Lipid emulsions for parenterally‐fed term and late preterm infants

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

The primary objective is to compare the effectiveness and safety of lipid formulations from different sources, including soybean oil‐based, multicomponent, olive oil‐based, and fish oil–based, in parenterally‐fed term and late preterm infants.

The secondary objective is to determine the effectiveness and safety of alternative lipid emulsions compared with soybean oil‐based lipid emulsions in relation to clinical conditions (surgical patients) and patients with established parenteral nutrition‐associated liver disease.

Background

Description of the condition

Maintaining adequate growth of critically ill infants is challenging (Ehrenkranz 2000; Hay 2008). Complications in the critically ill late preterm or term infant (particularly those with congenital or acquired disease causing gastrointestinal failure) make routine enteral feeding problematic, and at times impossible. These infants do not receive enough protein and energy to achieve adequate growth (Hay 2008). Frequently, they require total parenteral nutrition (TPN) or partial parenteral nutrition (PPN) to provide all or part of their caloric requirements. Lipid emulsions have been a vital component of parenteral nutrition in critically ill infants since their introduction in the 1960s. Lipids are an attractive energy source because of their high energy density and their supply of essential fatty acids necessary for central nervous system development (Vlaardingerbroek 2012). In addition, lipids are needed to prevent essential fatty acid deficiency.

Parenteral nutrition (PN) is not without its risks. The central access needed to deliver PN is sometimes associated with nosocomial infection (Stoll 2002). In infants and children who receive PN for more than 14 days, the incidence of PN‐associated cholestasis (PNAC) and intestinal failure‐associated liver disease is high. Parenteral nutrition‐associated liver disease (PNALD) in infants can range from 40% to 85% with long‐term use of TPN (Park 2015).

Farnesoid X receptor (FXR) is a nuclear receptor expressed at high levels in the liver and intestine. FXR activity is a major regulator of excretion and metabolism of conjugated bilirubin and bile acid. Studies have demonstrated that plant‐derived phytosterols, mainly present in lipid components of PN, act as a potent inhibitor of FXR (Al‐Shahwani 2017).

Description of the intervention

Lipid emulsions serve as a source of high density energy and a source of essential fatty acids, i.e. linoleic acid (omega‐6 fatty acid) and alpha‐linolenic acid (omega‐3 fatty acid). These are precursors for eicosanoids, which are active in numerous physiological mechanisms, such as platelet function, immune response, inflammation, and early visual and neural development (Driscoll 2008; Koletzko 2001; Lapillonne 2013; SanGiovanni 2000).

Pure soybean oil‐based lipid emulsions (S‐LEs; for example Intralipid, Ivelip, Liposyn III) have been the standard lipid emulsions used in neonatal intensive care units (NICUs) worldwide for the last few decades (de Meijer 2009). However, there is evidence to suggest that S‐LEs may have harmful effects due to excessive polyunsaturated fatty acid (PUFA) and linoleic acid content (Sala‐Vila 2007). Newer lipid emulsions aim to decrease the excessive omega‐6 fatty acid content by using lipids from sources other than soybean oil.

Medium‐chain triglyceride (MCT)‐based lipid emulsions (coconut oil‐derived) decrease the omega‐6 content by adding MCT to lipid emulsions, for example, Lipovenoes MCT and 20% Lipofundin MCT/long‐chain triglyceride (LCT) are a 1:1 mix of MCT and LCT (Vanek 2012). Structured lipid emulsions (for example, Structolipid) are a modification of MCT‐LCT‐based lipid emulsions and are formed by re‐esterification of medium‐ and long‐chain fatty acids (Waitzberg 2006). Olive oil‐based lipid emulsions which are rich in the monounsaturated fatty acid, oleic acid (18:1; omega‐9), have been available since the 1990s. For example, ClinOleic is an olive oil‐based lipid emulsion with a 4:1 ratio of olive to soybean oil and one‐third of the PUFA content compared with S‐LE (e.g. 20% Intralipid). Fish oil‐containing lipid emulsions (e.g. Omegaven), which are rich in omega‐3 fatty acids and have a low ratio of omega‐6 to omega‐3, have also been developed (Wanten 2007).

More recently, lipid emulsions derived from multiple sources have become available for clinical use. SMOFLipid is one such lipid emulsion; it is a 30:30:25:15 mix of soybean oil, MCT, olive oil, and fish oil (Sala‐Vila 2007). Lipoplus, also known as Lipidem in some countries, is a mix of 50% MCT, 40% soybean oil, and 10% fish oil.

How the intervention might work

Currently available lipid emulsion formulations differ in the source of lipid, fatty acid profile, anti‐oxidant levels, and presence of additional components (Wanten 2007).

Conventional S‐LEs contribute to parenteral nutrition‐associated liver disease (PNALD) in term and preterm infants (de Meijer 2009; Xu 2012); phytosterols, present in soybean oil, may have harmful effects on liver function (de Meijer 2009). Histologically cholestatic changes in the liver can be detected in the liver as early as two weeks after commencement of PH, and evidence of fibrotic changes can be observed within six weeks (Nandivada 2013; Park 2015). The incidence of PNALD increases as the duration of the PN increases; other contributing risk factors include absence of enteral feeding, low birth weight and prematurity (Park 2015).

High amounts of linoleic acid and alpha‐linolenic acid in S‐LEs may lead to substrate inhibition of Δ⁶desaturase (Gobel 2003), resulting in decreased formation of arachidonic acid and docosahexaenoic acid, which are crucial for visual and cognitive development in premature infants (Heird 2005; Lehner 2006). S‐LEs also lead to an increase in pro‐inflammatory prostaglandins and leukotrienes (Wanten 2007), may increase the risk of sepsis (Palmblad 1991), and may adversely affect phagocytic and lymphocytic functions (Gogos 1995).

Soybean oil‐based lipid emulsions have excessive amounts of PUFA (up to 60%) and linoleic acid (50%) (Sala‐Vila 2007), which exceeds the daily preterm linoleic acid requirement of 0.25 g/kg/day and adds to oxidative stress (Koletzko 2005; Pitkanen 1991).

Medium‐chain triglyceride (coconut oil‐derived) and LCT (soybean oil‐derived)‐based lipid emulsions (MS‐LEs) may have advantages due to reduced omega‐6 content and the rapid metabolism of MCTs. Data suggest good tolerance in preterm infants with increased eicosapentaenoic acid levels and an equivalent essential fatty acid profile compared with S‐LEs (Lehner 2006). However, in vitro studies have raised concerns that MCTs may cause leucocyte activation, impair immune function, and decrease killing of Candida albicans (Waitzberg 2006; Wanten 2007). Use of MCT oil lipid emulsions has also been associated with impaired lung function and aggravation of tissue inflammation in adults with acute respiratory distress syndrome (Lekka 2004); they may also be ketogenic, which limits their utility in acidotic patients (Waitzberg 2006).

Structured lipid emulsions have an even distribution of medium‐chain fatty acids in the lipid droplets, aimed at reducing the immunological adverse effects of MS‐LEs. There is limited evidence to suggest that structured emulsions are well tolerated in critically ill patients; however, unlike MS‐LEs, they may not affect phagocyte function (Wanten 2007).

Borage oil‐soybean oil‐based lipid emulsions (BS‐LEs) substitute the soy content partially with borage oil, which is the highest source of gamma‐linolenic acid (18:3; omega‐6). The enzyme, Δ⁶desaturase, is essential in the conversion from linoleic acid to gamma‐linolenic acid and is considered the rate‐limiting step in the metabolism from linoleic acid to arachidonic acid. Borage oil‐based lipid emulsions were developed to potentially circumvent this enzymatic step. PFE 4501 (Pharmacia, Sweden) is a combination of borage oil (15%) and soybean oil (85%) with increased amounts of carnitine to prevent carnitine deficiency in preterm infants (Magnusson 1997).

Olive oil‐soybean oil‐based lipid emulsions (OS‐LEs) have generated interest due to the immune‐neutral nature of oleic acid (Reimund 2004), decreased PUFA content, higher alpha‐tocopherol content (Sala‐Vila 2007), and reduced peroxidability of low‐density lipoproteins, with an overall reduction in oxidative stress (Goulet 1999; Krohn 2006). OS‐LE (ClinOleic) has been reported to have a fatty acid composition similar to that of breast milk, and to result in higher alpha‐tocopherol levels in preterm infants when compared with S‐LE (Intralipid; Gobel 2003). Studies have reported decreased immunological disturbance, with lesser inhibition of T‐cell activation, lesser effect on interleukin‐2 production and decreased alteration in neutrophil responses with the use of OS‐LE compared with S‐LE (Buenestado 2006; Gawecka 2008a; Granato 2000). Olecanthol, a minor component in olive oil, has been shown to inhibit the cyclooxygenase pathway but not the 5‐lipoxygenase pathway, displaying "ibuprofen‐like" anti‐inflammatory activity (Beauchamp 2005). Use of OS‐LE may decrease the incidence of hyperglycaemia when compared with S‐LE (Intralipid) (Van Kempen 2006). Randomised controlled trials of critically ill neonates have shown OS‐LE to be as equally well‐tolerated as conventional S‐LE (Gawecka 2008a).

Fish oil‐containing lipid emulsions (F‐LEs) have increased omega‐3 PUFAs, resulting in inhibition of the cyclooxygenase pathway and preferential use of the lipoxygenase pathway, which in turn decreases pro‐inflammatory prostaglandins (Fürst 2000). Eicosapentaenoic acid (C20:5; omega‐3), present in fish oil, activates the peroxisome proliferator‐activated receptors, alpha and gamma, which in turn antagonise the nuclear factor‐κB signalling pathway, leading to reduced production of inflammatory mediators (Fürst 2000). Adult studies have indicated that, in sepsis, the use of F‐LE decreases the length of hospital stay, readmission rates, and rate of mechanical ventilation, and improves survival (Wanten 2007). Recently, a pure F‐LE (Omegaven) was shown to decrease and even reverse PNALD in infants, resulting in decreased mortality and lower levels of triglycerides, conjugated bilirubin and liver enzymes compared with S‐LE (20% Intralipid) (de Meijer 2009; Puder 2009).

Multisource lipid emulsions (MCT‐fish‐soybean oil‐based lipid emulsions (MFS‐LE) and MCT‐olive‐fish‐soybean oil‐based lipid emulsions (MOFS‐LE) derive the advantages of lipids from multiple sources, including MCTs (rapidly metabolised lipids), soybean oil (essential fatty acid source), olive oil (fewer immune effects), and fish oil (anti‐inflammatory effects). There is evidence of reduced hospital stay, better plasma elimination of triglycerides, better alpha‐tocopherol levels, and good tolerance profile with a MOFS‐LE (SMOFlipid) in adults (Grimm 2005; Wanten 2007). ClinOleic and Omegaven, in a 1:1 combination, have been shown to decrease cholestasis and the incidence of retinopathy of prematurity requiring laser therapy in preterm infants (Pawlik 2011). Recent meta‐analyses have shown significant decreases in incidence of cholestasis with the use of fish oil emulsions in preterm infants (Kotiya 2016; Vayalthrikkovil 2017).

There is emerging evidence from recent studies performed in mice that pure fish oil‐based lipid emulsions may have the least impact on hepatic steatosis (Nandivada 2017).

The abbreviation scheme used for the lipid emulsions is described in Appendix 1.

Why it is important to do this review

The introduction of life‐saving parenteral nutrition was a landmark in neonatal care, but it appears that the conventionally used S‐LEs are far from ideal. Despite their widespread use, conventional S‐LEs may have harmful effects in infants due to their high PUFA content and phytosterols, which may contribute to adverse outcomes, including mortality, PNALD and sepsis. The lipid emulsion of choice in infants would be one that is easy to metabolise, does not increase inflammatory or oxidative stress, is not immunosuppressive, has the least adverse effects, and has a proven safety profile. Therefore, we plan to undertake this Cochrane Review to compare the effectiveness of newer lipid emulsions to the conventionally used pure soybean oil‐based lipid emulsions in critically ill late preterm and term infants.

Other systematic reviews on the topic include the systematic review and meta‐analysis published by the European Society for Paediatric Gastroenterology Hepatology and Nutrition (ESPGHAN), which looks at the role of different LEs in the pathogenesis of cholestasis and PNALD, in a paediatric population including infants (ESPGHAN 2016). Another systematic review and meta‐analysis of randomised and observational studies has looked at the role of fish oil‐containing LE in reversing PNALD in newborns (Park 2015).

Objectives

The primary objective is to compare the effectiveness and safety of lipid formulations from different sources, including soybean oil‐based, multicomponent, olive oil‐based, and fish oil–based, in parenterally‐fed term and late preterm infants.

The secondary objective is to determine the effectiveness and safety of alternative lipid emulsions compared with soybean oil‐based lipid emulsions in relation to clinical conditions (surgical patients) and patients with established parenteral nutrition‐associated liver disease.

Methods

Criteria for considering studies for this review

Types of studies

We will include randomised controlled trials (RCTs) or quasi‐RCTs.

Types of participants

We will include term infants (gestational age 37 weeks or more) and late preterm infants (gestational age between 34 weeks and 0 days, and 36 weeks and 6 days) who received intravenous lipid emulsions as a part of either total parenteral nutrition (TPN) or partial parenteral nutrition (PPN) at any time in the first six months after birth and for any duration.

We will consider neonates who received parenteral nutrition who may have surgical conditions including: necrotising enterocolitis, gastroschisis, omphalocoele, tracheo‐oesophageal fistula, intestinal atresia, malrotation, short bowel syndrome, and meconium ileus.

We will also include term infants (gestational age 37 weeks or more) or late preterm infants (gestational age between 34 weeks and 0 days, and 36 weeks and 6 days) with parenteral nutrition‐associated liver disease (PNALD; conjugated bilirubin more than 2 mg/dL or 34.2 µmol/L), with or without raised liver enzymes within the first six months of life.

We will exclude infants with cholestasis due to inborn errors of metabolism, biliary atresia, or congenital infection.

Types of interventions

We will consider studies comparing various lipid emulsions, including newer lipid emulsions (lipids derived from olive oil, fish oil, and medium‐chain triglyceride (MCT), structured lipids, and multisource lipid emulsions), as well as conventional pure soybean oil‐based lipid emulsions, in term or late preterm infants.

Eligible lipid emulsions

Soybean oil‐based lipid emulsions (S‐LEs): lipid emulsions with 100% lipids derived solely from soybean oil.

• Intralipid

• Ivelip

• Liposyn III

Fish oil‐containing lipid emulsions: all fish oil‐containing lipid emulsions.

• MCT‐olive‐fish‐soybean oil‐based lipid emulsions (MOFS‐LEs), e.g. SMOFlipid

• Multisource lipid emulsions (MCT‐fish‐soybean oil‐based lipid emulsions (MFS‐LEs)), e.g. Lipidem

• Pure fish oil (F‐LE), e.g. Omegaven

We will also consider any LE or multi‐component LE that includes fish oil as one of the constituents.

Alternative lipid emulsions: all lipid emulsions with partial or complete substitution of soybean oil from other sources (decreased linoleic acid content), but not containing fish oil.

• Olive oil‐soybean oil‐based lipid emulsions (OS‐LEs), e.g. ClinOleic

• MCT‐soybean oil‐based lipid emulsions (MS‐LEs), e.g. Lipovenoes MCT

• Borage oil‐soybean oil‐based lipid emulsions (BS‐LEs), e.g. PFE 4501

• Structured lipids (structured MCT‐soybean oil), e.g. Structolipid

See Appendix 1 for a list of abbreviations for eligible lipid emulsions.

We will consider the following comparisons

• Fish oil lipid emulsion versus all non‐fish oil lipid emulsions

• Fish oil lipid emulsion versus other fish oil lipid emulsion

• Alternative lipid emulsion versus soybean oil‐based lipid emulsion (S‐LE)

• Alternative lipid emulsion versus other alternative lipid emulsion

Details of all possible comparisons are noted in Appendix 2.

For term infants with underlying clinical conditions that require parenteral nutrition (surgical conditions) or a change in parenteral nutrition solutions (established cholestasis), we will include studies using lipid emulsions as a part of TPN or PPN at any time. We will not place restrictions on minimum or maximum dose of lipid emulsions.

We will not place restrictions on co‐interventions of amino acids, minerals, trace elements or vitamins for parenteral nutrition and expressed breast milk or formula feeds via a nasogastric tube for PPN.

Types of outcome measures

Primary outcomes

• Parenteral nutrition‐associated liver disease (PNALD) (conjugated bilirubin ≥ 2 mg/dL or 34.2 µmol/L) with or without raised liver enzymes (alanine aminotransferase (ALT) > 45 IU/L, alkaline phosphatase > 420 IU/L) in the absence of other causes (Christensen 2007; Robinson 2008), in late preterm and term infants without parenteral nutrition‐associated liver disease at study entry

• Resolution of PNALD (conjugated bilirubin < 2 mg/dL or < 34.2 µmol/L), in late preterm and term infants with established PNALD (Lam 2014)

• Physical growth

  • Days to regain birth weight

  • Growth rate (grams/kg/day) during study period and hospital stay (Fenton 2017)

Secondary outcomes

• Head growth

  • Head circumference below 3% at discharge

  • Head growth velocity (cm/week)

• Length

  • Rate of growth

  • Length velocity (cm/week)

• Body composition: measured at corrected term gestation by magnetic resonance (MR) spectroscopy and magnetic resonance imaging (MRI) (Ahmad 2010; Roggero 2007; Uthaya 2016)

  • Intrahepatocellular lipid content (IHCL; intra‐hepatic lipid: water ratio) values

  • Non‐adipose tissue mass

• Proven sepsis: blood culture positive

• Necrotising enterocolitis (NEC) ≥ stage 2 on Bell's staging system (Bell 1978)

• Significant jaundice: requiring treatment with phototherapy or exchange transfusion

• Duration of phototherapy (days)

• Duration of ventilation (total days)

• Duration of supplemental oxygen (total days)

• Need for home oxygen therapy

• Duration of hospital stay (days)

• Thrombocytopenia (platelets < 50,000)

• Hypertriglyceridaemia, defined by serum triglyceride levels > 200 mg/dL (2.25 mmol/L; Putet 2000)

• Mean conjugated bilirubin levels (µmol/L)

• Mean gamma glutamyl transferase (GGT) levels (IU/L)

• Mean alanine aminotransferase (ALT) levels (IU/L)

• Mean alkaline phosphatase (ALP) levels (IU/L)

• Mean triglyceride levels (mmol/L)

• Time to development of PNALD (days)

• Time to resolution of PNALD (days)

• Hyperglycaemia (blood sugar level > 8.3 mmol/L or > 150 mg/dL) (Sinclair 2011) or hypoglycaemia (blood sugar level < 2.6 mmol/L or < 46 mg/dL)

• Essential fatty acid (EFA) deficiency defined by triene/tetraene ratio > 0.05 (Cober 2010; Gura 2005)

• Other markers of EFA deficiency

• Need for liver transplantation due to PNALD related liver failure

• Death before discharge or neonatal death (within the first 28 days of life)

• Neurodevelopmental outcome (assessed by a standardised and validated assessment tool or a child developmental specialist) at any age reported (outcome data grouped at 12, 18, and 24 months if available)

Search methods for identification of studies

We will use the criteria and standard methods of Cochrane and Cochrane Neonatal (see the Cochrane Neonatal search strategy for specialized register). We will search for errata or retractions from included studies published in full‐text on PubMed (www.ncbi.nlm.nih.gov/pubmed), and we will report the date this was done within the review.

Electronic searches

We will conduct a comprehensive search including: the Cochrane Central Register of Controlled Trials (CENTRAL, current issue) in the Cochrane Library; MEDLINE via PubMed (1996 to current); Embase (1980 to current); and CINAHL (1982 to current). Details are noted in Appendix 3.

We will not apply language restrictions. We will search clinical trial registries for ongoing or recently completed trials (ClinicalTrials.gov, the World Health Organization's International Trials Registry and Platform, and the ISRCTN Registry).

Searching other resources

Additionally, we will review the reference lists of all identified articles for relevant articles not identified in the primary search.

Data collection and analysis

We will use the standard methods of Cochrane Neonatal for data collection and analysis. Data extraction forms will be specifically designed for this review, tested on two studies, further refined and then used to collect and collate data. For each included study, we will record details regarding the method of randomisation, allocation concealment, blinding, intervention, stratification, and whether the study was single‐centre or multi‐centred. We will extract data regarding participants, parenteral nutrition (PN) details, and reported outcomes.

We will record the selection process in sufficient detail to complete a PRISMA flow diagram (Moher 2009), and 'Characteristics of excluded studies' table.

Selection of studies

Two review authors (VK, MM) will independently search the databases to identify articles eligible for inclusion in the review. We will assess methodology with regard to blinding of randomisation, allocation concealment, intervention and outcome measurements, and completeness of follow‐up.

Data extraction and management

Two review authors (VK, MM) will separately extract the data for each study on data extraction forms. One review author (VK) will enter data into Review Manager 5 (Review Manager 2014), and the other review author (MM) will cross‐check the printout against his own data extraction forms. At each stage, any difference in opinion will be resolved by discussion.

Assessment of risk of bias in included studies

Two review authors (VK, MM) will independently assess the risk of bias (low, high, or unclear) of all included trials using the Cochrane 'Risk of bias' tool for the following domains (Higgins 2017).

• Sequence generation (selection bias)

• Allocation concealment (selection bias)

• Blinding of participants and personnel (performance bias)

• Blinding of outcome assessment (detection bias)

• Incomplete outcome data (attrition bias)

• Selective reporting (reporting bias)

• Any other bias

Any disagreements will be resolved by discussion or by a third review author (RS). See Appendix 4 for a more detailed description of risk of bias for each domain.

Measures of treatment effect

We will follow the recommendations of Cochrane Neonatal, and use a fixed‐effect model for meta‐analysis. We will estimate the treatment effects for categorical outcomes using the typical risk ratio (RR) and typical risk difference (RD) with 95% confidence intervals (CIs). We will estimate the number needed to treat for an additional beneficial outcome (NNTB) and number needed to treat for an additional harmful outcome (NNTH) if the RD is statistically significant. For continuous outcomes, we will use the mean difference (MD) with 95% CIs to describe the data.

Unit of analysis issues

The unit of analysis will be the participating infant in individually randomised trials and the neonatal unit for cluster‐randomised trials. If cluster‐randomised trials are included in the analyses, we will adjust their sample sizes using the methods described in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2017).

Dealing with missing data

We will contact the authors of included studies if clarifications or additional information are required. In the case of missing data, we will describe the number of participants with missing data in the results section and in the 'Characteristics of included studies' table.

Assessment of heterogeneity

We will estimate treatment effects in individual trials and examine heterogeneity between trials by inspecting forest plots and quantifying the impact of heterogeneity by using the I² statistic, a measure that describes the proportion of variation in point estimates that is due to variability across studies rather than sampling error (Deeks 2017). We will interpret results as follows.

• Less than 25%: no heterogeneity

• 25% to 49%: low heterogeneity

• 50% to 74%: moderate heterogeneity

• 75% to 100%: high heterogeneity

If we detect statistical heterogeneity, we plan to explore possible causes (e.g. differences in study quality, participants, intervention regimens or outcome assessments) by performing post hoc subgroup analyses.

Assessment of reporting biases

If there are more than 10 studies for any outcome in any comparison, we will use funnel plots to assess publication bias (Sterne 2017). We will identify and evaluate multiple reports of a single study (multiple publication bias) by comparing the reported baseline characteristics and the author details with clarifications requested from authors, if required, to avoid double‐counting.

Data synthesis

We will perform meta‐analyses using Review Manager 2014, Cochrane's software for preparing and maintaining systematic reviews. For estimates of typical RR and typical RD, we will use the Mantel‐Haenszel method (Higgins 2017). We will use the inverse variance method for measured quantities. We will carry out and report all primary meta‐analyses using the fixed‐effect model, according to the recommendations of Cochrane Neonatal.

Details of calculations and imputations

We will replace any standard error of mean by the corresponding standard deviation (SD). If the data are described in medians and interquartile ranges, we will substitute medians for means and impute the corresponding SDs by dividing interquartile ranges by 1.35. If the data are described in medians and ranges, then we will use the formulae proposed by Hozo and colleagues to impute the SD (Hozo 2005). We will pool the means and SDs of weekly observations in a group of study participants using the formulae for pooling means and variances (McNaught 1997). For combining multiple groups' means and SDs, the formulae we will use are as described for pooling means and SDs in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2017; Furukawa 2006).

Where we cannot perform meta‐analyses, we plan to present qualitative inferences as systematically as possible and explain why we cannot perform meta‐analyses. We will present the results for important outcomes in the 'Summary of findings' tables.

Quality of evidence

We will use the GRADE approach, as outlined in the GRADE Handbook (Schünemann 2013), to assess the quality of evidence for the following (clinically relevant) outcomes.

• PNALD (conjugated bilirubin ≥ 2 mg/dL or 34.2 µmol/L) with or without raised liver enzymes in the absence of other causes (Christensen 2007; Robinson 2008)

• Resolution of PNALD/cholestasis (in infants originally enrolled with underlying liver disease or cholestasis)

• Physical growth: weight

  • Days to regain birth weight

  • Growth rate (grams/kg/day) during study period and hospital stay (Fenton 2017)

• Head growth

  • Head circumference below 3% at discharge

  • Head growth velocity

• Death before discharge or neonatal death (within the first 28 days of life)

• Proven sepsis: blood culture positive

• Neurodevelopmental outcome (neurodevelopmental outcome assessed by a standardised and validated assessment tool or a child developmental specialist) at any age reported (outcome data grouped at 12, 18, and 24 months if available).

Two review authors will independently assess the quality of the evidence for each of the outcomes above. We will consider evidence from randomised controlled trials as high‐quality, but will downgrade our assessments of the evidence by one level for serious (or two levels for very serious) limitations based upon the following: design (risk of bias), consistency across studies, directness of the evidence, precision of estimates and presence of publication bias. We will use the GRADEpro Guideline Development Tool (GRADEpro GDT 2015) to create a ‘Summary of findings’ table to report the quality of the evidence.

The GRADE approach results in an assessment of the quality of a body of evidence according to one of the following four grades.

1. High quality: we are very confident that the true effect lies close to that of the estimate of the effect.

2. Moderate quality: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

3. Low quality: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect.

4. Very low quality: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

Subgroup analysis and investigation of heterogeneity

We will explore high statistical heterogeneity in the outcomes by visually inspecting the forest plots and by removing the outlying studies in the sensitivity analysis (Deeks 2017). Where statistical heterogeneity is significant, we will interpret the results of the meta‐analyses accordingly; and we will downgrade the quality of the evidence in the ‘Summary of findings’ tables, according to the GRADE recommendations.

We will consider the following groups for subgroup analysis where specific subgroup data are available.

• Age of enrolled infants (first week versus later)

• Gender

• Gestational age: term infants (gestational age 37 weeks or more); late preterm infants (gestational age between 34 weeks and 0 days, and 36 weeks and 6 days)

• Critically ill infants or those requiring surgery (surgical conditions in neonates include: necrotizing enterocolitis, gastroschisis, omphalocoele, tracheo‐oesophageal fistula, intestinal atresia, malrotation, short bowel syndrome, and meconium ileus)

• Lipid dosing (actual dose received daily; cumulative)

• Continuous intravenous lipid versus intermittent intravenous lipid

Sensitivity analysis

We will present results of the sensitivity analyses only if these are significantly different from the primary results. We will perform sensitivity analyses in the following situations.

• If there is unexplained moderate to high heterogeneity, we will explore this by removing the outlying study/studies causing heterogeneity (if feasible).

• If a study with high risk of (material) bias is included in the meta‐analysis of an outcome where the other studies have low risk of bias, we will remove the study with high risk of bias.

Appendices

Appendix 1. Abbreviations for lipid emulsions

Soybean oil‐based lipid emulsions (S‐LE): lipid emulsions with 100% lipids derived solely from soybean oil

• Intralipid

• Ivelip

• Liposyn III

Fish oil‐containing lipid emulsions: all fish oil‐containing lipid emulsions (MOFS‐LE, MFS‐LE, and pure F‐LE)

• MOFS‐LE (MCT‐olive‐fish‐soybean oil) e.g. SMOFlipid

• MFS‐LE (MCT‐fish‐soybean oil) e.g. Lipidem

• Pure F‐LE (pure fish oil) e.g. Omegaven

We will also consider any other combination LE that contain fish oil if available at the time of the review.

Alternative lipid emulsions: all alternative lipid emulsions with partial or complete substitution of soybean oil from other sources (decreased linoleic acid content)

• OS‐LE (olive‐soybean oil) e.g. ClinOleic

• MS‐LE (MCT‐soybean oil) e.g. Lipovenoes MCT

• BS‐LE (borage‐soybean oil) e.g. PFE 4501

• Structured lipids (structured MCT‐soybean oil) e.g. Structolipid

Specific lipid components have been denoted by the following letters: soy by ‘S’; MCT (from coconut oil) by ‘M’; fish oil by ‘F’; olive oil by ‘O’; borage oil by ‘B’. The abbreviations for the ‘alternative lipid emulsion' end in the letter ‘S’ (if containing soybean oil) for consistency in nomenclature and to indicate the common theme of substitution of soybean oil by lipids from alternative sources (e.g. olive‐soy is abbreviated as ‘OS‐LE’; MCT‐soy as ‘MS‐LE’; MCT‐olive‐fish‐soy as ‘MOFS‐LE’). Further, except the letter ‘S’ (which is always the last letter in the lipid emulsion abbreviations), the sequence of letters denoting the other lipid components are in the decreasing order of lipid percentage (as found in commonly available preparations) e.g. in MFS‐LE (e.g. Lipidem), the percentage of MCT > percentage of fish oil; and in MOFS‐LE (e.g. SMOFlipid) the percentage of MCT (30%) > percentage of olive oil (25%) > percentage of fish oil (15%).

Appendix 2. Comparisons of lipid emulsions

We will consider newer lipid emulsions with partial or complete substitution of soy by lipids from other sources in the intervention group. We will also look at newer lipid emulsions compared to each other.

We will consider the following interventions/comparisons.

1. Fish oil‐containing lipid emulsion versus all non‐fish oil lipid emulsion

• MOFS‐LE (MCT‐olive‐fish‐soybean oil) versus S‐LE

• MFS‐LE (MCT‐fish‐soybean oil) versus S‐LE

• Pure F‐LE (pure fish oil) versus S‐LE

• MOFS‐LE, e.g. SMOFlipid® versus OS‐LE, e.g. ClinOleic®

• MOFS‐LE, e.g. SMOFlipid® versus MS‐LE, e.g. Lipovenoes MCT®

• MOFS‐LE, e.g. SMOFlipid® versus BS‐LE, e.g. PFE 4501®

• MOFS‐LE, e.g. SMOFlipid® versus Structured LE, e.g. Structolipid®

• MFS‐LE, e.g. Lipidem® versus OS‐LE, e.g. ClinOleic®

• MFS‐LE, e.g. Lipidem® versus MS‐LE, e.g. Lipovenoes MCT®

• MFS‐LE, e.g. Lipidem® versus BS‐LE, e.g. PFE 4501®

• MFS‐LE, e.g. Lipidem® versus Structured LE, e.g. Structolipid

• Pure F‐LE, e.g. Omegaven® versus BS‐LE, e.g. PFE 4501®

• Pure F‐LE, e.g. Omegaven® versus Structured LE, e.g. Structolipid®

• Pure F‐LE, e.g. Omegaven® versus MS‐LE, e.g. Lipovenoes MCT®

• Pure F‐LE, e.g. Omegaven® versus OS‐LE, e.g. ClinOleic®

We will also consider any other combination LE containing fish oil in this comparison if available at the time of the review.

2. Fish oil LE versus other fish oil‐containing lipid emulsion

• MOFS‐LE, e.g. SMOFlipid® versus Pure F‐LE, e.g. Omegaven®

• MFS‐LE, e.g. Lipidem® versus Pure F‐LE, e.g. Omegaven®

• MOFS‐LE, e.g. SMOFlipid® versus MFS‐LE, e.g. Lipidem®

3. Alternative lipid emulsion versus soybean oil‐based lipid emulsion (S‐LE)

• OS‐LE (olive‐soybean oil) versus S‐LE

• MS‐LE (MCT‐soybean oil) versus S‐LE

• BS‐LE (borage‐soybean oil) versus S‐LE

• Structured lipids (structured MCT‐soybean oil) versus S‐LE

4. Alternative lipid emulsion versus other alternative lipid emulsion

• OS‐LE, e.g. ClinOleic® versus MS‐LE, e.g. Lipovenoes MCT®

• OS‐LE, e.g. ClinOleic® versus BS‐LE, e.g. PFE 4501®

• OS‐LE, e.g. ClinOleic® versus Structured LE, e.g. Structolipid®

• MS‐LE, e.g. Lipovenoes MCT® versus BS‐LE, e.g. PFE 4501®

• MS‐LE, e.g. Lipovenoes MCT® versus Structured LE, e.g. Structolipid®

• BS‐LE, e.g. PFE 4501® versus Structured LE, e.g. Structolipid®

Appendix 3. Search Strategy

MEDLINE Database: Ovid MEDLINE(R) Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R) <1946 to Present> Search Strategy: ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ 1 exp Parenteral Nutrition/ 2 infusions, intravenous/ 3 Fat Emulsions, Intravenous/ 4 (parenteral$ adj2 (fed or feed$ or nutrition$)).ti,ab. 5 (TPN or PPN or PN).mp. 6 (intravenous adj2 (infus$ or emulsion$)).ti,ab. 7 ("i.v." adj2 (infus$ or emulsion$)).ti,ab. 8 1 or 2 or 3 or 4 or 5 or 6 or 7 9 exp Lipids/ 10 (coconut$ or borage$ or fish$ or olive$ or soy$ or soybean$).mp. 11 ((alternative or conventional or multisource) adj LE).mp. 12 (alternative‐LE or conventional‐LE or multisource‐LE).mp. 13 structured MCT$.mp. 14 (arachidon$ or BS‐LE or clinoleic$ or DHA or docosahexaenoic acid$ or eicosapentaenoic acid$ or EPA or F‐LE).mp. 15 (intralipid$ or ivelip$).mp. 16 (LCT$ or linolenic$ or linoleic$ or lipidem$ or lipoplus$ or liposyn$ or lipovenoes$ or lipofundin$).mp. 17 (MCT‐fish or MCT‐olive or MCT‐soy or MFS‐LE or MOFS$ or MOFSLE$ or MCT$ or MS‐LE or MUFS$ monounsaturated).mp. 18 (omega‐6$ or omega‐3$ or omegaven$ or OS‐LE).mp. 19 (PFE 4501$ or PFE4501$ or polyunsaturated$ or PUFA$).mp. 20 (S‐LE or SMOF$ or structolipid$ or triacylgl$ or triglyc$).mp. 21 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 22 8 and 21 23 exp Infant, Newborn/ 24 Premature Birth/ 25 (neonat$ or neo nat$).ti,ab. 26 (newborn$ or new born$ or newly born$).ti,ab. 27 (preterm or preterms or pre term or pre terms).ti,ab. 28 (preemie$ or premie or premies).ti,ab. 29 (prematur$ adj3 (birth$ or born or deliver$)).ti,ab. 30 (low adj3 (birthweight$ or birth weight$)).ti,ab. 31 (lbw or vlbw or elbw).ti,ab. 32 infan$.ti,ab. 33 (baby or babies).ti,ab. 34 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 35 22 and 34 36 randomised controlled trial.pt. 37 controlled clinical trial.pt. 38 randomized.ab. 39 placebo.ab. 40 drug therapy.fs. 41 randomly.ab. 42 trial.ab. 43 groups.ab. 44 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 45 exp animals/ not humans.sh. 46 44 not 45 47 35 and 46

Appendix 4. Risk of bias tool

We will use the standard methods of Cochrane and Cochrane Neonatal to assess the methodological quality of the trials. For each trial, we will seek information regarding the method of randomisation, blinding and reporting of all outcomes of all the infants enrolled in the trial. We will assess each criterion as being at a low, high, or unclear risk of bias. Two review authors will separately assess each study and we will resolve any disagreement by discussion. We will add this information to the 'Risk of bias' tables that form part of the 'Characteristics of included studies' tables. We will evaluate the following issues and enter the findings into the 'Risk of bias' table.

1. Random sequence generation (checking for possible selection bias). Was the allocation sequence adequately generated?

For each included study, we will categorise the method used to generate the allocation sequence as:

• low risk (any truly random process e.g. random number table; computer random number generator);

• high risk (any non‐random process e.g. odd or even date of birth; hospital or clinic record number); or

• unclear risk.

2. Allocation concealment (checking for possible selection bias). Was allocation adequately concealed?

For each included study, we will categorise the method used to conceal the allocation sequence as:

• low risk (e.g. telephone or central randomisation; consecutively numbered sealed opaque envelopes);

• high risk (open random allocation; unsealed or non‐opaque envelopes, alternation; date of birth); or

• unclear risk

3. Blinding of participants and personnel (checking for possible performance bias). Was knowledge of the allocated intervention adequately prevented during the study?

For each included study, we will categorise the methods used to blind study participants and personnel from knowledge of which intervention a participant received. We will assess blinding separately for different outcomes or class of outcomes. We will categorise the methods as:

• low risk, high risk or unclear risk for participants; and

• low risk, high risk or unclear risk for personnel.

4. Blinding of outcome assessment (checking for possible detection bias). Was knowledge of the allocated intervention adequately prevented at the time of outcome assessment?

For each included study, we will categorise the methods used to blind outcome assessment. We will assess blinding separately for different outcomes or class of outcomes. We will categorise the methods as:

• low risk for outcome assessors;

• high risk for outcome assessors; or

• unclear risk for outcome assessors.

5. Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations). Were incomplete outcome data adequately addressed?

For each included study and for each outcome, we will describe the completeness of data including attrition and exclusions from the analysis. We will note whether attrition and exclusions were reported, the numbers included in the analysis at each stage (compared with the total randomised participants), reasons for attrition or exclusion where reported, and whether missing data were balanced across groups or were related to outcomes. Where sufficient information is reported or supplied by the trial authors, we will re‐include missing data in the analyses. We will categorise the methods as:

• low risk (< 20% missing data);

• high risk (≥ 20% missing data); or

• unclear risk.

6. Selective reporting bias. Are reports of the study free of suggestion of selective outcome reporting?

For each included study, we will describe how we investigated the possibility of selective outcome reporting bias and what we found. For studies in which study protocols were published in advance, we will compare prespecified outcomes versus outcomes eventually reported in the published results. We will assess the methods as:

• low risk (where it is clear that all of the study's prespecified outcomes and all expected outcomes of interest to the review have been reported);

• high risk (where not all the study's prespecified outcomes have been reported; one or more reported primary outcomes were not prespecified outcomes of interest and are reported incompletely and so cannot be used; study fails to include results of a key outcome that would have been expected to have been reported); or

• unclear risk.

7. Other sources of bias. Was the study apparently free of other problems that could put it at a high risk of bias?

For each included study, we will describe any important concerns we had about other possible sources of bias (for example, whether there was a potential source of bias related to the specific study design or whether the trial was stopped early due to some data‐dependent process). We will assess whether each study was free of other problems that could put it at risk of bias as:

• low risk;

• high risk;

• unclear risk

If needed, we plan to explore the impact of the level of bias through undertaking sensitivity analyses.

Contributions of authors

For the original review (Kapoor 2015), all review authors were involved in conceiving and designing the protocol.

RS, VK and MM performed the update of the protocol (2018) to address the requests of UK guideline developers of the National Institute for Health Research (NIHR).

Sources of support

Internal sources

• No sources of support supplied

External sources

• Vermont Oxford Network, USA.

  Cochrane Neonatal Reviews are produced with support from Vermont Oxford Network, a worldwide collaboration of health professionals dedicated to providing evidence‐based care of the highest quality for newborn infants and their families.

• National Institute for Health Research, UK.

  Editorial support for Cochrane Neonatal has been funded with funds from a UK National Institute of Health Research (NIHR) Cochrane Programme Grant (16/114/03). The views expressed in this publication are those of the authors and not necessarily those of the National Health Service, the NIHR or the UK Department of Health.

Declarations of interest

VK has no interests to declare. MM has no interests to declare. RS has no interests to declare.

Notes

This protocol and the companion protocol of "Lipid emulsions for parenterally‐fed preterm infants" will replace the published review titled "Alternative lipid emulsions versus pure soybean oil‐based lipid emulsions for parenterally fed preterm infants" (Kapoor 2015).

New