Parenteral nutrition and refeeding syndrome — branching viva

Opening — framing the problem

The examiner begins: a 15-year-old with anorexia nervosa is admitted for refeeding after six weeks of negligible intake and a 15 per cent weight loss, with a body mass index of 15. On day two the phosphate has fallen and she is oedematous. Talk me through your approach. [5]

I would frame this as a high-risk refeeding scenario. The combination of prolonged negligible intake, significant weight loss and a very low body mass index places her at the highest risk of refeeding syndrome. My priorities would be to confirm the risk, to have given thiamine before feeding, to have corrected the phosphate, potassium and magnesium, to start at a conservative calorie level and escalate slowly, and to check the electrolytes at least daily for the first week, because the phosphate fall and the oedema tell me the syndrome is already declaring itself. [5] [6]

Branch A — definition and biochemistry

Define refeeding syndrome for me and explain the mechanism. [5]

Refeeding syndrome is the metabolic and fluid disturbance that follows the reintroduction of carbohydrate to a starved body. In starvation the body runs on fat and ketones with low insulin, and the intracellular stores of phosphate, potassium and magnesium are depleted even though the serum levels may look normal. When carbohydrate reappears, insulin surges, driving phosphate, potassium and magnesium out of the serum and into the cells just as those cells begin to consume them in bulk. Thiamine, a cofactor for carbohydrate metabolism, is exhausted, and the serum phosphate falls within the first 72 hours. [5] [6]

Branch B — risk stratification

Why did you call her high risk, and what are the ASPEN criteria? [5]

The ASPEN consensus stratifies risk by intake, weight loss and body mass index. This girl has had negligible intake for well over five days, she has lost 15 per cent of her body weight, and her body mass index is 15, each of which marks the severe end of the risk spectrum. Any one of these would put a child at risk; the combination places her at the highest risk and demands the most conservative start and the closest monitoring. [5]

Branch C — prevention and the slow start

What should you have done before and during feeding, and what will you do now that the phosphate has fallen? [6]

Before feeding I would have given thiamine, corrected any phosphate, potassium or magnesium deficit, restricted the sodium and fluid, and started at a conservative intake around 10 to 20 kcal per kg per day, escalating slowly over several days. Now that the phosphate has fallen, I would correct it with intravenous replacement, slow or pause the calorie escalation, manage the fluid overload with restriction and diuresis if needed, and check the electrolytes more frequently than daily until the trend stabilises. The slow start is the whole point of the management. [5] [6]

Branch D — the oedema and the heart

Why is she oedematous with a rising heart rate, and what is the danger? [6]

The insulin surge retains sodium at the kidney and reactivates the renin-angiotensin-aldosterone system, and the starved heart with its wasted myocardium cannot handle the extra volume, so she has tipped toward acute cardiac failure. The oedema and the rising heart rate are the bedside warning, and the danger is arrhythmia from the electrolyte shifts and frank pump failure from the fluid load. The response is to restrict the fluid, correct the electrolytes, slow the calories, and monitor her closely with an electrocardiogram. [6]

Branch E — the premature infant and the liver

Now a premature infant has been on parenteral nutrition for four weeks with a soy-based lipid at 3 g/kg/day and two line infections, and the conjugated bilirubin is rising. What is this and what is your plan? [9]

This is intestinal failure-associated liver disease, and a rising conjugated bilirubin in an infant on parenteral nutrition is IFALD until proven otherwise. The two dominant modifiable drivers are the parenteral lipid load and recurrent sepsis. I would reduce or modify the lipid, restricting it toward around 1 g per kg per day and shifting to a mixed emulsion such as SMOFlipid or an omega-3 fish-oil emulsion, because the phytosterols in the soy-based emulsion drive the cholestasis. I would tighten the line care and treat any infection, advance trophic enteral feeding to restore bile flow, and monitor the conjugated bilirubin serially. I would also actively exclude biliary atresia, because a cholestatic infant needs that work-up in parallel. [8] [9]

Closing — the prescription and the principle

Name the parenteral streams and give the infant targets. [1]

The four streams are glucose as a glucose infusion rate, amino acids, lipid, and electrolytes with vitamins and trace elements. The glucose infusion rate should not exceed the oxidative capacity of about 11 to 12 mg per kg per min in a term infant. Amino acids advance to around 3 to 3.5 g per kg per day. Lipid advances to around 3 g per kg per day but is restricted to protect the liver. The energy target is around 90 to 120 kcal per kg per day, lower than the enteral target because every calorie is absorbed. The governing principle is to feed the gut first, use the vein only when the gut cannot, and feed the starved child slowly. [1] [3]