GI Bleeding, Pancreatitis, Hepatic Failure and Nutrition

[9]

Overview & definition

The GI and the nutritional disorders are the common ICU complications — from the stress ulcer prophylaxis to the acute variceal bleed, from the mild pancreatitis to the severe necrotising, from the early enteral nutrition to the acute liver failure with the encephalopathy. Each has a specific, the evidence-based management, and the intensivist's role is the recognition, the resuscitation, and the targeted therapy.[1][2]

The upper GI bleeding

The upper GI bleeding presents with the haematemesis, the melaena, or the coffee-ground aspirate. The two commonest causes are the peptic ulcer disease (the Helicobacter, the NSAID) and the oesophageal varices (the portal hypertension from the cirrhosis).[1]

The resuscitation. The ABCDE — the airway (the aspiration risk in the massive haematemesis — the early intubation), the breathing (the oxygenation), the circulation (the two large-bore cannulae, the fluid, the blood — the damage-control principles, the restrictive transfusion at the Hb of 70, the correction of the coagulopathy and the platelets).[2]

The peptic ulcer bleed. The proton pump inhibitor infusion (the pantoprazole 80 mg bolus then 8 mg/h), the upper GI endoscopy within 24 hours (the adrenaline injection, the clipping, the thermal coagulation), and the Helicobacter eradication. The risk stratification (the Glasgow-Blatchford score for the admission; the Rockall score for the post-endoscopy risk).[1]

The variceal bleed. The vasoactive drug (the terlipressin or the octreotide — the splanchnic vasoconstriction), the prophylactic antibiotic (the ceftriaxone — the cirrhotic patient is immunocompromised and the infection worsens the outcome), the band ligation (the endoscopic), and the TIPSS (the transjugular intrahepatic portosystemic shunt) for the refractory. The balloon tamponade (the Sengstaken-Blakemore) is the bridge to the definitive therapy.[1][2]

The acute pancreatitis

The acute pancreatitis is the inflammation of the pancreas from the gallstone, the alcohol, the ERCP or the other cause. The severity ranges from the mild (the majority, self-limiting) to the severe (the necrotising, the infected, the multi-organ failure).[1]

The management. The aggressive fluid resuscitation (the lactated Ringer's preferred — the goal-directed), the analgesia (the opiate), the early enteral nutrition (the modern standard — NOT the prolonged fasting; the nasogastric or the nasojejunal if the gastric feeding not tolerated), the ERCP for the gallstone with the cholangitis or the obstruction, and the antibiotics ONLY for the proven or the suspected infected necrosis (NOT the prophylactic — the previous practice of the prophylactic carbapenem is disproven). The necrosectomy (the minimally invasive, the delayed — the 4-week rule for the walling-off) for the infected necrosis.[1][2]

The acute liver failure

The acute liver failure is the severe, the acute hepatocellular injury with the coagulopathy and the encephalopathy — the commonest cause in the developed world is the paracetamol (the acetaminophen) overdose, followed by the drug-induced, the viral and the indeterminate.[1]

The management. The N-acetylcysteine (the NAC — for the paracetamol overdose AND, increasingly, for the non-paracetamol ALF — the antioxidant, the glutathione replenishment), the encephalopathy management (the lactulose, the rifaximin — the ammonia reduction; the intubation for the grade 3 to 4 encephalopathy; the raised ICP management — the mannitol, the hypertonic saline), the coagulopathy (the vitamin K, the fresh frozen plasma only for the bleeding or the procedure — NOT routinely, for the INR is a prognostic marker not a treatment target), and the liver transplant for the poor prognosis (the King's College criteria).[1][2]

The King's College criteria for the paracetamol ALF transplant: the arterial lactate above 3.5 after the early resuscitation, or above 3.0 after the fluid resuscitation; OR the arterial pH below 7.25; OR all three of the INR above 6.5, the creatinine above 300, and the grade 3 to 4 encephalopathy.[1]

The stress ulcer prophylaxis

The stress ulcer bleeding was historically common in the ventilated ICU patient. The modern practice is the selective — the acid suppression (the PPI or the H2 blocker) for the high-risk patient: the mechanical ventilation over 48 hours, the coagulopathy, the shock, the sepsis, the major burns, the traumatic brain injury, the history of the GI bleeding. The routine prophylaxis for the low-risk patient is NOT recommended (the increased pneumonia risk from the gastric alkalinisation).[1][2]

The nutritional support

The early enteral nutrition (within 24 to 48 hours) is the standard for the ICU patient — it maintains the gut mucosal integrity, reduces the bacterial translocation, modulates the immune response, and improves the outcome. The parenteral nutrition is reserved for the patient who cannot tolerate the enteral (the prolonged ileus, the high-output fistula, the mesenteric ischaemia) or the malnourished patient whose enteral nutrition is inadequate after 5 to 7 days.[1][2]

The refeeding syndrome. The malnourished patient started on the nutrition is at risk of the refeeding syndrome — the hypophosphataemia, the hypokalaemia, the hypomagnesaemia, the fluid retention, the cardiac arrhythmia. The prevention is the slow escalation (the 10 to 15 kcal/kg/day for the first week in the high-risk), the electrolyte replacement, the thiamine, and the monitoring.[1][2]

Enteral versus parenteral nutrition: the route question

The enteral nutrition (EN) is preferred over the parenteral nutrition (PN) in nearly all critically ill patients who have a functioning gut. The biological rationale is robust: enteral feeding maintains the gut mucosal integrity, supports the gut-associated lymphoid tissue (GALT), preserves the intestinal brush border enzyme activity, attenuates the stress-induced immune response, and reduces the bacterial translocation that drives the nosocomial infection and the multi-organ failure. The enteral route is cheaper, safer (no central line, no catheter-related bloodstream infection, no metabolic derangement), and is associated with fewer infectious complications in the meta-analytic data.[9][10]

The CALORIES trial (Harvey 2014, NEJM) is the landmark randomised study on the route. In 2400 mechanically ventilated ICU patients, early EN was compared with early PN within 36 hours of admission and continued for up to 5 days. There was no difference in the 30-day mortality between the two groups (EN 39% vs PN 40%). However, the EN group had fewer episodes of hypoglycaemia and a lower mean glycaemic load. The trial settled the question: when the gut works, the EN and the PN give equivalent mortality — but the EN is cheaper, safer, and physiologically superior, so it remains the default.[4]

[9] [10]

Timing of nutrition: the early EN question

The early enteral nutrition (within 24 to 48 hours of the ICU admission) is the modern standard for the critically ill patient who is no longer in active unresuscitated shock. The evidence base is now extensive: the early EN reduces the infection, the ICU-acquired weakness, the length of stay, and possibly the mortality (the effect is modest but consistent across the meta-analyses).[9][10]

The NUTRIREA-2 trial (Reignier 2018, Lancet) is the pivotal study on the timing AND the route in shock. It randomised 2410 mechanically ventilated adults with shock (all severities, all causes) to the early EN versus the early PN, both started within 24 hours and isocaloric. The result: there was no difference in the 28-day mortality (EN 37% vs PN 35%) or the infection rate. Crucially, the early PN did NOT improve the outcome — and the EN group had a SIGNIFICANTLY HIGHER rate of the digestive ischemia (1.0% vs 0.0%, p=0.03) and the bowel ischaemia requiring surgery. The takeaway: in the ventilated patient with the shock, the early EN is the standard, but the clinician must be vigilant for the mesenteric ischaemia — and the early PN confers NO benefit.[3]

[3]

Caloric and protein targets in the ICU

The caloric and the protein targets are the two pillars of the ICU nutrition prescription. The cardinal error is the overfeeding — which causes the hyperglycaemia, the hypercapnia (raising the minute ventilation and the work of breathing), the hepatic steatosis, the azotaemia, and the increased infection — and the underfeeding — which causes the loss of the lean body mass, the ICU-acquired weakness, the impaired wound healing, and the prolonged ventilation.[10][11]

The caloric target. The gold standard is the indirect calorimetry (the measurement of the oxygen consumption and the carbon dioxide production to derive the resting energy expenditure, REE). Where it is unavailable, the weight-based predictive equation of 25 to 30 kcal/kg/day is used (the ESPEN and the SCCM/ASPEN guidelines). The older Harris-Benedict equation overestimates the energy expenditure in the critically ill and is NOT recommended. The permissive underfeeding (the 70 to 80% of the target for the first week in the well-nourished patient, then advancing to the full target) is supported by the contemporary trials (see below).[10][11]

The protein target. The critically ill patient is in the catabolic state — the nitrogen losses of 20 to 30 g/day (the equivalent of 0.7 to 1 kg of the lean mass per week) are common. The protein target is 1.2 to 2.0 g/kg/day, with the higher end (1.5 to 2.0 g/kg/day) for the patient with the AKI on the CRRT (the CRRT loses the amino acids), the burns, the trauma, and the prolonged ventilation. The protein is delivered as the whey-based or the casein-based enteral formula, or as the amino-acid solution in the PN.[10][11]

[9] [10] [11]

Permissive underfeeding: the evidence

The permissive underfeeding (the delivery of approximately 50 to 70% of the caloric target, with the full protein delivery, for the first week of the ICU stay) is the contemporary practice in the well-nourished patient. The rationale: the early full feeding in the catabolic patient does not reverse the catabolism (the catabolism is hormonally driven, not substrate-limited) and may worsen the autophagy (the cell-clearance of the damaged mitochondria and the proteins, which is essential for the recovery). The underfeeding respects the autophagy, avoids the overfeeding harm, and delivers the full protein to mitigate the lean mass loss.[5][12]

[5] [12]

Practical aspects of enteral feeding

The practical delivery of the enteral nutrition — the route of the tube, the prokinetic, the gastric residual volume (GRV), the position, the monitoring — is the everyday work of the ICU team. The cardinal principle: start the EN early, advance gradually, monitor the tolerance, and use the prokinetics and the post-pyloric route when needed.[9][10]

The tube position — the gastric vs the post-pyloric. The gastric feeding (the nasogastric tube, the NG) is the default — it is easy, the tube is placed at the bedside, and the vast majority tolerate it. The post-pyloric feeding (the nasojejunal tube, the NJ, or the nasoduodenal) is reserved for the patient with the documented gastric intolerance (the high GRV, the vomiting) despite the prokinetic, the severe pancreatitis (the historical, now debated — the NG often works), the severe gastroparesis, the recurrent aspiration. The NJ does NOT reduce the ventilator-associated pneumonia in the meta-analysis — it is not the first choice.[9]

The prokinetics. The metoclopramide (10 mg IV every 8 hours) and the erythromycin (200 mg IV every 12 hours, or 250 mg NG every 8 hours) are the prokinetic agents. The combination (the metoclopramide + the erythromycin) is more effective than either alone for the high GRV. The erythromycin carries the QT-prolongation and the cardiac arrhythmia risk — monitor the ECG, avoid in the long-QT and the concurrent macrolide. The prokinetics are the first-line for the GRV intolerance before the NJ.[9][10]

The gastric residual volume (GRV). The routine GRV monitoring is NO LONGER recommended. The NUTRIREA-2 trial (and the earlier REGANE and the meta-analyses) showed that abandoning the routine GRV monitoring did NOT increase the pneumonia or the outcome harm — and it increased the caloric delivery (fewer interruptions). The current practice: the clinical assessment of the tolerance (the vomiting, the distension, the abdominal examination), the head of bed elevated 30 to 45°, the prokinetic if symptomatic. The GRV is measured only if the patient is symptomatic (the distension, the vomiting, the high aspirate on the suction).[3][9]

[3] [9] [10]

Refeeding syndrome: prevention and management

The refeeding syndrome is the potentially fatal metabolic disturbance that occurs when the chronically malnourished or starved patient is fed (the enteral or the parenteral) too rapidly. The mechanism: the refeeding drives the insulin surge, the insulin drives the intracellular shift of the phosphate, the potassium and the magnesium (depleting the serum levels), the glucose becomes the fuel (depleting the thiamine-dependent pathways), and the fluid retention causes the heart failure and the pulmonary oedema. The result: the hypophosphataemia (the weakness, the respiratory failure, the rhabdomyolysis, the haemolysis), the hypokalaemia (the arrhythmia), the hypomagnesaemia (the arrhythmia, the seizures), the thiamine deficiency (the Wernicke, the heart failure), and the fluid overload (the cardiac and the respiratory compromise).[1][10]

The high-risk patient. Anyone with the BMI below 16, the unintentional weight loss over 15% in the 3 to 6 months, the little or no nutritional intake for over 10 days, the low baseline phosphate/potassium/magnesium, the history of the alcohol misuse or the anorexia. These patients need the slow start (10 to 15 kcal/kg/day), the aggressive electrolyte replacement, the thiamine before the feeding, and the daily monitoring.[10]

[1] [10] [11]

Stress ulcer prophylaxis: the SUP-ICU evidence

The SUP-ICU trial (Krag 2018, NEJM) is the contemporary landmark on the stress ulcer prophylaxis. In 3298 ICU patients at risk of the GI bleeding (the mechanical ventilation over 48 hours, the INR over 1.5, the platelet below 50, the renal replacement therapy), the pantoprazole 40 mg IV daily was compared with the placebo. The result: the clinically important GI bleeding was significantly reduced (2.5% pantoprazole vs 4.2% placebo; RR 0.58, 95% CI 0.40 to 0.82), but there was no difference in the 90-day mortality (31% vs 31%), the infection, or the ICU-acquired weakness. The selective prophylaxis in the high-risk patient is confirmed — but the routine prophylaxis in the low-risk patient is NOT recommended (the increased pneumonia from the gastric alkalinisation, the C. difficile risk).[7]

[7]

Management: the integrated approach

1. The GI bleed — the resuscitation, the endoscopy (the PPI, the band ligation, the terlipressin/antibiotics for the varices), the risk stratification.[1]
2. The pancreatitis — the fluid (the lactated Ringer's), the early enteral, the ERCP for the gallstone, the antibiotics only for the infected necrosis.[1]
3. The ALF — the NAC, the encephalopathy/ICP, the coagulopathy (selective), the transplant (the King's College).[1]
4. The stress ulcer prophylaxis — the selective (the ventilated, the coagulopathic, the shocked).[2]
5. The nutrition — the early enteral (within 24 to 48 hours); the parenteral for the intolerant; the refeeding prevention.[1]

Monitoring

• The GI bleed — the haemoglobin, the haemodynamics, the endoscopy findings, the rebleeding.
• The pancreatitis — the CRP (the severity marker at 48 hours — above 150 suggests the severe), the CT severity index (the Balthazar), the organ failure (the Marshall score), the infection markers.
• The ALF — the INR (the prognostic), the ammonia (the encephalopathy), the lactate, the arterial pH, the King's College criteria.
• The nutrition — the electrolytes (the K, the Mg, the phosphate — the refeeding), the gastric residual volumes (the tolerance), the weight, the albumin (the trend, not the single value).[3]

Prognosis

The peptic ulcer bleeding mortality is 5 to 10 per cent. The variceal bleeding mortality is 15 to 20 per cent per the episode (higher with the rebleeding). The severe pancreatitis mortality is 10 to 30 per cent (higher with the infected necrosis and the multi-organ failure). The ALF mortality without the transplant is 50 to 80 per cent (with the transplant, 20 to 30 per cent). The nutritional status is the independent predictor of the ICU outcome.[1][2]

Short-answer questions

[3] [6]

Red flags

Clinical pearls: the high-yield exam points

Red flags: additional critical points

Compare: the EN vs PN summary at a glance

[3] [4] [10]

References

1. [1]Karnad DR, Nor MBM, Richards GA, et al. Intensive care in severe malaria: Report from the task force on tropical diseases by the World Federation of Societies of Intensive and Critical Care Medicine. Journal of critical care, 2018.PMID 29132978
2. [2]Kazma JM, van den Anker J, Allegaert K, et al. Anatomical and physiological alterations of pregnancy. Journal of pharmacokinetics and pharmacodynamics, 2020.PMID 32026239
3. [3]Reignier J, Boisramé-Helms J, Brisard L, et al. (NUTRIREA-2) Enteral versus parenteral early nutrition in ventilated adults with shock: a randomised, controlled, multicentre, open-label, parallel-group study (NUTRIREA-2). Lancet, 2018.PMID 29128300
4. [4]Harvey SE, Parrott F, Harrison DA, et al. (CALORIES) Trial of the route of early nutritional support in critically ill adults. New England Journal of Medicine, 2014.PMID 25271389
5. [5]Arabi YM, Aldawood AS, Haddad SH, et al. (Permissive Underfeeding) Permissive underfeeding or standard enteral feeding in critically ill adults. New England Journal of Medicine, 2015.PMID 25992505
6. [6]Hego A, Boulet M, Lassalle G, et al. (NUTRIREA-3) Impact of early low-calorie low-protein versus standard-calorie standard-protein feeding on outcomes of ventilated adults with shock: design and conduct of a randomised, controlled, multicentre, open-label, parallel-group trial (NUTRIREA-3). BMJ Open, 2021.PMID 33980526
7. [7]Krag M, Marker S, Perner A, et al. (SUP-ICU) Pantoprazole in patients at risk for gastrointestinal bleeding in the ICU. New England Journal of Medicine, 2018.PMID 30354950
8. [8]Rice TW, Mogan S, Hays MA, et al. (EDEN trial follow-up) One year outcomes in patients with acute lung injury randomised to initial trophic or full enteral feeding: prospective follow-up of EDEN randomised trial. BMJ, 2013.PMID 23512759
9. [9]Reintam Blaser A, Starkopf J, Alhazzani W, et al. Early enteral nutrition in critically ill patients: ESICM clinical practice guidelines. Intensive care medicine, 2017.PMID 28168570
10. [10]Singer P, Blaser AR, Berger MM, et al. ESPEN guideline on clinical nutrition in the intensive care unit. Clinical Nutrition, 2019.PMID 30348463
11. [11]McClave SA, Taylor BE, Martindale RG, et al. (SCCM/ASPEN) Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.). JPEN Journal of Parenteral and Enteral Nutrition, 2016.PMID 26773077
12. [12]Zhao S, Sun Y, Wang H, et al. Efficacy of permissive underfeeding for critically ill patients: an updated systematic review and trial sequential meta-analysis. Journal of Intensive Care, 2024.PMID 38254228