The neonate is not a small adult —
and not even a small child.
Between 24 weeks of gestation and 2 years of age the body changes its composition, its liver enzymes and its kidney function more than at any other time of life. Every dosing rule in neonatal practice is a consequence of that. This guide makes the consequence visible: move the baby, and watch the dose, the interval and the concentration curve move with it.
The maturation engine
Six curves generate every neonatal dosing rule you will ever meet. Water and extracellular fluid fall; fat, glomerular filtration, albumin and conjugating enzymes rise — each on its own timetable. Set a baby below and read what follows.
Set the infant
PMA = gestational age + postnatal age — the primary determinant of clearanceWhat that infant should receive
NeoFax interval chart| Drug | Preterm neonate | Term neonate | Adult | What follows |
|---|---|---|---|---|
| Gentamicin | 8–12 h | 4–6 h | 2–3 h | Interval 24–48 h, indexed to PMA |
| Ampicillin / penicillin G | 3–5 h | 2–3 h | ≈1 h | q12h in week 1, q8h thereafter |
| Vancomycin | 6–10 h | 4–6 h | 4–6 h | Interval 6–18 h by PMA; troughs mandatory |
| Meropenem | 3.82 h | 1.58 h | ≈1 h | q12h if <32 wk and <14 d; else q8h |
| Caffeine | 50–100 h | 40–70 h | ≈5 h | Once daily; CYP1A2 is last to mature |
| Fluconazole | 30–90 h | 25–30 h | ≈25 h | q48h for 2 weeks, then q24h |
| Metronidazole | 20.5 h | 16.7 h | ≈8 h | The interval chart stretches to q24h at 24–25 wk |
Only two parameters are real
Loading dose comes from volume of distribution. Maintenance rate comes from clearance. Half-life is derived from those two, never the other way round — which is why a long half-life can mean either wide distribution or slow elimination, with different clinical consequences.
Loading dose = Vd × target
How big the dose is. Neonatal Vd of a water-soluble drug is larger, so the mg/kg dose is larger. Gentamicin Vd ≈ 0.5 L/kg against 0.25 in an adult — that alone is why a neonate gets 4–5 mg/kg.
Maintenance rate = CL × target
How much per unit time. Neonatal clearance is smaller, so the interval is longer. Only the first dose comes from Vd; at steady state the regimen is governed entirely by clearance.
t½ = 0.693 × Vd ÷ CL
How far apart doses can be, and how long steady state takes (4–5 half-lives). In the neonate Vd is up and CL is down — both push half-life the same way, which is why neonatal half-lives are long.
Concentration over one steady-state interval
One-compartment, IV infusion, steady state — driven by the infant set in §01| Index | Killing | Drugs | Dosing consequence |
|---|---|---|---|
| Cmax/MIC | Concentration-dependent; target ratio 8–10 | Aminoglycosides, daptomycin, colistin, metronidazole, amphotericin B | One large dose at a long interval. Maximises the peak and allows a low-concentration window that limits toxicity. |
| T > MIC | Saturates above ≈4× MIC | All β-lactams (penicillins ≈50 %, cephalosporins 60–70 %, carbapenems ≈40 %), clindamycin, linezolid, flucytosine | Divide the same daily dose more often, or extend the infusion to 3–4 h. Raising the peak achieves nothing. |
| AUC₀₋₂₄/MIC | Total exposure | Vancomycin (≥400), fluoroquinolones, azithromycin, tetracyclines, fluconazole, echinocandins | Only the daily exposure matters; the schedule is then chosen for convenience and safety. |
Why every neonatal chart is a table of intervals
Larger Vd raises the mg/kg dose; smaller clearance lengthens the gap. That is the entire structure of the standard NeoFax β-lactam chart — and the live cell below is the one your infant from §01 falls into.
| Postmenstrual age | Postnatal age | Interval | Applies to |
|---|---|---|---|
| ≤ 29 wk | 0–28 d | q12h | Ampicillin · penicillin G · nafcillin · oxacillin · cloxacillin · cefazolin · ceftazidime · aztreonam · clindamycin · piperacillin–tazobactam |
| ≤ 29 wk | > 28 d | q8h | |
| 30–36 wk | 0–14 d | q12h | |
| 30–36 wk | > 14 d | q8h | |
| 37–44 wk | 0–7 d | q12h | |
| 37–44 wk | > 7 d | q8h | |
| ≥ 45 wk | all | q6h |
The sepsis cascade, one stage at a time
This is the section that separates a good answer from an average one. Work through it in seven steps: how the organism arrives, why the host cannot stop it, what the alarm does, what happens to the blood-vessel lining, how organs fail, and what follows the storm.
Infection → SIRS → MODS → immunoparalysis
Step through, or press PlayClotting, from scratch — and why the newborn decompensates so fast
Blood must stay liquid inside vessels and turn solid instantly at a cut. There is an accelerator and a brake, permanently balanced. Thrombin is the central character: everything either makes it or stops it — and thrombomodulin is the switch that decides which personality thrombin has.
| Test | What it measures | In DIC |
|---|---|---|
| Platelets | Consumed in the microthrombi | Low — often the earliest and most persistent sign |
| PT | Extrinsic + common pathway | Prolonged |
| aPTT | Intrinsic + common pathway | Prolonged |
| Fibrinogen | Raw material for fibrin | Low — used up |
| D-dimer / FDP | Fragments of broken-down clot | High — proof clots formed and partly dissolved |
The mechanism map
An antimicrobial works because it attacks something the microbe has and we do not. The size of that difference predicts the drug's toxicity — which is the whole hierarchy of antimicrobial safety in one idea. Click a target.
Where each class binds
Eight targets · click to open| Class | Target | Cidal/static | PK/PD | CSF | Main neonatal toxicity |
|---|---|---|---|---|---|
| Penicillins | transpeptidase (PBP) | cidal | T > MIC | only if inflamed | hypersensitivity; seizures at high dose; Na⁺/K⁺ load |
| Cephalosporins | transpeptidase (PBP) | cidal | T > MIC | 3rd & 4th gen, good | ceftriaxone → kernicterus + calcium precipitate |
| Carbapenems | transpeptidase (PBP) | cidal | T > MIC | meropenem, good | seizures (imipenem); resistance selection |
| Glycopeptides | D-Ala–D-Ala substrate | cidal | AUC/MIC | poor | nephrotoxicity; red man syndrome |
| Aminoglycosides | 30S ribosome | cidal | Cmax/MIC | inadequate | nephrotoxicity (reversible), ototoxicity (irreversible) |
| Macrolides | 50S ribosome | static | AUC/MIC | poor | pyloric stenosis <6 wk; CYP3A4 inhibition |
| Clindamycin | 50S ribosome | static | T > MIC | poor | C. difficile colitis |
| Linezolid | 23S rRNA, 70S initiation | static | AUC/MIC | good | thrombocytopenia; neuropathy after 28 d |
| Chloramphenicol | 50S peptidyl transferase | static | T > MIC | excellent | grey baby syndrome; aplastic anaemia |
| Co-trimoxazole | folate, two steps | cidal (combo) | AUC/MIC | good | kernicterus; haemolysis in G6PD deficiency |
| Metronidazole | DNA strand breaks | cidal | Cmax/MIC | excellent | neuropathy with prolonged use |
| Rifampicin | RNA polymerase | cidal | AUC/MIC | excellent | hepatotoxicity; enzyme induction; orange secretions |
| Polymyxins | lipid A | cidal | Cmax/MIC | negligible | nephrotoxicity and neurotoxicity |
Why fungi are harder: exactly four differences
Fungi are eukaryotes — nucleus, mitochondria, 80S ribosomes, exactly as we have. Almost every antibacterial target simply does not exist. There are precisely four exploitable differences, and they define the four antifungal classes. That single fact explains why antifungals are systematically more toxic.
β(1,3)-D-glucan + chitin
Not peptidoglycan — and mammalian cells have no wall at all. Echinocandins (micafungin, caspofungin) inhibit glucan synthase. The target has no human counterpart, so these are the best-tolerated antifungals — but they penetrate urine, CSF and vitreous poorly, which is the key neonatal limitation.
Ergosterol, not cholesterol
Polyenes bind it and punch pores (amphotericin B — fungicidal, broadest spectrum, and CSF penetration of 40–90 % of serum in the preterm neonate). Azoles stop it being made. Binding is not perfectly selective for ergosterol over cholesterol — hence nephrotoxicity.
Cytosine permease + deaminase
Flucytosine is selectively taken in and converted to 5-fluorouracil. Human cells lack cytosine deaminase — but gut bacteria have it, which is how systemic 5-FU is generated and the marrow is suppressed. Never used alone; resistance is rapid.
Lanosterol 14-α-demethylase
Azoles inhibit it — though not with perfect selectivity, which is exactly why they also inhibit human CYP3A4 and cause drug interactions. Fluconazole is the least potent inhibitor of the class.
Contraindicated or restricted — and the mechanism for each
Marks are earned for the reason, not the list. Each of these is a specific consequence of neonatal physiology from §01.
Ceftriaxone
Highly protein-bound — displaces bilirubin from albumin → kernicterus. And it forms an insoluble precipitate with calcium, which has caused fatal pulmonary and renal precipitates. Use cefotaxime. One exception: a single dose for gonococcal ophthalmia in a non-jaundiced infant not receiving calcium.
Sulphonamides / co-trimoxazole
The historical, best-documented cause of drug-induced kernicterus. Also haemolysis in G6PD deficiency and crystalluria from poorly soluble acetylated metabolites. Exceptions: Pneumocystis, toxoplasmosis, refractory Listeria.
Tetracyclines
Chelate calcium and deposit in calcifying tissue → permanent yellow-brown tooth staining, enamel hypoplasia, depressed bone growth.
Chloramphenicol
Grey baby syndrome. Inactivated by glucuronidation, then the conjugate is excreted renally — both steps immature. It accumulates, inhibits mitochondrial protein synthesis, and causes vomiting, abdominal distension, ashen-grey cyanosis, hypotension, collapse and death. Plus idiosyncratic aplastic anaemia.
Fluoroquinolones
Cartilage erosion in juvenile animals; tendinopathy; QT prolongation. Human neonatal data have not confirmed clinically significant arthropathy — so this is a considered exception for multi-drug-resistant infection, not an absolute bar.
Oral erythromycin
Motilin-receptor agonism → infantile hypertrophic pyloric stenosis, with a ~30-fold increased risk at 0–13 days of age. Use azithromycin for pertussis in this age group for exactly this reason.
Benzyl alcohol-preserved preparations
Metabolised to benzoic acid, which the neonate cannot conjugate to hippurate → gasping syndrome: metabolic acidosis, gasping respiration, encephalopathy, seizures, intraventricular haemorrhage, collapse, death. Benzoate also displaces bilirubin. Use preservative-free.
Propylene glycol vehicles
In IV lorazepam, phenytoin, phenobarbitone, some digoxin and multivitamin preparations: hyperosmolality, raised anion-gap lactic acidosis, seizures, arrhythmia, haemolysis, nephrotoxicity.
Aminoglycosides
Nephrotoxicity is reversible; ototoxicity is not. A mitochondrial 12S rRNA A1555G mutation causes profound, dose-independent deafness after a single dose — ask about a maternal family history of deafness after antibiotics.
Therapeutic drug monitoring
TDM is justified when four conditions hold together: a narrow therapeutic index, wide unpredictable variability, a measurable concentration that tracks effect or toxicity, and no easy clinical end-point. The neonate satisfies the first two for almost every drug — which is why TDM is used far more here than in adult practice.
| Drug | Peak | Trough | Why |
|---|---|---|---|
| Gentamicin / tobramycin | 5–12 mg/L (Cmax/MIC > 8:1) | 0.5–1 mg/L | Peak drives efficacy; trough drives nephro- and ototoxicity. Measure if therapy exceeds 48 h. |
| Amikacin | 20–35 mg/L | < 3–8 mg/L | As above. Avoid prolonged peaks above 35. |
| Vancomycin | — | 10–15 mg/L; 15–20 for meningitis, osteomyelitis, endocarditis | Trough, not peak, tracks both efficacy and AKI. 15 mg/kg hit a 10–20 trough in only ~74 % of infants; 10 mg/kg in 14 %. |
| Flucytosine | 50–80 µg/mL | — | Marrow suppression above 100 µg/mL. Amphotericin-induced renal impairment raises levels — the price of the synergy. |
| Chloramphenicol | 15–25 mg/L | 5–15 mg/L | Grey baby syndrome; dose-related marrow suppression. |
| Phenobarbitone | — | 15–40 mg/L | Long half-life, wide variability. |
| Phenytoin | — | total 10–20 mg/L, free 1–2 mg/L | Protein-binding displacement makes total levels unreliable; kinetics are saturable. |
| Caffeine citrate | — | 5–25 mg/L | Very long half-life, wide index — levels rarely needed. |
Self-test
Twenty questions drawn from the notes' own high-yield one-liners. Answer, then read why — recall beats re-reading.
Question 1 of 20
Where each claim on this page comes from
This guide was first written from the coursework notes alone. It has since been checked against two general pharmacology textbooks, chunked and indexed into a searchable knowledge base — 382 chunks · 3.20 M characters · 914 figures. The checking turned up something worth stating plainly, because it decides what you can cite for what.
Lippincott Illustrated Reviews: Pharmacology, 6e
680 pages → 347 section chunks, split on the book's own outline. Every mechanism claim on this page — β-lactams acylating the transpeptidase, aminoglycosides being the one bactericidal protein-synthesis inhibitor, amphotericin B binding ergosterol, gray baby syndrome — is verifiable here, page-anchored. Cited inline above.
Pharmacology — An Illustrated Review (Thieme, Simmons 2011)
35 chapter chunks, 203 figures. Used as an independent second statement of the same mechanisms — where the two books agree, the claim is safe; where the emphasis differs, the difference is worth knowing.
NeoFax 2020 · Cloherty 8e · AIMS NICU Manual
Reached through the coursework notes, not through the two textbooks. Every dose, every PMA interval, every neonatal Vd and half-life on this page belongs to this tier.
Every abbreviation, in plain words
The simulators are teaching models built on the neonatal Vd and half-life values documented in those notes; they show the right shape and magnitude and nothing more. Every dose must be confirmed against the current NeoFax monograph and the local unit protocol before prescribing.