How much should you feed a patient in shock, and when?

Published 2026-09-15 · Updated 2026-09-15 · Last checked against sources 2026-09-15

Before I looked it up

ICU nutrition is always one of the things I agonise over. I mostly see shock patients, and for about the first day I give almost no calories (of course, if the blood sugar gets too low I give some..). From day 2–3 I start TPN, and once the norepinephrine comes down to 0.1 (µg/kg/min) or below I start feeding and adjust the EN composition to the patient. Depending on the case I add trace elements, thiamine and so on… This time I had a look at the recent trends in ICU nutrition.

What I found

The note I started from was a calorie-selection scheme — under 70% in the acute phase, 70–100% from day 3, full energy after a week, protein “more important than calories”, 1.0 versus 1.5 kcal/mL by fluid allowance. It cited ESPEN, ASPEN, the Japanese and the French guidelines. I read all four in full and the trials they lean on, with the question my own practice raises: a patient on norepinephrine — when, which route, how much, and does protein matter?

• Four guidelines, four different first weeks. ESPEN 2023: acute phase no more than 70% of measured energy expenditure, 80–100% after day 3, protein 1.3 g/kg/day reached progressively. ASPEN 2022: 12–25 kcal/kg/day for the first 7–10 days, protein 1.2–2.0. The French SRLF 2025: 6–8 kcal/kg/day and 0.2–0.9 g/kg protein for the whole first week. The Japanese JCCNG 2024: against deliberately feeding below energy expenditure in the acute phase, and protein above 1.2. They read the same trials.
• The trials point one way — feeding less in the first week does not kill anyone. NUTRIREA-3 (3,044 ventilated patients in shock): 6 kcal/kg with 0.2–0.4 g/kg protein versus 25 kcal/kg with 1.0–1.3 for seven days — 90-day mortality 41.3% versus 42.8%, readiness for ICU discharge a day sooner, less vomiting, less diarrhoea, half the bowel ischaemia, less liver dysfunction. PermiT, EDEN and EPaNIC each found the lower arm the same or better. A 2026 post hoc analysis: kidneys were no worse either.
• When to start enteral feeding in shock. ESPEN: delay EN while shock is uncontrolled; start low-dose EN as soon as it is controlled with fluids and vasopressors, watching for bowel ischaemia. SCCM/ASPEN 2016: withhold until resuscitated, consider it with caution while vasopressors are being weaned. JCCNG 2024: against EN during haemodynamic instability (weak, very low certainty). No guideline gives a norepinephrine number. The one number in the literature is observational — 52,563 Japanese patients: early EN was associated with lower 28-day mortality at norepinephrine under 0.1 and at 0.1–0.3 µg/kg/min, and with no difference at 0.3 or more.
• Early EN at full dose in shock does harm. NUTRIREA-2 (2,410): normocaloric EN versus PN within 24 hours of intubation — mortality the same (37% versus 35%), but with EN vomiting 34% versus 20%, bowel ischaemia 2% versus under 1%, more colonic pseudo-obstruction. The problem was the amount, not the route — which is what NUTRIREA-3 then showed.
• When to start parenteral nutrition. ESPEN: if EN is contraindicated, PN within three to seven days (Grade B). Its sepsis section: try EN for the first three to four days; if that is impossible or insufficient after day 3, PN up to about half of requirements. ASPEN 2022: PN or EN in the first week gives the same outcomes (strong recommendation), and no supplemental PN before day 7 (strong). The Early PN trial (1,372 patients with short-term contraindications to EN): PN within an hour versus standard care (nutrition after a mean of 2.8 days) — no difference in death or infection, 0.47 fewer ventilator days. EPaNIC (4,640): supplemental PN at 48 hours versus day 8 — the late group had fewer infections (22.8% versus 26.2%) and left sooner. ESPEN’s reading: what hurt in EPaNIC was early PN driven to a full, equation-derived target.
• Measuring energy. Indirect calorimetry is the reference (ESPEN Grade B; JCCNG weak), but the French guideline says not to use it routinely with the aim of improving outcomes. Without it, REE = VCO₂ × 8.19 (respiratory quotient assumed 0.85; validated in 84 patients — better than equations, worse than calorimetry); without that, 20–25 kcal/kg (ESPEN) or 12–25 (ASPEN). Predictive equations are off by up to 60%.
• Formula density. TARGET (3,957): 1.5 versus 1.0 kcal/mL at the same rate — 90-day mortality 26.8% versus 25.7%, no difference in anything. Density is chosen for fluid, not survival.
• “Protein matters more than calories” is a 2016 sentence. Three trials since: EFFORT Protein (1,301 patients, 2.2 or more versus 1.2 g/kg or less) — no gain in discharge alive, a harm signal in acute kidney injury and high SOFA scores (a Bayesian reanalysis put the probability of harm at 72%); PRECISe (935, 2.0 versus 1.3) — worse quality of life and more gastrointestinal intolerance with high protein; NUTRIREA-3 — 0.2–0.4 g/kg in the first week no worse than 1.0–1.3. ESPEN 2023 now writes that there is no support for doses above 1.3 g/kg. In one observational cohort (PROTINVENT), under 0.8 g/kg for the first three days, then 0.8–1.2, then over 1.2 after day 5 carried the lowest mortality.
• Refeeding. For high-risk patients (NICE: BMI under 16, little or no intake for over five days, alcohol, low phosphate, potassium or magnesium): no more than 10 kcal/kg/day (5 in extreme risk) or 10–20 kcal/kg for the first 24 hours, then up by a third of goal every one to two days (ASPEN 2020). Thiamine 100 mg before feeding and 100 mg a day for five to seven days (ASPEN), or 200–300 mg (NICE). Phosphate, potassium and magnesium two to three times a day (ESPEN). My note’s “start at 50–75%” and “sodium under 1 mmol/kg/day” — I could not find a primary source for either; 50% is the NICE figure for over five days without intake, the rest circulates in protocol tables.
• Gastric residual volume. Not measured routinely (SCCM 2016, SRLF 2025); if measured, do not hold feeds for under 500 mL (SCCM), or delay feeding above 500 mL per 6 hours (ESPEN). The 300 in my note’s “300–500” is habit.
• Micronutrients and thiamine. With PN, daily (ESPEN Grade B). High-dose vitamin C, thiamine, selenium and vitamin D trials have all failed to change outcomes (ESPEN 2023). Replete in deficiency, and after more than two weeks of continuous renal replacement therapy.

What I was reading

The ESPEN 2023 practical guideline (full text, recommendation numbers and grades as printed), ASPEN 2022 (full text; five GRADE questions), SCCM/ASPEN 2016 (full text), the Japanese Critical Care Nutrition Guideline 2024 and the French SRLF–GFRUP 2025 consensus (both full text); the two shock trials, NUTRIREA-2 and NUTRIREA-3 (with the 2026 renal post hoc analysis); three PN-timing trials — EPaNIC, CALORIES, the Early PN trial; three calorie trials — TARGET, PermiT, EDEN; two protein trials — EFFORT Protein (with its 2024 Bayesian reanalysis) and PRECISe; three observational studies (Ohbe 2020 by vasopressor dose, Zusman 2016 on calorimetry-based adequacy, PROTINVENT); the VCO₂ validation (Stapel 2015); and two refeeding documents (ASPEN 2020 consensus, Mehanna 2008 summarising NICE). The note I started from cited “ASPEN/SCCM 2024” — the current ASPEN critical care guideline is the 2022 edition (JPEN 2022;46:12); there is no 2024 edition. What follows is what I read in September 2026.

0. The patient in shock — when to start

Evidence: guideline wording (ESPEN Grade B and good practice point; SCCM expert consensus; JCCNG weak/very low), two randomised trials (NUTRIREA-2, NUTRIREA-3), one propensity-matched observational study (Ohbe 2020).

• The guidelines say “low dose once controlled”, not “once the vasopressor is off”. ESPEN 2023, recommendation 11: delay EN “if shock is uncontrolled and haemodynamic and tissue perfusion goals are not reached, whereas low dose EN can be started as soon as shock is controlled with fluids and vasopressors/inotropes, while remaining vigilant for signs of bowel ischaemia.” SCCM/ASPEN 2016, B5: withhold EN in haemodynamic instability until resuscitated and/or stable; “initiation/reinitiation of EN may be considered with caution in patients undergoing withdrawal of vasopressor support.” JCCNG 2024, CQ1-6: against EN in haemodynamically unstable patients (2D) — while noting that in practice small amounts are often given.
• There is no number in any guideline. None of them says how many micrograms per kilogram per minute of norepinephrine counts as “controlled”. The only number comes from one observational study — 52,563 ventilated patients with shock in the Japanese national database (Ohbe 2020), EN within two days of ventilation versus later, propensity-matched within vasopressor strata: 28-day mortality −2.9% (−4.5 to −1.3) at under 0.1, −6.8% (−9.6 to −4.0) at 0.1–0.3, and −1.4% (−7.4 to +4.7, no difference) at 0.3 or more. Early EN was associated with benefit at medium doses, not only low ones.
• Early EN at a proper dose is harmful in shock. NUTRIREA-2: normocaloric (20–25 kcal/kg) EN versus PN within 24 hours of intubation — 28-day mortality 37% versus 35% (p = 0.33), ICU-acquired infection 14% versus 16% (no difference); in the EN group vomiting 34% versus 20%, diarrhoea 36% versus 33%, bowel ischaemia 19 versus 5 patients (2% versus under 1%; hazard ratio 3.84), colonic pseudo-obstruction 11 versus 3. The PN group switched to EN once shock had resolved (24 hours off vasopressors, lactate under 2).
• Hence NUTRIREA-3. Same population (3,044): 6 kcal/kg plus 0.2–0.4 g/kg protein versus 25 kcal/kg plus 1.0–1.3 g/kg for seven days — 90-day mortality 41.3% versus 42.8% (−1.5%, −5.0 to +2.0); median time to readiness for ICU discharge 8 versus 9 days (hazard ratio 1.12); less vomiting (0.77), diarrhoea (0.83), bowel ischaemia (0.50) and liver dysfunction (0.92). The 2026 post hoc analysis (3,036): acute kidney disease 44.6% versus 46.1% (no difference), lower peak urea in the low group, no difference in renal replacement — the same in patients with pre-existing renal dysfunction.
• When to add PN. ESPEN 2023, recommendation 17: “In case of contraindications to oral and EN, PN should be implemented within three to seven days” (Grade B, consensus 89%). Its sepsis commentary: keep EN as the first choice for the first three to four days; if not feasible or insufficient after three days, prescribe PN up to approximately half of the predicted or measured energy needs, and EN as soon as the condition permits. The Early PN trial (Doig 2013; 1,372 patients with short-term relative contraindications to EN): PN within 44 minutes of randomisation versus standard care (EN or PN after a mean of 2.8 days) — no difference in 60-day mortality or infection, 0.47 fewer days of ventilation, less muscle and fat loss. EPaNIC (4,640; PN supplementing EN): within 48 hours versus day 8 — the late group 6.3% more likely to leave the ICU alive early, infections 22.8% versus 26.2%, less cholestasis, shorter renal replacement, same mortality. ESPEN’s reading: EPaNIC’s harm may have come from full targets set by predictive equations in the acute phase. ASPEN 2022, questions 3 and 4: in the first week PN and EN give the same outcomes (high certainty, strong), and supplemental PN should not start before day 7 (high, strong) — conclusions that excluded malnourished patients.
• Amount over route. CALORIES (2,400 patients who could be fed either way): PN versus EN, 30-day mortality 33.1% versus 34.2%, infections the same, less hypoglycaemia and vomiting with PN. Most patients in both arms never reached target.

1. Calorie targets by phase — where the guidelines split

Evidence: four guidelines (grades as printed), the trials behind them (NUTRIREA-3, PermiT, EDEN, EPaNIC, TARGET), one observational cohort (Zusman 2016).

• My note’s staging (≤70% → 70–100% → 100%) is ESPEN’s. ESPEN’s reason for staying below: in the acute phase endogenous energy production runs at 500–1,400 kcal/day, so full feeding from the start is overfeeding (R17 commentary). In the calorimetry cohort (Zusman 2016, 1,171 patients) mortality fell as delivered/measured energy rose towards 70% and rose again above it — a non-linear curve — and above 70% stays and ventilation were longer.
• Reading the same trials, France went lower and Japan went the other way. SRLF lifted NUTRIREA-3’s 6 kcal/kg straight into a recommendation; JCCNG read the absence of a mortality benefit as “no reason to restrict deliberately”. Neither says that feeding less in the first week kills.
• Trophic starts are fine (my note): EDEN (1,000 patients with acute lung injury) — six days of trophic feeding (about 400 kcal/day, 10–20 mL/h) versus full (about 1,300) — same ventilator-free days, 60-day mortality and infections; the full group had more vomiting, high residuals, constipation, higher glucose and insulin. PermiT (894): 40–60% of non-protein calories versus 70–100%, protein equal — 90-day mortality 27.2% versus 28.9%, no difference.
• “From day 3–7 calorie deficit starts to matter” (my note) is observational (Zusman: under 50% was bad). In the trial, seven days of low intake did no harm (NUTRIREA-3). The recovery-phase “100% of REE, about 25 kcal/kg” is ESPEN’s 20–25 kcal/kg weight-based figure.

2. Measuring energy — what with

Evidence: ESPEN R15 (B) and statement 2 (consensus); JCCNG 2B; SRLF R4; one validation study (Stapel 2015, n = 84).

3. Formulas — density and type

Evidence: one randomised trial (TARGET); guideline wording (SCCM 2016 E4, JCCNG 2024, ESPEN 2023); water content per manufacturer.

• Density follows fluid (my note’s formula: required calories ÷ maximum deliverable volume = minimum density) — 1.0 kcal/mL is about 84% water, 1.5 about 76%, 2.0 about 70% (manufacturer figures, product-dependent). TARGET (3,957 ventilated patients): 1.5 versus 1.0 kcal/mL at the same rate (1 mL/kg/h) — 1,863 versus 1,262 kcal/day (+601), 90-day mortality 26.8% versus 25.7% (RR 1.05, 0.94–1.16), no difference in survival time, organ support or infection. Putting in more calories changed nothing; SRLF 2025 R9 goes a step further and advises against energy- or protein-enriched formulas.
• Type. My note’s algorithm (digestive capacity → protein fraction → fluid → special situations) is habit. Peptide formulas for digestive failure: SCCM 2016 E4b suggests small-peptide formulas for persistent diarrhoea, suspected malabsorption or failure of fibre (expert consensus); JCCNG 2024 suggests against oligomeric or elemental formulas as a nitrogen source (2D).
• Disease-specific formulas. My note’s “diabetes → low-carbohydrate formula”: JCCNG 2024 suggests against high-fat, low-carbohydrate EN (2C); ESPEN makes no recommendation. “Liver failure → BCAA-enriched”, “renal failure → electrolyte-adjusted”: no ICU outcome data (habit). “ARDS → omega-3”: ESPEN 2023 allows omega-3 EN within nutritional doses but not high-dose boluses; SCCM 2016 against routine use in ARDS; JCCNG 2024 suggests omega-3-enriched EN (2C) — they disagree.
• Fibre. SCCM 2016 E4b: consider a mixed-fibre formula for persistent diarrhoea; avoid both soluble and insoluble fibre in patients at high risk of bowel ischaemia or severe dysmotility (expert consensus). SRLF 2025 R20: fibre for diarrhoea, insufficient evidence for routine prevention. My note’s “contraindicated in haemodynamic instability” is a translation of “high risk of bowel ischaemia”.

4. Protein — where “more important than calories” came from, and where it is now

Evidence: four guidelines (divided); three randomised trials (EFFORT Protein, PRECISe, NUTRIREA-3) plus one Bayesian reanalysis; one observational cohort (PROTINVENT).

• The sentence comes from SCCM/ASPEN 2016, C4 — protein “appears to be the most important macronutrient”, 1.2–2.0 g/kg, more in burns and multitrauma. Its evidence was observational (achieving 1.3 g/kg associated with outcome). PermiT’s design — cut calories, keep protein — is the same era’s thinking.
• The 2023–24 trials point the other way. EFFORT Protein (1,301 high-nutritional-risk ventilated patients; ≥2.2 versus ≤1.2 g/kg started within 96 hours): alive discharge by day 60 46.1% versus 50.2% (hazard ratio 0.91, 0.77–1.07), 60-day mortality 34.6% versus 32.1% — a harmful direction in the acute kidney injury and high-SOFA subgroups. The 2024 Bayesian reanalysis: 72% probability of harm for mortality, 92% for delayed discharge, 97% and 95% probability of interaction with baseline creatinine and SOFA. PRECISe (935, double-blind; 2.0 versus 1.3 g/kg): EQ-5D health utility lower with high protein (−0.05, p = 0.031), mortality hazard ratio 1.14 (0.92–1.40), gastrointestinal intolerance odds ratio 1.76. NUTRIREA-3: 0.2–0.4 g/kg for the first week was no worse than 1.0–1.3 (section 0).
• How the guidelines took it. ESPEN 2023 R22 keeps “1.3 g/kg progressively” but its commentary reads: “there is no support to recommend higher doses than 1.3 g/kg/d … medium doses (not low and not high) are associated with the best outcomes”, and stable-isotope work shows amino-acid utilisation only improves after the early acute phase. SRLF 2025 R7: 0.2–0.9 g/kg in the first week, then 1.0–1.3. ASPEN 2022: 1.2–2.0 retained, “cannot make a new recommendation”. JCCNG 2024: above 1.2 suggested, very low certainty.
• Timing. My note’s “under 0.8 g/kg on days 1–2, then increase” comes from an observational study (PROTINVENT; 455 patients ventilated seven days or more): under 0.8 before day 3 and over 0.8 after was associated with the lowest 6-month mortality (adjusted hazard ratio 0.61); high protein in the first three to five days with higher long-term mortality; overall low protein worst of all. Same direction as ESPEN’s “progressively”. Not confirmed by a trial.
• “Permissive underfeeding — 80% of calories, 100% of protein” (my note): the 80% is an observational figure (Zusman’s 70–100%); the “100% protein” is 2016 thinking. PermiT went down to 40–60% of calories with protein equal in both arms (57 versus 59 g/day). Counting propofol lipid (1.1 kcal/mL) in the energy total is right — both ESPEN and ASPEN ask for non-nutritional calories to be included.

5. Refeeding — the numbers, with their sources

Evidence: two consensus documents (ASPEN 2020; NICE as summarised by Mehanna 2008), ESPEN 2023 recommendations 54–55, SRLF 2025 R6. No randomised trial — the definition itself was only agreed in 2020.

• High risk (NICE): BMI under 16, weight loss over 15% in three to six months, little or no intake for over five days, or low phosphate, potassium or magnesium — any one; or two of BMI under 18.5, 10% loss, ten days without intake, alcohol or insulin or diuretics. My note’s list (prolonged fasting, low weight, alcohol, electrolyte disturbance) is the same list.
• How much to start with. NICE (via Mehanna 2008): no more than 10 kcal/kg/day (5 if BMI under 14 or over two weeks without intake), increased over four to seven days; no more than 50% of requirements after five days without intake. ASPEN 2020 consensus: 100–150 g of dextrose or 10–20 kcal/kg for the first 24 hours, then advance by 33% of goal every one to two days; if electrolytes drop steeply, cut by 50% and resume the 33% steps. My note’s “50–75%” and “sodium under 1 mmol/kg/day” appear in neither document — I have left the 75% and the sodium limit as protocol-table figures without a source, and used the two sets above.
• In the ICU. ESPEN 2023, 54–55: with refeeding hypophosphataemia (under 0.65 mmol/L or a fall of over 0.16), measure electrolytes two to three times a day and replace; restrict energy for 48 hours, then increase gradually. SRLF 2025 R6: hypophosphataemia under 0.65 on standard feeding — reduce to 20 kcal/h (480 kcal/day) for two days, then increase. ESPEN: electrolyte shifts on refeeding are not confined to patients with overt malnutrition.
• Thiamine. ASPEN 2020: 100 mg before feeding or before dextrose-containing fluids, and 100 mg a day for five to seven days or longer in severe starvation, chronic alcohol use or signs of deficiency. NICE: 200–300 mg a day orally plus vitamin B compound and a daily multivitamin and trace elements. Potassium 2–4, phosphate 0.3–0.6, magnesium 0.2 (IV) or 0.4 (oral) mmol/kg/day (Mehanna’s table).

What I take from it

1. Feeding less in the first week does not kill anyone — the conclusion of a 3,044-patient trial in shock. No guideline claims otherwise.
2. In shock, EN is “low dose once controlled” — not “once the vasopressor is off”. No guideline gives a number; the one observational study saw benefit up to 0.3.
3. Full-dose early EN doubled bowel ischaemia — the harm was the amount, not the route.
4. PN within three to seven days, up to half of needs — ESPEN’s sepsis prescription. A full early PN target was EPaNIC’s harm.
5. Calorimetry → VCO₂ × 8.19 → 20–25 kcal/kg — but the French say not to use calorimetry routinely. Not filling the target matters more than measuring it.
6. 1.5 kcal/mL just puts in more calories — survival is the same (TARGET). Choose density for fluid.
7. “Protein matters more than calories” is a 2016 sentence — the 2023–24 trials found no benefit from high or early-high protein, and harm in kidney injury. 1.3 g/kg, reached progressively after the first week, is where the guidelines stand; France says 0.2–0.9 in week one.
8. Do not stop feeds for a residual under 500 mL — and 2016 and 2025 both say not to measure it routinely.
9. Refeeding: 10 kcal/kg or 10–20 kcal/kg on day one, thiamine 100 mg first — “50–75%” and “sodium under 1 mmol/kg” are numbers without a source.
10. Micronutrients daily with PN — nothing above that, in high-dose trials, has changed an outcome.

Where this leaves me

I know protein matters, but honestly I don’t think I paid much(?) attention to it. In surgical ICUs there is a lot of interest and they give it aggressively, but in medical ICUs (not just me..) we didn’t pay much attention to protein. From now on I should..

Sources

Guidelines

• Singer P, Blaser AR, Berger MM, et al. ESPEN practical and partially revised guideline: Clinical nutrition in the intensive care unit. Clin Nutr. 2023;42(9):1671-1689. doi:10.1016/j.clnu.2023.07.011 · PMID 37517372
• Singer P, Blaser AR, Berger MM, et al. ESPEN guideline on clinical nutrition in the intensive care unit. Clin Nutr. 2019;38(1):48-79. doi:10.1016/j.clnu.2018.08.037 · PMID 30348463
• Compher C, Bingham AL, McCall M, et al. Guidelines for the provision of nutrition support therapy in the adult critically ill patient: The American Society for Parenteral and Enteral Nutrition. JPEN J Parenter Enteral Nutr. 2022;46(1):12-41. doi:10.1002/jpen.2267 · PMID 34784064
• McClave SA, Taylor BE, Martindale RG, et al. Guidelines for the Provision and Assessment of Nutrition Support Therapy in the Adult Critically Ill Patient: SCCM and ASPEN. JPEN J Parenter Enteral Nutr. 2016;40(2):159-211. doi:10.1177/0148607115621863 · PMID 26773077
• Nakamura K, Yamamoto R, Higashibeppu N, et al. The Japanese Critical Care Nutrition Guideline 2024. J Intensive Care. 2025;13(1):18. doi:10.1186/s40560-025-00785-z · PMID 40119480
• Reignier J, Gaillard-Le Roux B, Dequin PF, et al. Expert consensus-based clinical practice guidelines for nutritional support in the intensive care unit: the French Intensive Care Society (SRLF) and GFRUP. Ann Intensive Care. 2025;15(1):99. doi:10.1186/s13613-025-01509-0 · PMID 40665004
• Berger MM, Shenkin A, Schweinlin A, et al. ESPEN micronutrient guideline. Clin Nutr. 2022;41(6):1357-1424. doi:10.1016/j.clnu.2022.02.015 · PMID 35365361

Trials in shock

• Reignier J, Boisramé-Helms J, Brisard L, et al. Enteral versus parenteral early nutrition in ventilated adults with shock: a randomised, controlled, multicentre, open-label, parallel-group study (NUTRIREA-2). Lancet. 2018;391(10116):133-143. doi:10.1016/S0140-6736(17)32146-3 · PMID 29128300
• Reignier J, Plantefeve G, Mira JP, et al. Low versus standard calorie and protein feeding in ventilated adults with shock: a randomised, controlled, multicentre, open-label, parallel-group trial (NUTRIREA-3). Lancet Respir Med. 2023;11(7):602-612. doi:10.1016/S2213-2600(23)00092-9 · PMID 36958363
• Schleef M, Mayet V, Le Gouge A, et al. Low versus standard calorie and protein feeding and renal dysfunction in ventilated adults with shock: a NUTRIREA-3 post hoc analysis. Intensive Care Med. 2026;52(4):651-661. doi:10.1007/s00134-026-08369-7 · PMID 41989569

Timing and route of parenteral nutrition

• Casaer MP, Mesotten D, Hermans G, et al. Early versus late parenteral nutrition in critically ill adults. N Engl J Med. 2011;365(6):506-517. doi:10.1056/NEJMoa1102662 · PMID 21714640
• Harvey SE, Parrott F, Harrison DA, et al. Trial of the route of early nutritional support in critically ill adults. N Engl J Med. 2014;371(18):1673-1684. doi:10.1056/NEJMoa1409860 · PMID 25271389
• Doig GS, Simpson F, Sweetman EA, et al. Early parenteral nutrition in critically ill patients with short-term relative contraindications to early enteral nutrition: a randomized controlled trial. JAMA. 2013;309(20):2130-2138. doi:10.1001/jama.2013.5124 · PMID 23689848

Calorie dose

• Chapman M, Peake SL, Bellomo R, et al.; TARGET Investigators, for the ANZICS Clinical Trials Group. Energy-Dense versus Routine Enteral Nutrition in the Critically Ill. N Engl J Med. 2018;379(19):1823-1834. doi:10.1056/NEJMoa1811687 · PMID 30346225
• Arabi YM, Aldawood AS, Haddad SH, et al. Permissive Underfeeding or Standard Enteral Feeding in Critically Ill Adults. N Engl J Med. 2015;372(25):2398-2408. doi:10.1056/NEJMoa1502826 · PMID 25992505
• Rice TW, Wheeler AP, Thompson BT, et al.; NHLBI ARDS Clinical Trials Network. Initial trophic vs full enteral feeding in patients with acute lung injury: the EDEN randomized trial. JAMA. 2012;307(8):795-803. doi:10.1001/jama.2012.137 · PMID 22307571
• Zusman O, Theilla M, Cohen J, et al. Resting energy expenditure, calorie and protein consumption in critically ill patients: a retrospective cohort study. Crit Care. 2016;20(1):367. doi:10.1186/s13054-016-1538-4 · PMID 27832823
• Stapel SN, de Grooth HJ, Alimohamad H, et al. Ventilator-derived carbon dioxide production to assess energy expenditure in critically ill patients: proof of concept. Crit Care. 2015;19:370. doi:10.1186/s13054-015-1087-2 · PMID 26494245

Protein dose

• Heyland DK, Patel J, Compher C, et al. The effect of higher protein dosing in critically ill patients with high nutritional risk (EFFORT Protein): an international, multicentre, pragmatic, registry-based randomised trial. Lancet. 2023;401(10376):568-576. doi:10.1016/S0140-6736(22)02469-2 · PMID 36708732
• Haines RW, Granholm A, Puthucheary Z, et al. The effect of high protein dosing in critically ill patients: an exploratory, secondary Bayesian analyses of the EFFORT Protein trial. Br J Anaesth. 2024;133(6):1192-1200. doi:10.1016/j.bja.2024.08.033 · PMID 39455305
• Bels JLM, Thiessen S, van Gassel RJJ, et al. Effect of high versus standard protein provision on functional recovery in people with critical illness (PRECISe): an investigator-initiated, double-blinded, multicentre, parallel-group, randomised controlled trial in Belgium and the Netherlands. Lancet. 2024;404(10453):659-669. doi:10.1016/S0140-6736(24)01304-7 · PMID 39153816
• Koekkoek WACK, van Setten CHC, Olthof LE, Kars JCNH, van Zanten ARH. Timing of PROTein INtake and clinical outcomes of adult critically ill patients on prolonged mechanical VENTilation: The PROTINVENT retrospective study. Clin Nutr. 2019;38(2):883-890. doi:10.1016/j.clnu.2018.02.012 · PMID 29486907

Vasopressors and early enteral nutrition

• Ohbe H, Jo T, Matsui H, Fushimi K, Yasunaga H. Differences in effect of early enteral nutrition on mortality among ventilated adults with shock requiring low-, medium-, and high-dose noradrenaline: A propensity-matched analysis. Clin Nutr. 2020;39(2):460-467. doi:10.1016/j.clnu.2019.02.020 · PMID 30808573

Refeeding

• da Silva JSV, Seres DS, Sabino K, et al. ASPEN Consensus Recommendations for Refeeding Syndrome. Nutr Clin Pract. 2020;35(2):178-195. doi:10.1002/ncp.10474 · PMID 32115791
• Mehanna HM, Moledina J, Travis J. Refeeding syndrome: what it is, and how to prevent and treat it. BMJ. 2008;336(7659):1495-1498. doi:10.1136/bmj.a301 · PMID 18583681