, ergonomic, and economic benefits for MCBs compared with hospital pharmacy compounded bags or multibottle systems.²⁵ In 6 out of 8 studies reporting this outcome, a lower risk of infectious complications was found with MCBs compared with other systems. The lower risk of infections translated into a shorter duration of hospital stay in half of the studies, and most studies identified a potential cost benefit for MCBs. Importantly, nutritional efficacy was found to be equivalent for these different delivery systems.²⁵ Whether one system is preferred over the other varies geographically (see statement 4 in the summary article).⁹ Whereas the use of MCBs in the US (typically 2CBs with ILE given separately) is often seen in smaller centers with lower PN volumes,²⁶ the use of MCBs (generally 3CBs) is very common in large European centers, including major academic institutions, mainly for safety and cost-effectiveness reasons. As a result of their wide acceptance, a broad variety of MCB formulations are available in Europe, covering the nutritional needs of different types of patient groups.

To illustrate these differences, in the US only one adult 3CB for central use and one for peripheral use are approved currently, though a wider variety of 2CBs for central or peripheral use are available (an ILE has to be given separately). In Europe, experts reported that, at their institutions, they can generally select between about 10 different 3CBs for adults, allowing them to adapt PN to each individual’s requirements. However, in the US only a 3CB with soybean oil (SO) as the sole ILE source is available, but in Europe there is a choice of 3CBs with different ILE types, including second-, third- and fourth-generation ILEs containing medium-chain triglycerides (MCT), olive oil (OO), and fish oil (FO).

These differences are reflected in a survey comparing MCBs to compounding in the US and in different European countries. In an international cross-sectional survey, Pironi et al²⁷ investigated the use of MCBs versus customized, compounded preparations for home PN (HPN) users with chronic intestinal failure. A total of 3,239 patients were enrolled in 65 centers from 22 countries, including the US and 15 European countries. The use of MCBs differed greatly among countries, HPN centers, and patients with benign and malignant chronic intestinal failure. Within Europe, up to 82% of patients received MCBs (with or without added fluid and electrolytes), but in the US 90% of HPN patients received compounded PN preparations. Overall, MCBs were more frequently used when the local pharmacy was the HPN provider, while compounded admixtures were more frequently used when the HPN provider was a home care company. Moreover, customized admixtures tailored to patients’ needs were mainly used in cases of benign chronic intestinal failure, while MCBs were used mainly in cases of malignant chronic intestinal failure. It is worth noting in this context that a survey of hospital pharmacy practices in the US found that the percentage of hospitals using 2CBs as the predominant form of PN preparations increased from 36% (in 2011) to 45% (in 2017), and that this increase was driven mainly by smaller institutions (<200 beds).²⁶

Daily use of ILEs.

ILEs are an integral part of PN, serving as an energy-dense source of calories, reducing the glucose/dextrose load and providing essential fatty acids (EFAs). Moreover, ILEs facilitate the delivery of lipid-soluble vitamins and modulate several biologic functions, including inflammatory and immune responses, coagulation, and cell signaling. Therefore, ILEs should be given regularly—if possible on a daily basis—to patients receiving PN in the hospital or home setting, and preferably using modern, mixed-oil ILEs containing FO.^28-30 In the US, historical concerns of using ILEs on a daily basis, especially in the ICU setting, came largely from a single randomized trial in critically ill trauma patients published in 1997.³¹ This study reported that SO-based ILE given during the early postinjury period (10 days) increased susceptibility to infection, prolonged pulmonary failure, and delayed recovery. However, it is unclear whether these less favorable outcomes were because of the SO-based ILE or the lower caloric intake in the control group. More recently, retrospective studies have consistently shown that giving ILEs early on a daily basis did not increase infection risk, time to recovery, or mortality, both in critically and noncritically ill hospitalized patients.^32-35 Moreover, experts in the field endorse the daily use of ILEs as an integral part of PN.^28,29,36

The recent 2022 update of the ASPEN ICU guidelines reflects the trend towards endorsing the daily use of ILEs as an integral part of PN.³⁷ However, the previous version recommended withholding or, in case of risk of EFA deficiency, limiting ILE during the first week after starting PN in critically ill patients.³⁸ Such restrictions on lipid supply in the first week of PN in the ICU are no longer imposed (though this is classified in the guidelines as a weak recommendation).³⁷ To help avoid potential errors, ASPEN published guidance concerning dosing, timing of administration, monitoring, stability, and compatibility of different types of ILEs in AIO admixtures used in adult and pediatric patients.^30,39 The adult recommendations are summarized in a fact sheet available from the ASPEN website⁴⁰ and also in Box 2.³⁰ It is also worth noting that there have been concerns about the continued use of ILEs with SO as the sole lipid source, particularly in critically ill patients, owing to an association with inflammatory and immunosuppressive effects.³⁰ Thus, other ILE products may be given to critically ill adults, reducing the SO ILE content and reducing the likelihood of inflammatory and immunosuppressive effects associated with pure SO ILE.³⁰ Another US issue with the use of ILEs is related to the appropriate use of filters, though this practice is not widespread in other parts of the world.⁴¹ In the US, a filter (1.2 µm) is recommended with the administration of ILE-containing and lLE-free PN formulations. If ILEs are infused through a different vascular access device from the PN, each will need to be filtered through a 1.2-µm filter.^30,39

Box 2. American Society for Enteral and Parenteral Nutrition Guidance on Safe Use of Intravenous Lipid Emulsions in Parenteral Nutrition³⁰.

Dosing and prescribing • Recommended doses vary with the type of ILE:  ◦ 100% SO: 1 g/kg BW/day (<1 g/kg BW/day in critical illness)  ◦ SO/MCT/OO/FO: 1-2 g/kg BW/day  ◦ SO/OO: 1 to 1.5g/kg BW/day • Maximum daily dose: 2.5 g/kg BW/day • Maximum infusion rate: 0.11 g/kg BW/hour • Dosing for maintenance therapy or long-term therapy: ~1 g/kg BW/day (stable patients) • To prevent EFAD: at least 2% to 4% of energy from linoleic acid and 0.25% to 0.5% from α-linolenic acid. Monitoring • TGs should be checked when starting PN, weekly for hospitalized patients, and monthly for those receiving long-term PN. • Withold or limit ILE if serum TG levels exceed 400 mg/dL. • Avoid withholding ILE for more than 2 weeks owing to the associated risk of EFAD. Stability and compatibility • Stability and compatibility of the ILE with other components in the AIO admixture should be reviewed by a trained pharmacist. • ILE manufacturers should provide data on the acceptable ranges of macronutrients and micronutrients for stability of the AIO admixture. Preparation and administration • Maximum infusion time for ILEs:  ◦ 24 hours as a part of an AIO admixture  ◦ 12 hours as a separate infusion • Use DEHP-free tubing. • Use a 1.2-µm filter for AIO admixtures and for a separate ILE infusion. • Avoid administration of medications through the venous catheter for PN.

Abbreviations: AIO, all-in-one; BW, body weight; DEHP, diethynlhexyl phthalate; EFAD, essential fatty acid deficiency; FO, fish oil; MCT, medium-chain triglyceride; OO, olive oil; SO, soybean oil; TG, triglyceride.

Newer ILE formulations: potential benefits.

In Europe and many other parts of the world, newer ILEs containing omega-3 polyunsaturated fatty acids (PUFAs) from FO enjoy wide acceptance and are endorsed by guidelines.^36,42-44 However, opinions seem to differ in the US.^28,37 In the US, there are currently 4 ILE products with an adult indication that are available, including 2 standard SO ILEs (Intralipid, Fresenius Kabi, Uppsala, Sweden; Nutralipid, B. Braun Medical Inc., Bethlehem, PA), a third-generation SO/OO-based emulsion (ClinOleic, Baxter Healthcare Corp., Deerfield, IL), and a fourth-generation product with SO/MCT/OO/FO (SMOFlipid, Fresenius Kabi). In Europe, additional third- and fourth-generation ILEs are available. These include mixtures of SO/MCT (Lipofundin, B. Braun, Melsungen, Germany), SO/MCT/FO (Lipoplus/Lipidem, B. Braun, Melsungen, Germany), and SO/MCT/OO/FO (Finomel, Baxter Healthcare Ltd, Thetford, Norfolk, UK). As reported by experts from Switzerland and Belgium, fourth-generation ILEs with omega-3 PUFAs from FO with reduced SO content are preferred because of their beneficial properties, including immunomodulatory actions and resolution of inflammation across a wide range of groups (eg, surgical, cancer, and critically ill patients).^29,36,45,46 In contrast, as revealed by a survey conducted in February 2021 among critical care dietitians and pharmacists, 60% of US centers reported that they were not using newer mixed-oil ILEs (ie, SO/OO or SO/MCT/OO/FO blends) (data provided by P. E. Wischmeyer, Duke University School of Medicine, February 2021). Nevertheless, in some US centers, ILEs containing FO have become well accepted.^28,47-50

The 2022 update of the ASPEN ICU guidelines did not differentiate between the provision of pure SO ILEs and newer alternative ILEs, with the guideline authors basing their conclusions on a meta-analysis of RCTs that compared PN using with ILEs with and without FO.³⁷ The guideline update proved controversial. Many clinicians and experts within the field criticized the methods used and a lack of transparency in the reporting, as well as disagreeing with some conclusions such as the lack of guidance concerning ILE choice for clinicians.^51,52 For example, most other meta-analyses have found clear clinical benefits for the use of FO in PN, including the recently published Canadian Critical Care Nutrition guidelines⁵³ and associated meta-analyses⁴⁴ (see Table 2).^37,42-44 Moreover, the results of the largest meta-analysis on this topic, including 49 RCTs and 3,641 patients, showed the use of omega-3 PUFAs was associated with 40% fewer infections (P < 0.00001), an approximately 2 days’ shorter hospital stay (P < 0.00001), about 2 days’ shorter ICU stay (P = 0.01), and 56% reduction in sepsis (P = 0.0004).⁴² Retrospective observational studies in the US have also shown favorable effects in a large cohort of hospitalized patients including a subgroup of critically ill adult and pediatric patients or in a small cohort of longer-term HPN patients when using mixed-oil ILEs containing FO compared with pure SO-based ILEs used previously (see Table 3).^47,48,50 A follow-up to the 2022 guidelines has been planned in the form of a “clinical recommendations paper,” which will hopefully address these clinical concerns, but this has not yet been published. During the summit, the experts discussed the lipid issue and encouraged the use of newer-generation ILEs that include FO (statement 11b).⁹

Table 2.

Recent meta-analyses comparing PN with various ILEs.^37,42-44

MA	Patients	Number of RCTs/patients	ILE comparison	Main results (clinical outcomes)
Pradelli et al., 202042	Adult hospitalized patients	49 studies 3641 patients	Containing FO vs not containing FO	Reduced risk of infection by 40% (p<0.00001). Reduced mean length of ICU stay by 1.95 days (p=0.01). Reduced length of hospital stay by 2.14 days (p<0.00001). Reduced risk of sepsis by 56% (p=0.0004). Non-significant trend of 16% reduction in mortality rate (p=0.15).
Pradelli et al., 202043	Adult hospitalized ICU patients (subgroup of Pradelli 202042)	24 studies 1421 patients	Containing FO vs not containing FO	Reduced risk of infection by 38% (p=0.004). Reduced mean length of ICU stay by 1.89 days (p=0.01). Reduced length of hospital stay by 3.05 days (p=0.003). Non-significant trend of 43% reduced risk of sepsis (p=0.13). No significant differences in duration of mechanical ventilation or mortality rates.
Compher et al., 202237	Adult critically ill patientsa	10 studies 919 patients	Containing FO vs not containing FO	Decreased incidence of pneumonia (∼5% vs ∼9%; p=0.03). No significant differences in incidences of catheter-related infections, length of ICU stay, days on mechanical ventilation, or mortality rates.
Notz et al., 202244	Adult critically ill patients	26 studies 1733 patients	Different comparisons all aiming to reduce omega-6 FA content: • SO/MCT vs SO • SO/OO vs SO or SO/MCT • FO-containing ILE vs SO, SO/MCT, or SO/OO	Results for FO-containing PN vs SO, SO/MCT or SO/OO: • reduced risk of infection complications by 35% (p=0.03) • reduced mean length of ICU stay by 3.53 days (p=0.009) • trend towards a reduced length of hospital stay by 5.93 days (p=0.11) • trend towards a reduced risk of one-month mortality by 26% (p=0.06) • no significant differences concerning duration of mechanical ventilation or overall mortality rates.

^aOnly studies reporting clinical outcomes

FA, fatty acid; FO, fish oil. ICU, intensive care unit; ILE, intravenous lipid emulsion; LOS, length of stay; MCT, medium-chain triglyceride; OO, olive oil; PN, parenteral nutrition; PUFA, polyunsaturated fatty acid; SO, soybean oil.

Table 3.

Retrospective Studies Assessing the Effects of Transitioning from Soybean Oil to Mixed Lipid Emulsions Containing Fish Oil in The United States (US).^47,48,50

Study	Patients	Main results
Mundi et al., 202050	Adults receiving HPN (n=17)	Post-switch (after ≥12 months) vs pre-switch period: • reduction in hospital length of stay (IRR 0.97; p=0.039). • increased ILE energy supply from 8% to 22%, and energy supplied by glucose/dextrose reduced from 66% to 54%. • improvements in the following parameters: ALT/AST, total bilirubin, and α-tocopherol levels • stable alkaline phosphatase and triglyceride levels.
Haines et al., 202247	Adult hospitalized patients (n=341 pre-switch; n=859 post-switch)	Post-switch vs pre-switch period: • reduction in hospital length of stay (IRR 0.97; p=0.039). • non-significant trend towards reduced risk of any infection and urinary tract infection (p=0.16 and p=0.12, respectively) • no significant differences in 30-and 90-day readmission and hospital mortality rates.
Haines et al., 202247	Adult ICU patients (n=110 pre-switch; n=337 post-switch)	Post-switch vs pre-switch period: • reduction in length of hospital stay (IRR 0.91; p<0.0001). • reduction in length of ICU stay (IRR 0.90; p=0.036). • reduced risk of urinary tract infection by 50% (p=0.038) • non-significant trend towards reduced risk of 30-day hospital readmission (p=0.067) • no significant differences in any infection, pneumonia, 90-day readmission, and hospital mortality rates.
Wischmeyer et al., 202248	Critically ill pediatric patients (n=379 pre-switch; n=278 post-switch)	Post-switch vs pre-switch period: • reduced risk of pneumonia by 76% (p=0.03). • Reduction in length of hospital stay by 27% (p<0.0001). • non-significant trend towards reduced risk of bacteremia/sepsis and all-cause infections (p=0.099 and p=0.091, respectively).

ALT, alanine aminotransferase; AST, aspartate aminotransferase; HPN, home parenteral nutrition; ICU, intensive care unit; ILE, intravenous lipid emulsion; IRR, incident rate ratio; LOS, length of stay.

Vitamin and mineral/trace element products.

In Europe, recommended daily doses of micronutrients can generally be met by using commercial market-authorized multivitamin and trace element preparations, which cover patients’ daily requirements (addition of 1 vial per day). However, only one vitamin and one trace element mixture are currently available in the US, as there have been constant and severe shortages in recent years. Moreover, the US experts previously called for modifications to the available trace element product to meet current recommendations,⁵⁴ though this is no longer an issue.

Another issue mentioned during the summit related to the provision of adequate amounts of phosphate, magnesium, and calcium in PN being limited by physical compatibility issues.¹⁵ Calcium salts are highly reactive compounds, readily forming insoluble products with a number of additives, primarily inorganic phosphate salts.¹⁵ The use of organic phosphates may considerably increase the calcium/phosphate compatibility, since organic sodium glycerophosphate is much more stable in solution than inorganic phosphates, regardless of pH, temperature, type of calcium, or amino acid or glucose concentrations.^55,56 Regrettably, organic phosphate products for PN (eg, Glycophos, Fresenius Kabi AG, Bad Homburg, Germany) are not approved in the US, thus limiting flexibility when compounding AIO admixtures.^57,58

Macronutrients and micronutrients shortages.

Shortages of various PN components have been an issue since the mid-1980s, when the first shortages of intravenous multivitamins occurred. Several PN component shortages were reported in the following years, but most were temporary. However, the problem of PN component shortages worsened such that in recent years almost every component has been in short supply at least once. In addition, shortages became more frequent and longer lasting, with some ongoing for almost 10 years.⁵⁹ Shortages of almost every PN component (with the exception of glucose/dextrose) have become a global issue.

Shortages of one or more of the critical components of PN admixtures contribute to the risks associated with this complex therapy, especially when multiple components are in limited supply.⁵⁹ Patients require PN as a life-sustaining measure that they cannot do without. Shortages can delay or compromise therapy for these patients and lead prescribers to use alternatives that may be less effective or pose additional risks. As shortages may increase the risk of medication errors with PN, they can be a significant threat to the health and welfare of patients, worsening patient outcomes and consuming healthcare resources.^30,60,61 Patients receiving HPN, in particular, are at risk of harm related to product shortages because they are monitored less frequently than those receiving PN in a hospital.⁶² Shortages can result from many factors, including supply disruptions of PN components, the manufacturer slowing or stopping production, and damage to production sites caused by natural disasters.⁶⁰ As an example, severe shortages of PN components in the US were caused by hurricanes in Puerto Rico that disrupted Baxter’s manufacturing plants for saline and amino acid infusion products.⁶³

In the US, information on ongoing drug shortages can be obtained from the drug shortage websites of ASPEN⁶⁴ and ASHP⁶⁵ and the US Food and Drug Administration (FDA) website on current and resolved drug shortages.⁶⁶ As of March 2022, seven shortages related to ILEs, multivitamins, and electrolytes were identified in the FDA database under the category of “total parenteral nutrition.” According to section 506C-1 of the FDA Safety and Innovation Act, FDA is obliged to prepare an annual report of shortages for the US Congress. Specifically, this report provides a background on drug shortages and FDA’s efforts to address them. Importantly, these data refer to all kinds of drugs, including over-the-counter preparations and certain cosmetics. The number of new drug shortages per calendar year reported by FDA declined from a high of 250 in 2011 to 43 in 2020. Of note, FDA helped to prevent large numbers of drug shortages, most recently 154 in 2018 and 199 in 2020. The highest number of shortages prevented was 282 in 2012. In Europe, the European Medicines Agency (EMA) publishes information on medication shortages that affect (or are likely to affect) more than one European Union member state. EMA also provides a list of national registers containing information on medicine shortages in individual countries.⁶⁷ As of March 2022, nine ongoing shortages were reported in the EMA database, none of them pertaining to PN components.

ASPEN has developed PN shortage considerations in order to assist clinicians in coping with PN shortages for their patients,⁶⁸ and European experts also rely on these well-made recommendations. For the most up-to-date product shortage information, the ASPEN website provides regular updates and strategies on shortages.⁶¹ As a practical example, a new PN use process was created at the Mayo Clinic, Rochester, MN, to facilitate safe PN delivery during shortages following the amino acid solution shortage (because of hurricanes Maria and Irma in Puerto Rico in 2017).⁶⁹ The target was to reduce the use of amino acid solutions by 75% per day by shifting from compounded PN to commercial MCBs whenever possible. Proposed strategies to conserve amino acid solutions in the face of shortages are shown in Box 3.⁶⁹ These strategies could also be applicable to shortages of other PN components (see statement 12 in the summary article).⁹ Strategies for coping with HPN shortages are discussed in a separate article that forms part of this supplement.⁷⁰

Box 3. Strategies to Conserve Amino Acid Solutions: Experience From The Mayo Clinic, Rochester, MN⁶⁹.

• Use MCBs in metabolically stable patients and restrict compounded PN to special cases (approval by the nutrition support service required), such as:  ◦ Volume limitation  ◦ Cardiopulmonary instability  ◦ Renal replacement therapy  ◦ Severe electrolyte derangements  ◦ High protein requirements  ◦ Other (determined by the nutrition support service) • Reduce new PN, either by delaying or deferring. • Discontinue PN as soon as enteral feeding is possible. • Decrease the dose of amino acids used for compounding. • Continuously review the PN process for ability to transition from compounded PN to MCBs.

Abbreviations: MCBs, multichamber bags; PN, parenteral nutrition.

Conclusion

PN is a complex medication associated with a high risk of error along the entire PN process, particularly during the preparation and administration steps. Despite the availability of PN guidelines, there is still room for improvement to ensure better safety. Standardization of the entire process, preferably with the help of advanced technologies guiding the prescription and preparation of PN and performing automated safety checks, is considered key for decreasing the risk of errors. However, these technologies are still not used consistently, and available systems frequently lack the functionalities to perform these tasks. In addition, 5 further PN process and formulation aspects were identified at the meeting as areas for major improvement (Figure 2). These included a need for a larger selection of MCBs, particularly with regard to protein content and ILE composition, to allow for adjustment to the individual needs of each patient. Importantly, it was recognized that MCBs do not meet every patient’s needs: MCB customization or individually compounded PN may be necessary, and, moreover, MCB use does not minimize the need for the careful evaluation of each patient’s nutritional and electrolyte requirements. However, the choice of MCBs is particularly limited in the US, which is reflected by the fact that their use in the US, in hospitals and particularly in the HPN setting, is much less common than in Europe. Another key concern of the experts was the need for better education on the importance and safety of ILEs as a part of a complete PN formulation. It is hoped that the anticipated ASPEN clinical recommendations paper for PN in critical illness will provide clinical guidance on ILE choice and facilitate the establishment of consistent daily lipid use as standard practice in the US. The experts also encouraged the use of newer-generation ILEs that contain FO. Extending the availability of vitamin and mineral/trace element product ranges, particularly in the US, to better deal with any potential shortages was also considered a key area for improvement. Finally, shortages of PN components remain a global concern, and effective strategies to deal with shortage situations are key in minimizing potential harm to the patient.

Data availability

No new data were generated or analyzed in support of this article.

Disclosures

Fresenius Kabi Deutschland GmbH and Fresenius Kabi USA provided financial support to organize and invite experts to participate as speakers, based on knowledge and international reputation within the areas of clinical nutrition, to the International Safety and Quality of Parenteral Nutrition Summit, as well as financial support for the development of this review. Fresenius Kabi had no involvement in the study design; collection, analysis, and interpretation of data; or writing of the manuscript. Dr. Martina Sintzel (mcs medical communication services, Erlenbach, Switzerland) drafted the manuscript, and Dr. Richard Clark (freelance medical writer, Dunchurch, UK) provided editorial and consultancy services; all were funded by Fresenius Kabi. These services complied with international guidelines for Good Publication Practice (GPP2022). Dr. Wischmeyer has received investigator-instigated grant funding related to his work from the National Institutes of Health, US Department of Defense, Abbott, Baxter, and Fresenius Kabi, consulting fees from Abbott, Baxter, Fresenius Kabi, Mend Inc., and Nutricia, and honoraria or travel expenses for CME lectures on improving nutrition care from Abbott, Baxter, DSM, Fresenius Kabi, Nestlé, and Nutricia. He has received an unrestricted gift donation for nutrition research from Musclesound and DSM. Dr. Klek has received speaker’s honoraria from Baxter, Braun, Fresenius Kabi, Nestlé, Nutricia, Shire, and Vipharm, and acted as an advisory board member for Fresenius Kabi, Shire, and Tracheron. Dr. Berger has received speaker’s honoraria from Baxter, Fresenius Kabi, and Nestlé. Dr. Berlana has received consulting fees and speaker’s honoraria from Baxter and Fresenius Kabi. Dr. Gray has received consulting fees from Fresenius Kabi. Dr. Ybarra has received consulting fees from Baxter, Heron Therapeutics, and Fresenius Kabi. Dr. Ayers is a consultant and speakers bureau member for Fresenius Kabi, and has received speaker’s honoraria from Baxter.