Vitamin K Injection at Birth: What Parents Should Know Before Saying No

Vitamin K prophylaxis at birth is one of the most studied and effective neonatal interventions in modern medicine. A single 0.5–1 mg intramuscular dose has been the standard since the 1960s, virtually eliminating a once-fatal disease called Vitamin K Deficiency Bleeding (VKDB) of the newborn. Despite that record, refusal of the injection has risen in some U.S. communities, and pediatric hospitals have reported clusters of late-onset VKDB — including infant deaths and permanent neurological disability — that would not have occurred had the injection been given.

Why Newborns Need It

Vitamin K is a clotting cofactor; without it, factors II, VII, IX, and X cannot be activated. Newborns are vitamin K deficient by default: vitamin K crosses the placenta poorly, breastmilk contains very little, and the infant gut has not yet been colonized with K-producing bacteria. This deficiency window is widest for exclusively breastfed infants and lasts for the first several months.

Three Forms of VKDB

Early VKDB (first 24 hours) is uncommon and almost exclusively associated with maternal anticonvulsant or anticoagulant use. Classic VKDB (2–7 days) presents as bleeding from the umbilical stump, mucous membranes, or circumcision sites. Late VKDB (2 weeks to 6 months) is the most dangerous form — about half of cases present with intracranial hemorrhage. Without prophylaxis, the AAP estimates late VKDB rates of 4.4–7.2 per 100,000 live births; with IM prophylaxis, the rate is essentially zero.

The 1990 Confusion

In 1990, a single observational paper by Golding et al. in BMJ reported a possible association between IM vitamin K and childhood leukemia. Subsequent and much larger case-control studies (Roman 2002, Fear 2003, Parker 1998 among others) found no such association, and the original signal was attributed to confounding by injection use in higher-risk newborns. The AAP, AAFP, and ACOG explicitly endorse IM vitamin K as safe; the leukemia hypothesis is considered refuted.

Oral Alternatives Don't Work as Well

Some parents request oral vitamin K. Multiple oral protocols exist (most commonly 2 mg at birth, 2 mg at 1 week, 2 mg at 4 weeks for breastfed infants). All oral regimens have lower late-VKDB protection than a single IM dose. Denmark and the Netherlands have used oral protocols, but late-VKDB cases have been documented even with full adherence. The AAP recommends IM prophylaxis as first-line and considers oral acceptable only when parents refuse IM, with explicit informed consent about the residual risk.

The Tennessee Cluster

A 2013 CDC MMWR report described five Nashville-area infants with late VKDB and intracranial hemorrhage between February 2013 and September 2013. All had refused vitamin K injection at birth. Three of the five required emergency neurosurgery. The cluster was widely cited as evidence that refusal is not a low-stakes choice. Similar clusters have since been reported from Oregon, Iowa, and the U.K.

Bottom Line for Parents

The IM injection delivers a single dose of phytonadione (vitamin K1) within the first hour of life, preventing all three forms of VKDB with near-100% efficacy and no demonstrated long-term harms. Refusing the injection is a position with consequences that current evidence does not justify.