Vitamin K for newborns — why it's offered

The question

Should my newborn get the vitamin K injection, or are the oral drops just as good?

Short answer

Newborns are born with very low vitamin K, which blood needs to clot — and without it, a rare but dangerous bleeding disorder can strike, including brain bleeds in the first months of life [9]. A single intramuscular injection after birth prevents classic bleeding (days 2–7), as shown in randomized trials [1], and surveillance data suggest it also prevents the later, more dangerous form [2]. Oral drops are an alternative, but they have only been tested for blood markers, never for actual bleeding, in trials [1] — so the evidence suggests the injection protects better, especially against late bleeding [2].

What the strongest evidence says

Why newborns are running on empty

Vitamin K helps blood clot, but the body's stores are very low at birth: only small amounts pass through the placenta, the gut bacteria that make vitamin K in adults aren't present yet in the newborn's intestines, and breast milk is also low in vitamin K [9]. That last point matters — late bleeding tends to occur in exclusively breastfed babies who didn't receive the shot [9]. (That doesn't mean breastfeeding causes the problem; it means breastfed babies without prophylaxis are the ones left unprotected.)

The three faces of the bleeding

The timing matters because the evidence is strongest for the middle column and thinnest for the last [1][2][9].

The injection prevents classic bleeding — shown in trials

A Cochrane review found two randomized trials comparing a single intramuscular vitamin K dose with placebo or nothing that measured actual bleeding: the injection reduced clinical bleeding at days 1–7, including bleeding after circumcision, and improved blood-clotting measures [1]. This is the strongest evidence in the field — randomized, measuring real bleeding, not just lab values.

Late bleeding: the danger, and what the data show

Late VKDB is the form parents fear most, because 30–60% of late cases involve bleeding inside the skull [2]. There are no randomized trials of either route against late bleeding — running one today would mean deliberately leaving babies unprotected, which ethics committees won't allow. What we have instead is national surveillance data, pooled in a systematic review: late VKDB at a median of 8.8 per 100,000 births in high-income countries without prophylaxis, and a pooled relative risk of 0.02 for IM prophylaxis versus none [2]. That's strong observational evidence, but it is observational: countries differ in more ways than vitamin K policy, so say suggests, not proves.

Injection vs drops: an untested comparison

Eleven randomized trials compared oral regimens — but none measured actual bleeding, only blood markers of clotting [1]. Surveillance data suggest a single oral dose performs worse than the injection against late bleeding, while repeated oral schedules come closer, but the estimates are imprecise [2]. The honest summary: the injection's advantage over oral drops for real bleeding has never been tested in a trial. If you choose drops, you're relying on lab values and surveillance, not trials. The AAP's 2022 policy statement is blunter than the trials allow this page to be: oral prophylaxis is "not as effective" as the injection at preventing late-onset VKDB and therefore "cannot be recommended" [10].

The cancer scare was tested — and didn't hold up

An early-1990s concern that the injection might cause childhood leukaemia was investigated in a UK national case–control study (2,530 children with cancer, including 1,174 with leukaemia, and 4,487 controls) and in a pooled analysis of six case–control studies (2,431 cases, 6,338 controls). Both found no convincing association, regardless of the route given [4][5]. The pooled analysis was careful: small effects can't be entirely ruled out, but there is no convincing evidence of any link [5].

What it means for the parents

This decision lands in the first hours after birth, when you are exhausted, possibly in pain, and being asked to consent to something you've never thought about. A few things worth knowing:

• Awareness gaps are real. In a CDC investigation of four bleeding cases in Tennessee in 2013 — all in babies whose parents had declined the injection — investigators found parents' awareness of the bleeding risk was "incomplete or absent," and the reasons for declining included the long-debunked cancer claim, a belief the shot was unnecessary, and a wish to minimize "toxins" [3].
• The oral route shifts work onto you. One common UK oral regimen is three doses (after birth, at days 4–7, and around one month for fully breastfed babies), but schedules vary by trust [8] — confirm the exact schedule with your own midwife. The injection is one and done. If life with a newborn makes you doubt you'll complete a multi-dose schedule, that matters — a missed dose is an unprotected baby.
• Declining is increasingly common, and worth discussing openly. In the US, nonreceipt rose from 2.92% in 2017 to 5.18% in 2024 [7]. Whatever you decide, the evidence supports making the decision with full information about the bleeding risk — not discovering it in an emergency department at 10 weeks.
• The decision can land on the birth partner. In the minutes after birth, the birth partner may be the one asked to consent. If vitamin K matters to you, discuss it before labour — the birth room is a bad place for a first conversation about bleeding risks.
• Guilt is not evidence. Whatever you decide, make the decision with the numbers above — not with fear of the needle or fear of the bleed. Neither feeling tells you which choice the evidence supports.

What remains uncertain

• Whether oral drops prevent actual bleeding (classic or late) as well as the injection: untested in trials [1].
• Whether a single oral dose vs repeated oral doses differ meaningfully for late bleeding: surveillance estimates are imprecise [2].
• Minor harms such as injection-site redness have not been systematically quantified in published studies.
• Match the wording to the rating: the evidence suggests the injection prevents late bleeding and suggests it beats oral drops — it does not prove it, because the late-bleeding evidence is observational.

Benefits and risks in absolute terms

Without any vitamin K prophylaxis:

• Early/classic bleeding: 0.25–1.7% of births — about 2.5 to 17 in every 1,000 babies [3].
• Late bleeding: 4.4–7.2 per 100,000 in older US/UK-era estimates [3]; a systematic review's high-income-country median is 8.8 per 100,000 [2]. Estimates vary by era and definition; the order of magnitude is consistent.
• Late bleeding is the dangerous form: 30–60% of late cases involve intracranial bleeding [2]. In a synthesis of 40 published case reports and series (1,486 babies, mostly from countries without routine prophylaxis), 20.5% died and 9.8% had significant neurological sequelae [6] — but case reports select the worst outcomes, so do not read these as population risks.

With the intramuscular injection:

• The CDC reports infants who do not receive the shot are 81 times more likely to develop late VKDB than those who do (an estimate sourced to a 1991 British prospective study, not a trial) [9].

Harms:

• Brief pain from the injection itself. Other minor harms have not been systematically quantified in published studies.
• The childhood-cancer hypothesis has been tested in large case–control studies that found no convincing evidence of any link [4][5].

Practical considerations

• The injection is a single 1 mg intramuscular dose given shortly after birth [8]; if you prefer oral drops, confirm the exact schedule (doses and timing) with your own midwife or trust, as it varies by formulation [8].
• Oral doses must actually be swallowed and kept down — if your baby spits one up, tell your midwife rather than assuming it counted.
• If you are leaning toward declining, say so early and ask for a proper conversation about the bleeding risk — the data show many parents who declined did so without full information [3][7].
• Put the decision in your birth plan or birth preferences document, whichever way you lean — it removes one decision from the busiest hour of your life.
• If your baby is premature or unwell, ask the neonatal team about vitamin K — this page covers the standard offer for well term babies.
• You can ask for a short delay. To protect immediate skin-to-skin bonding, the injection can be given up to 6 hours after birth [9] — you don't have to choose between bonding and protection.
• There are often no warning signs. In most cases of VKDB there is no advance notice before a life-threatening bleed begins [9] — which is the core argument for prevention over watchful waiting.

When to talk to your doctor, midwife, or pediatrician

Seek urgent care if your baby shows signs that could indicate bleeding problems [6][9]: blood in the stool or vomit, black or tarry stools, unusual bruising (especially around the head and face), nosebleeds, bleeding from the umbilical stump or injection sites, paler skin than before, unusual sleepiness or floppiness, irritability, seizures, or a lot of vomiting. Crucially, most cases have no warning signs at all before a life-threatening bleed [9] — so if you declined vitamin K and anything seems wrong in the first months, say so immediately. It changes the assessment.

References

1. Puckett RM, Offringa M. Prophylactic vitamin K for vitamin K deficiency bleeding in neonates. Cochrane Database Syst Rev. 2000;(4):CD002776. https://pmc.ncbi.nlm.nih.gov/articles/PMC6718236/ — A
2. Sankar MJ et al. Vitamin K prophylaxis for prevention of vitamin K deficiency bleeding: a systematic review. J Perinatol. 2016;36(9):701–10. https://pmc.ncbi.nlm.nih.gov/articles/PMC4862383/ — B
3. CDC. Notes from the Field: Late Vitamin K Deficiency Bleeding in Infants Whose Parents Declined Vitamin K Prophylaxis — Tennessee, 2013. MMWR. 2013;62(45):901–2. https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6245a4.htm — C
4. Fear NT et al. Vitamin K and childhood cancer: a report from the United Kingdom Childhood Cancer Study. Br J Cancer. 2003;89(7):1228–31. https://www.nature.com/articles/6601278 — B
5. Roman E et al. Vitamin K and childhood cancer: analysis of individual patient data from six case–control studies. Br J Cancer. 2002;86(1):63–69. https://pmc.ncbi.nlm.nih.gov/articles/PMC2746550/ — B
6. Dineen L, Hagan K. Refusal of intramuscular vitamin K by parents of newborns: a review. https://pmc.ncbi.nlm.nih.gov/articles/PMC7041551/ — C (case-report synthesis; severe outcomes over-represented)
7. Scott et al. Trends in Vitamin K Administration Among Infants. JAMA. 2025. https://doi.org/10.1001/jama.2025.21460 — B
8. University Hospitals Coventry & Warwickshire NHS Trust. Guidance for parents about vitamin K (patient leaflet). https://www.uhcw.nhs.uk/download/clientfiles/files/Patient%20Information%20Leaflets/Women%20and%20Children_s/Maternity/Guidance%20for%20parents%20about%20Vitamin%20K.pdf — D (practice guidance)
9. Centers for Disease Control and Prevention. About Vitamin K Deficiency Bleeding. https://www.cdc.gov/vitamin-k-deficiency/about/ (accessed 2026-09-08). — D (public-health information page; cites AAP 2022, Zipursky 1999, Sutor 1999, McNinch & Tripp 1991)
10. Hand I et al. Vitamin K and the Newborn Infant. Pediatrics. 2022;149(3):e2021056036. https://publications.aap.org/aapnews/news/19601/Vitamin-K-policy-addresses-birth-dose-for-term — B (guideline/policy statement; oral-vs-IM recommendation based on surveillance review plus expert consensus, not an RCT)

Changelog

• 2026-09-08: Topic created (draft). Awaiting independent review. (Superseded 2026-09-09: author-review model, no external review before v1.)
• 2026-09-09: Full author review (v1 author-review model; reviewer Guille, executed by AI agent under his explicit delegation of 2026-09-09). Short answer corrected ("early form" → "classic bleeding, days 2–7") to match the Cochrane trials' actual outcome window. Added the AAP 2022 policy statement (Hand I et al., Pediatrics): oral prophylaxis "not as effective" as IM for late-onset VKDB and "cannot be recommended" — the institutional position behind the draft's conservative "suggests" language. Softened "refuted" to "found no convincing evidence of any link" for the cancer question (case–control studies can't fully exclude small effects). All nine author flags resolved; rating B stands. Verdict: approve.