Vitamin K Deficiency Bleeding

Overview and Considerations

Caroline W. Burke, MSN, RN, CPNP


J Pediatr Health Care. 2013;27(3):215-221. 

In This Article

Case Study Answers

1. What are the differential diagnoses for this infant?

This 4-week old infant presented with a history of irritability, increased sleepiness, decreased oral intake, fever, and projectile emesis. Without a clinically apparent source of infection on initial examination, the initial concern in this neonate is occult infection, with bacteremia, sepsis, bacterial meningitis, urinary tract infection, and pneumonia as the differential diagnoses (Baraff, 2008). Physical examination findings were not suggestive of septic shock. Urinary tract infections are the most common serious bacterial infection in children younger than 2 years, and laboratory analysis is crucial to rule out this diagnosis (American Academy of Pediatrics [AAP], 2011). In this case, the infant's urinalysis and urine culture findings were negative. An LP is an appropriate diagnostic tool in the setting of an infant with fever; however, contraindications for LP include concern for increased ICP and uncorrected coagulopathies. This infant's history and neurologic examination required the deferral of the LP while an intracranial mass, hemorrhage, or midline shift were ruled out to prevent possible uncal herniation (Mann & Jackson, 2008).

Given his irritability, emesis, and full fontanels, an expanded differential that includes neurologic processes such as elevated ICP is appropriate. Increased ICP is commonly the result of traumatic brain injury, hydrocephalus, masses or tumors, hypoxic/ischemic brain injury. or intracranial hemorrhage (Singhi & Tiwari, 2009). Although assessment can be challenging in infants, it is important to be alert for the early signs and symptoms of increased ICP, including irritability, emesis, and pupillary sluggishness. Later signs include bulging of fontanels, sun-setting eyes, seizures, cranial nerve dysfunction, decreased spontaneous movement, posturing, pupillary inequality or dilation, hypertension, bradycardia, and irregular respirations (Keefe & LeFlore, 2005; Holleman et al, 2012, Lam, 2012). In an infant, the presence of intracranial hemorrhage is likely due to a traumatic process, a coagulation disorder, or a vascular abnormality (Hubbard & Tobias, 2006). Given this infant's age, elevated PT and PTT, and lack of vitamin K prophylaxis at birth, the concern for VKDB is high. In the setting of vitamin K deficiency, an abnormal form of coagulation factor II, also referred to as "protein induced by vitamin K absence" (PIVKA-II), is released into the bloodstream and can be directly measured (Gopakumar, Sivii, & Rajiv, 2010). PIVKA-II is unique in that it can help identify early or subclinical vitamin K deficiency, and because of its long half-life, it can be used to identify VKDB on retrospective analysis (Clarke & Shearer, 2007a). Given the severity of this infant's presentation, PIVKA-II levels were not assessed in this situation. In retrospect, they could have been obtained to further support the child's diagnosis.

The presence of an intracranial hemorrhage in an infant raises the concern for nonaccidental trauma. Victims of severe head trauma are more likely to present with subdural and subarachnoid hemorrhages than are victims of accidents, along with multiple subdural hematomas in various stages of healing, skull fractures, retinal hemorrhages, or associated skin and skeletal injuries (Kellogg, 2007). Any infant or child with an unexplained brain injury or skull fracture should be evaluated for the possibility of nonaccidental trauma. It is reassuring that in this case the infant did not have any obvious injury, no skull fracture was seen on the CT scan, and the diagnosis of VKDB corresponded with the patient's age, presentation, and history. Nevertheless, it is important to keep the possibility of nonaccidental trauma in the differential diagnosis, because abusive head trauma and VKDB often present with similar, nonspecific symptoms (Kellogg, 2007).

2. What is the function of vitamin K in the body?

Vitamin K is a fat-soluble vitamin that is a critical part of the clotting cascade. It is necessary for synthesis in the liver of four clotting factors (II, VII, IX, and X), as well as the anticoagulation proteins C and S (Ardell, Offringa, & Soll, 2010). Two types of natural vitamin K exist: K1 (phylloquinone), which is plant based and typically is found in green leafy vegetables and oils, and K2 (menaquinone), which is produced through endogenous synthesis from intestinal flora (Ardell et al., 2010). Vitamin K works to facilitate binding of the procoagulation clotting factors II, VII, XI, and X with surface phospholipids through calcium ion channels, which initiates the thrombotic process (Sarnaik, Kamat, & Kannikeswaran, 2010). Deficits of vitamin K affect both the intrinsic and extrinsic clotting cascade and can result in prolonged clotting times and hemorrhage. Deficiency can occur relatively quickly because of the short half-life of the vitamin K–dependent coagulation factors. Deficiency is seen not only in newborns but also in persons with malabsorption syndromes, decreased production of bile salts, insufficient intake, or decreased intestinal flora related to antibiotic use (Ardell et al., 2010).

3. What is VKDB? What populations are at greatest risk?

VKDB, formerly known as hemorrhagic disease of the newborn, is a bleeding disorder caused by low levels of vitamin K–dependent clotting factors. Although vitamin K deficiency can occur in older children and adults, it is most common in newborns who have limited stores of vitamin K and immature gastrointestinal tracts. Placental transfer of vitamin K is low, and the serum levels of vitamin K–dependent factors have been found to be as low as 50% those of adults (Ardell et al., 2010). Diagnosis of VKDB can be made in infants younger than 6 months who have spontaneous bleeding, bruising, or intracranial hemorrhage with a prolonged clotting time but with a normal or elevated platelet count. The exception to this scenario is infants with an inherited coagulopathy or disseminated intravascular coagulation (Gopakumar et al., 2010).

VKDB can be classified into three distinct groups. Early VKDB typically occurs within the first 24 hours of life and is not affected by vitamin K prophylaxis at birth. This early-onset bleeding is related to medications taken by women in the intrapartum period that affect vitamin K storage and function in the newborn, including Warfarin, anticonvulsants, Rifampin, and Isoniazid (Hubbard & Tobias, 2006). Women taking these medications should be given 5 mg of oral vitamin K daily during their third trimester of pregnancy to prevent the onset of early VKDB. Classic VKDB presents between day 2 and 7 of life and typically involves gastrointestinal, nasal, skin, or circumcision site bleeding. This classic form is related to the low placental transfer of vitamin K, low concentration in breast milk, lack of gastrointestinal flora in the newborn gut, and the poor oral intake that commonly occurs in the newborn period as breastfeeding is initiated (Hubbard & Tobias, 2006). Classic VKDB has been reported to occur in as many as 0.25% to 1.7% of infants without prophylaxis (AAP, 2003). Late onset of VKDB occurs between 7 days and 6 months and is seen primarily in infants who are exclusively breastfed and who have not received prophylaxis, although it may be related to other complications that interfere with synthesis or storage of clotting factors. Although few data are available on the incidence of late VKDB in the United States, studies from Europe and Asia have reported a prevalence of 4.4 to 7.2 infants per 100,000 births (AAP, 2003). Common symptoms of late VKDB in infants include vomiting (44%), bulging fontanelles (40%), pallor (40%), decreased appetite (32%), seizures (40%), and "warning bleeding" occurring at other sites such as the nares, mucosa, or umbilicus (36%; Misirlioglu et al., 2009). Up to 50% of infants with late VKDB present with intracranial hemorrhage (Van Hasselt, et al., 2008), and up to 69% of these infants have multifocal hemorrhaging. As such, late VKDB carries a significant morbidity and mortality rate, with mortality as high as 20% to 50% in various studies (Cekinmez, Cemil, Cekinmez, & Altmors, 2008). Follow-up from a 2009 study of 29 infants with VKDB-associated intracranial hemorrhaging found that 57.1% were developmentally normal and 42.8% had neurologic defects including hydrocephalus, cerebral atrophy, encephalopathy, epilepsy, and developmental delay (Misirlioglu et al., 2009).

Breastfeeding has been implicated as a risk factor for the development of VKDB, because breast milk has very low levels of vitamin K (< 5–15 μg/L) when compared with formula (50–60 μg/L; Van Hasselt et al., 2008). Children with hepatic or intestinal disease that interferes with absorption of vitamin K also are at increased risk of developing VKDB (Van Hasselt et al., 2008). For some persons, VKDB is the first indicator of an underlying disorder such as malabsorption, infectious diarrhea, cystic fibrosis, alpha1-antitrypsin deficiency, hepatic dysfunction, or celiac disease (Van Hasselt et al., 2008).

4. What is the appropriate management for persons with VKDB?

Once the diagnosis of VKDB is established, the priority is to rapidly correct the coagulopathy to prevent ongoing hemorrhage and minimize complications. Administration of vitamin K is the most appropriate means to correct a vitamin K deficiency and has been shown to normalize PT and PTT in as little as 4 to 6 hours (Cekinmez et al., 2007). One study showed a 30% to 50% correction of PT within 1 hour (Clarke & Shearer, 2007b). Doses of 1 to 3 mg of vitamin K have been administered in the setting of VKDB via intramuscular, IV, or subcutaneous injection (Clarke & Shearer, 2007b). Studies have shown that parenteral administration of a single dose of vitamin K in the setting of VKDB is sufficient to correct coagulation disturbance (Ardell et al., 2010). Intramuscular vitamin K has been shown to be effective in correcting the PT and reversing the coagulopathy, but the concern exists for hematoma associated with intramuscular injection in a person with coagulopathy. IV vitamin K administration carries less risk of hematoma formation, but reports have been made of allergic and anaphylactic reactions to this form of administration. Less risk of hematoma or anaphylaxis occurs with subcutaneous administration; however, the drug absorption with subcutaneous administration been shown to be inconsistent (Gopakumar et al., 2010).

FFP also is used frequently to urgently correct coagulopathies in infants with VKDB (Sarnaik et al., 2010). FFP is plasma taken from a unit of whole blood, which contains all coagulation factors in their normal concentrations and can correct the PT. Transfusion is indicated in patients presenting with a coagulation deficiency who are actively bleeding or undergoing invasive procedures (Gopakumar et al., 2010). Dosed at 10 to 20 mL/kg, FFP ideally requires cross-matching before administration, although AB plasma can be administered in emergent situations (Helfaer & Nichols, 2009). Although Hubbard and Tobias (2006) expressed a concern regarding the possibility for delay in administration because of the necessity of thawing FFP before administration and the concern for volume overload in the pediatric population, a review of the literature reveals that administration of FFP remains standard practice in the setting of severe VKDB (Gopakumar et al., 2010; Hubbard & Tobias, 2006;Sarnaik et al., 2010). The use of recombinant factor VIIa, a newer clotting agent currently approved by the Food and Drug Administration for use in patients with hemophilia and factor deficiencies, was discussed in one case study. Results showed reversal of coagulopathy after administration of recombinant factor VIIa. Further analysis of its use in the pediatric population continues (Hubbard & Tobias, 2006).

After correction of the underlying coagulopathy, the priority is the management of hemorrhage-associated sequelae, which are largely dependent on the location and severity of the bleeding. In an infant or child with a complicated intracranial hemorrhage, management includes hospitalization with intensive care monitoring, management of ICP, neurosurgical evaluation with surgical decompression and evacuation of clots once coagulopathy has resolved, and a neurology evaluation for seizure management.

5. What are the current recommendations for vitamin K prophylaxis? What are the concerns?

After the identification of an association between vitamin K and spontaneous hemorrhage in newborns, the AAP began recommending in 1961 that all newborns receive vitamin K prophylaxis to prevent hemorrhage, either as a one-time intramuscular dose or as a series of 3 oral doses (AAP, 1961). In the 1950s and 1960s, concern was raised regarding the increased incidence of hemolytic anemia and kernicterus associated with the use of menadione, or VK3, a water-soluble form of vitamin K used at the time (Hubbard & Tobias, 2006). In 1990, a British study linked administration of intramuscular vitamin K at birth with an increased risk of childhood cancers (Golding, Paterson, & Kinlen, 1990). A follow-up study in 1992 showed no increased risk with oral vitamin K but supported the association between intramuscular prophylaxis and the development of childhood cancer. This study showed an increased rate of leukemia in children who had received intramuscular vitamin K, and the authors subsequently recommended prophylaxis exclusively with oral vitamin K to eliminate the increased risk of leukemia (Golding, Greenwood, Birmingham, & Mott, 1992). An early preparation of intramuscular vitamin K (Synkavit) included the emulsifier polyethoxylated castor oil, as well as the preservatives propylene glycol and phenol, which was shown to cause tumors in mice (Clarke & Shearer, 2007a) and was cited as being administered to infants in the 1992 Golding study (Golding et al., 1992). Newer formulations of vitamin K, Konakion and AquaMEPHYTON, contain more natural emulsifiers and have an "improved safety profile" with less risk of jaundice and anaphylaxis and no known carcinogenic properties (Clarke & Shearer, 2007a).

Multiple studies subsequent to the 1990 and 1992 studies by Golding and colleagues found no correlation between intramuscular vitamin K administration and an increased incidence of childhood cancers (AAP, 2003;Draper & Stiller, 1992;Hubbard & Tobias, 2006; Ross & Davies, 2000). In 2003, the AAP Vitamin K Ad Hoc Task Force concluded that no evidence exists to support an increased risk of cancer with the administration of parenteral vitamin K, and the AAP continues to recommend the administration of intramuscular vitamin K at a dose of 0.5 to 1 mg for all newborns (AAP, 2003).

Historically, much debate has ensued about the best route and effectiveness of oral versus intramuscular vitamin K prophylaxis. The benefits of prophylaxis for prevention of VKDB are well established, and studies have shown that prophylactic administration, regardless of route, is more cost-effective than management of VKDB sequelae (Ross & Davies, 2000). Practice has varied during the past 50 years and continues to vary from country to country. Oral vitamin K prophylaxis has the benefit of being minimally invasive and less expensive than the intramuscular version. Because oral administration generates lower peak levels and lower sustained levels than parenteral administration, an oral regimen requires repeated dosing and includes the risk of missed doses. In the 1990s, the Netherlands and Denmark both offered examples of oral prophylaxis models. Infants in the Netherlands received 1 mg of vitamin K orally at birth and 25-μg daily doses for 3 months. In 2005, a study showed a discouraging 3.2/100,000 incidence of late VKDB (Ijland, Pereira, & Cornelissen, 2008). Danish infants were given 2 mg orally at birth and weekly 1-mg doses for 3 months. From 1992 to 2000, no VKDB cases were reported in a population in which 71% of mothers reported breastfeeding (Clarke & Shearer, 2007a). Intramuscular vitamin K administration is more invasive than oral administration, but it is a one-time dose with no concerns for ongoing compliance. Although early concerns were expressed regarding jaundice risk and more recent concerns have regarded the carcinogenicity of the intramuscular preparation, no evidence has supported this concern since 1992, and the modern formulation has been shown to be safer than the one previously used. The AAP at this time recommends only intramuscular administration, but it does cite a need for continued research regarding the optimal dosing and safety of oral prophylaxis (2003).

Equally if not more pressing is the need for promotion of the risks of VKDB related to missed prophylaxis. Although parents have a right to make medical choices for their children, parental refusal of vitamin K administration has been associated with 31% to 70% of cases of VKDB (Clarke & Shearer, 2007b). The AAP recommends administration of prophylactic vitamin K to all newborns, regardless of the location of their birth (2003). Although midwifery regulations vary from state to state, formal state regulations and standards of practice typically include a recommendation for vitamin K administration (Department of Regulatory Agencies, 2012;New York State Department of Health, 2010). Although no concrete data exists regarding the incidence of vitamin K administration for infants born at home, it is clear that the frequency of home births is increasing. In 2010 in the United States, 0.8% of births took place in the home (U.S. Department of Health & Human Services, 2012), an increase from 0.73% in 2009 (U.S. Department of Health & Human Services, 2011) and 0.6% in 2007 (U.S. Department of Health & Human Services, 2010). No direct evidence exists to link home birth to an increased rate of vaccination and vitamin K refusal. However, one qualitative study of women who chose home birth found that women choosing to give birth at home tended to do so in an attempt to "redefine authoritative knowledge," enhance their control of the birthing process, and create a greater sense of intimacy (Cheyney, 2008, p. 265). Although this finding does not directly indicate an increased likelihood to refuse vaccinations, it does reflect an overall questioning of the "validity of mainstream. metanarratives" regarding the birth experience as a whole (Cheyney, 2008, p. 265).

Refusal of vaccinations and prophylactic medications such as vitamin K has continued associated risks that create challenges for the provider and health system. Discussions that include the risk and benefits of safe vitamin K administration may correct any misconceptions and misinformation that would allow parents to make a fully informed decision (AAP, 2005).