The NICU Survival Guide

What it is

This is not a guideline, a review or a trial. It is an orientation manual written by junior doctors for the junior doctors coming after them, and that is precisely what makes it interesting. Guidelines tell you what a specialty has agreed. A survival guide tells you what a unit actually worries about at three in the morning — which order set to open, which consultant does not want a physical examination in the note, who to telephone about an echo. It runs through respiratory, fluids and nutrition, haematology, cardiovascular, neurology, infection, gastroenterology, renal and endocrine, then appends the admission orders, a guideline for the death of a baby, and two full policies on pain and sedation and on palliative care. Reading it is the closest most of us will get to standing in someone else's unit for a fortnight.

What transfers

A surprising amount, once you strip the local scaffolding away. The respiratory distress tree on page nine — upper airway, pulmonary, cardiac, thoracic, other — is a better differential than most of us carry in our heads. The PPHN section gives you the oxygenation index and the rules of 20s and 60s in four lines. The HIE criteria set out the six-hour cooling window explicitly, with the sentinel event, the Apgar, the pH and the base excess all quantified. Small things recur usefully: that a raised potassium on a capillary sample may simply be haemolysis, that urine output belongs in mL/kg/h, that hyperthermia in an incubator may be withdrawal or intracranial haemorrhage rather than a warm room. And the pain and sedation policy is genuinely sophisticated — ten painful procedures a day as the baseline, the finding that more than seven days of combined opioid and benzodiazepine exposure in extremely preterm infants tracks with worse two-year cognitive, motor and language scores, and a reasoned preference for dexmedetomidine over both.

What to read sceptically

This is where a journal club earns its keep, because the document's authority is local and its reach is not. The units are American throughout — glucose in mg/dL, weights in pounds on the order screens — and a misread threshold is a real hazard. Drugs and products differ: Survanta and Curosurf, nirsevimab under a US schedule, the Kaiser sepsis calculator, an ampicillin-and-gentamicin default that may not match your local antibiogram. Some advice would not survive contact with an Australian neonatal unit at all: rice cereal and famotidine for reflux sit oddly against current evidence, and the instruction to open the General Adult Admission order set for a newborn is a lesson in what electronic records do to us rather than a clinical recommendation. Much of it is uncited. None of that makes the document bad — it makes it a local artefact, which is what its own disclaimer says on every page. The question worth arguing about is whether the parts that are excellent, particularly the guideline for the death of a baby, are excellent because they were written locally rather than in spite of it.