Diamond 2017.

Methods	Design: multicentre blinded RCT Setting: NICUs at multiple sites including Hospital for Sick Children, Toronto, ON, Canada; McMaster Children's Hospital, Hamilton, ON, Canada; Alberta Children's Hospital, Calgary; Stollery Children's Hospital, Edmonton, AB, Canada; and CHU Sainte‐Justine, Montreal, QC, Canada. Study enrolment: January 2009 to September 2011 I. Allocation concealment: yes II. Blinding of intervention: yes III. Blinding of outcome measurement(s): yes IV. Complete follow‐up: yes
Participants	Population: 26 preterm infants with hepatic dysfunction due to surgical conditions Inclusion criteria: infants aged < 24 months with short bowel syndrome or intestinal failure who received substantial PN support (> 40% total calories) and was demonstrating early hepatic dysfunction (Cbil: 17–50 µmol/L (1–3 mg/dL)) in the absence of sepsis. Although the age of inclusion was < 24 months, all the included infants were in preterm or borderline preterm range with the outer range of the ages being < 6 months and therefore met the current review's inclusion criteria. Age, mean (range): 6.5 (4.3 to 8.7) weeks in SMOFlipid group; 5.3 (3.5 to 7.2) weeks in Intralipid group Gestational age, mean (range): 34.5 (32.4 to 36.7) weeks in SMOFlipid group; 35.2 (33.2 to 37.1) weeks in Intralipid group Exclusion criteria: sepsis or haemodynamic instability of any cause; coagulopathy (platelets ≤ 150 000 cells/μL, or INR ≥ 1.4); hypersensitivity to fish‐, egg‐ or soybean protein or to any of the active substances or excipients; current enrolment in another clinical trial involving a surgical or pharmacological intervention; serum Cbil > 50 μmol/L; hyperlipidaemia; treatment with intravenous N‐acetylcysteine or oral ursodeoxycholic acid; renal insufficiency; disorders of fluid balance; unstable medical conditions.
Interventions	26 infants randomised; results available for 24 infants (11 infants in SMOFLipid group; 13 in Intralipid group) MPFS‐LE vs S‐LE Group 1 (n = 11): MOFS‐LE; SMOFlipid Group 2 (n = 13): S‐LE; Intralipid TPN protocol: participants received trial lipid for up to 12 weeks. Participants also ended the trial if they achieved full enteral tolerance (autonomy from PN) prior to this time point or if they developed progressive liver disease defined by a serum Cbil > 100 µmol/L for > 14 days. Lipid dosing was according to a nomogram which adjusted the amount of the lipids proportional to the enteral intake. All type of enteral formulas were allowed except the enteral fish oil solution.
Outcomes	Primary outcome: last value of the Cbil the week the infant received the last dose of the trial lipid (i.e. at 12 weeks, at full enteral tolerance or on the development of the progressive liver disease). Secondary outcomes: liver markers other than Cbil in the blood, weight, length and head circumference were assessed at baseline, week 6 and post‐trial. A full blood count at weeks 0, 4 and 8 and post‐trial. INR, C‐reactive protein, immunological markers (IL‐1, IL‐6, IL‐8, IL‐10 and IL‐12; tumour necrosis factor‐α), nephelometry, serum cholesterol and serum TGs assessed at baseline, week 6 and post‐trial. RBC PL composition assessed at baseline, week 6 and post‐trial.
Notes	The investigators also had a provision of replacement of the study participants who discontinued PN prior to second week of study due to achievement of full enteral tolerance. Mean duration on trial was 8 weeks and did not differ according to treatment (P = 0.99). Infants who received SMOFlipid were more likely to have a decrease in serum Cbil to 0 µmol/L than those in the Intralipid group over the entire observation period (hazard ratio 10.6, 95% CI 1.3 to 86.9; P = 0.006). At the time the primary end point was achieved, 3 (27%) infants in the SMOFlipid group had a serum Cbil exceeding 50 μmol/L, while 9 (69%) infants in the Intralipid group had Cbil above this level (P = 0.04). The authors did not provide a specific definition for some adverse outcomes, e.g. sepsis. Beginning and end weights were described and growth rates could not be imputed due to lack of data on covariance.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The randomisation sequence was developed by the research support pharmacy using a random number table prior to enrolment of the first patient. The sequence was developed in blocks of variable size without investigator input or knowledge." Comment: done
Allocation concealment (selection bias)	Low risk	Quote: "Allocation concealment was achieved by the randomisation sequence only being known to the research support pharmacy at the Hospital for Sick Children. The group assignment was relayed to the dispensing pharmacist at the patient's institution only after enrolment had occurred." Comment: done
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "All participants, treating clinicians, and investigators were blinded to the treatment assignment. Only the research support pharmacist at the Hospital for Sick Children and the dispensing pharmacist at the patient's institution were aware of the group assignment." Comment: possibly done (by author consensus (VK, MM)).
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "All participants, treating clinicians, and investigators were blinded to the treatment assignment. Only the research support pharmacist at the Hospital for Sick Children and the dispensing pharmacist at the patient's institution were aware of the group assignment." Comment: possibly done
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "complete outcomes reported." Comment: done
Selective reporting (reporting bias)	Low risk	Quote: "This trial was registered at clinicaltrials.gov as NCT00793195." Comment: no significant concerns as per the outcome reporting.
Other bias	Unclear risk	Quote: "The investigators also had a provision of replacement of the study participants who discontinued the PN prior to second week of the study due to achievement of full enteral tolerance. Comment: The study investigators performed an per protocol analysis as two patients in the SMOFlipid arm were not analyzed as they reached PN within 14 days. However as this is < 20% of the data it may not make a significant difference. For this reason we have assigned unclear risk of bias.