Part One / In the Beginning
The First Three Days
Hospital choices, newborn screening, weight loss, and jaundice.
In this note
The one thing to remember
Almost everything in the hospital falls into one of two buckets — routine screening that is clearly worth doing, or a preference call where the stakes are genuinely small either way. The two things that cause the most panic, weight loss and jaundice, are normal biological processes. Both numbers are meaningless until you read them alongside the baby's age in hours.
Cutting through the noise#
You will be asked to decide things you have never thought about, on no sleep, by people who ask as though the answer is obvious. Some of these questions are medical (vitamin K, screening). Some are cultural dressed up as medical (circumcision, bath timing). And some — weight loss, jaundice — are presented as alarms when they are actually routine monitoring.
The useful mental move is to sort each question before answering it: Is this a real medical decision, a values decision, or a monitoring number I'm misreading as a verdict?
The 60-second recap#
- Newborn blood and hearing screens are worth it, no debate. They look for uncommon conditions where catching it in week one changes the whole outcome. PKU, for example, affects about 1 in 10,000 babies, and untreated it causes brain damage almost immediately — because both breast milk and formula contain the protein they can't process. Caught at birth, it's managed with diet.
- There is no medical reason to bathe a newborn right away. The blood mostly wipes off. Delaying may help skin-to-skin contact and early feeding, and there's nothing to lose.
- Circumcision has small real benefits and small real risks. Both are small enough that culture, family, and preference are legitimately the deciding factors. If you do it, insist on real pain relief.
- Rooming in has not been shown to improve breastfeeding long-term. The one randomized trial found nothing at six months. Using the nursery to sleep is not a failure.
- Nearly all newborns lose weight, and breastfed babies lose more. This is expected: your milk hasn't come in yet. A percentage without a timestamp tells you nothing.
- About half of newborns get jaundiced. The treatment thresholds are deliberately cautious, so being treated does not mean your baby was in danger.
- Delayed cord clamping is generally favorable, strongly so for premature babies. Get the vitamin K shot.
What to actually do#
MUST - HighComplete the newborn blood screen and hearing screen, and follow up promptly on anything abnormal
These catch rare conditions where early treatment prevents permanent damage. Zero risk to the baby — it's a heel prick and a sensor
MUST - HighGive the vitamin K shot unless a clinician identifies a specific contraindication
Without it, about 1.5% of infants have unexpected bleeding in week one, plus rarer serious bleeding later. The 1990s cancer scare came from small, methodologically weak studies and was refuted by follow-up work
MUST - HighLet the baby's clinician interpret significant weight loss, dehydration signs, or bilirubin
These numbers are only meaningful against age-specific curves and risk categories. This is exactly where DIY reading goes wrong
MUST - HighIf the baby is in your room, put them in their own safe sleep space before you might fall asleep
Falling asleep while holding or feeding a baby is the actual mechanism behind hospital bed-sharing injuries. Ask for help rather than fighting it
RECOMMENDED - HighDelay cord clamping when medically feasible, especially for a premature baby
For preemies, randomized trials show better blood volume, less anemia, fewer transfusions. For full-term babies the case is good but milder: more iron, less anemia later, slightly more jaundice
OPTION - ModerateRoom in, use the nursery, or mix the two
The single randomized trial (176 women) found no effect on breastfeeding at six months. The correlation people cite runs backwards: mothers who already intend to breastfeed are the ones who choose to room in
OPTION - ModerateCircumcision is a values decision
Roughly 100 circumcisions prevent 1 childhood UTI. Complication rates are low. Neither side of the ledger is big enough to override family, religion, or preference
OPTION - ModerateDelay the first bath; prefer a tub bath over a sponge bath if bathing in hospital
Sponge baths leave the baby wet and exposed longer, which causes temperature swings. The swings aren't dangerous, but they can be misread as infection and trigger unnecessary workups
OPTION - ModerateEye antibiotic ointment depends on maternal infection risk and state law
It prevents 85–90% of gonococcal eye infections, which cause blindness. But routine STI testing in pregnancy means most mothers are already known to be negative. Some states mandate it regardless
The two numbers that scare new parents#
Weight loss#
What's happening: in the womb the baby got nutrients through the cord. Now they have to learn to eat, and for the first few days there isn't much milk — just colostrum, and not a lot of it, especially with a first baby. Losing weight isn't a malfunction. It's the system working as designed.
Why it's monitored anyway: excessive loss can be an early signal that feeding isn't working, which risks dehydration, which makes feeding harder still. Monitoring catches that spiral early.
The mistake to avoid: treating a single percentage as a threshold. From a study of 160,000 births:
- At 48 hours, the average breastfed baby born vaginally has lost about 7% of birth weight. About 5% of them have lost more than 10%.
- Babies born by caesarean lose a bit more, and are typically in hospital longer.
- Formula-fed babies lose much less — about 3% at 48 hours — because formula doesn't need to "come in."
So a 10% loss at 72 hours is inside the normal range. A 10% loss at 12 hours would be a serious outlier. Same number, completely different meaning.
What to ask: "Where does this fall on the age-specific curve for a baby of this delivery and feeding type?" That's a better question than "is 10% bad?"
A parallel check you can do yourself: the actual worry is dehydration, and you can watch that directly. Is the baby peeing regularly? Is their mouth moist? If yes, dehydration is unlikely regardless of the percentage. If no, that matters even if weight loss looks fine.
If you do need to supplement: there's no good evidence that a short period of formula supplementation harms long-term breastfeeding. Oster found this genuinely reassuring, and it's worth knowing in advance because it will be presented to you at 2 a.m.
Jaundice#
What's happening: bilirubin is a byproduct of breaking down red blood cells. Newborns have three things working against them at once — more cells being broken down, an immature liver, and not much food moving through the gut, so bilirubin gets reabsorbed instead of excreted.
Why it's taken seriously: at very high concentrations bilirubin crosses into the brain and causes kernicterus, permanent brain damage.
Why you shouldn't panic: the effect is not a continuum. At low and moderate levels, bilirubin doesn't cross the blood–brain barrier at all — it isn't doing a little bit of damage. And the scale of the caution is worth knowing: the US sees 2 to 4 cases of kernicterus per year, while tens of thousands of babies are treated for jaundice every week. Treatment protocols are deliberately aggressive, accepting a lot of unnecessary treatment to avoid a single catastrophic case.
What determines treatment: bilirubin level, age in hours, gestational age, and risk category — not skin color. Oster's pediatrician put it well: people will tell you the baby is yellow, and you can ignore them.
Treatment itself: phototherapy is a blue-light bassinet. The light breaks bilirubin down into forms the body can pass in urine. Anywhere from a few hours to a few days, with breaks for feeding.
More likely if: exclusively breastfed, Asian heritage, mother-baby blood type mismatch, significant weight loss, or bruising during delivery.
Guidelines here evolve, and at the time of writing there was pressure to make them less aggressive. Reasonable to ask which guideline your clinician is using.
Numbers in context#
| Number | What it actually means |
|---|---|
| Hearing loss: 1–3 per 1,000 children | Screening catches 85–100% of cases but has a lot of false positives — about 4% of babies fail while only 0.1–0.3% truly have hearing loss. A failed first test is common. Ask to retest before you leave |
| UTI: 1% uncircumcised vs. 0.13% circumcised | Real and consistent, but in absolute terms you'd circumcise ~100 boys to prevent 1 UTI |
| Circumcision complications: ~1.5% minor, essentially none serious | Based on data including developing countries, so US rates are likely lower still |
| Penile cancer: ~1 in 100,000 men | Risk is genuinely higher without circumcision, but a large increase on a tiny base is still tiny |
| Phimosis needing later treatment: 1–2% | Uncommon but not unheard of; can mean circumcision at an older age |
| Kernicterus: 2–4 US cases/year vs. tens of thousands treated weekly | The gap between these two numbers is the safety margin |
Practical playbook#
Before delivery — decide these while you can think:
- Circumcision (and if yes, confirm the hospital uses a penile nerve block, not just sugar water — infants who feel pain during circumcision react more strongly to vaccination pain four to six months later)
- Bath timing, rooming in vs. nursery, eye ointment
- Ask whether delayed cord clamping and immediate skin-to-skin are routine
In the hospital:
- Write down birth weight, every later weight, the exact time of each, and the feeding method. You will not remember, and the timestamps are what make the numbers interpretable
- Ask what percentile the weight loss represents for this baby's circumstances
- Watch urine output and a moist mouth as your own dehydration check
- If the baby looks yellow, ask whether bilirubin was actually measured
- Ask for nursery help if exhaustion is making it unsafe for you to hold or feed the baby. This is what the nursery is for
Tools the book mentions:newbornweight.org and bilitool.org plot your baby against the same curves clinicians use. Treat them as context for a conversation, not as a reason to overrule anyone.
Where the evidence runs out#
- Whether rooming in has any long-run breastfeeding effect at all. The trial was small.
- How common some circumcision complications really are, and whether HIV/STI findings from settings with mostly heterosexual transmission transfer to the US.
- The right jaundice threshold for every well newborn. Guidelines are actively being revised.
Our family's notes#
- Choices we're leaning toward:
- What matters most to us here:
- Decisions to lock in before delivery:
- Questions for the hospital or pediatrician:
- What would make us change our minds: