Subjects and Methods

This is a retrospective study where charts of children diagnosed to have late HDN were scrutinized. Details regarding pregnancy and delivery, use of vitamin K prophylaxis, method of feeding, timing of presentation, underlying illness in the baby, laboratory findings and outcome were recorded. Any infant less than 6 months’ old presenting with hemorrhage without any evidence of infection and having a normal platelet count, pro-longed prothrombin time (PT) and activated partial thromboblastin time (APTT), which normalized within 12 to 24 hours after administering vitamin K, was considered to have late HDN.

Results

Of 14 infants diagnosed to have late HDN, 8 were boys. Eight were born in the hospital of which 4 received intramuscular vitamin K. The rest were delivered at home and did not receive any vitamin K prophylaxis. None of the mothers were on medications during pregnancy, which could alter the coagulation status of these infants. All 14 infants were exclusively breast-fed. The age at presenta-tion ranged between 1 and 3 months. Twelve infants did not have any underlying illness; 2 had suspected liver disease, with raised blood levels of aminotransferases and bilirubin. Liver biopsy was not done in these children. Convulsions (71%), poor feeding (50%), vomiting (57%) and altered sensorium (36%) were the common presenting symptoms. Fever was reported in 50% of the cases. Pallor was seen in all our infants (100%). Anterior fontanel was tense and bulging in 64% infants. CT scan showed intracranial hemorrhage in 11 (78%) cases. Two infants did not have CT scan but presented with clinical features suggestive of IC bleed. Only 1 infant had a warning bleed (umbilical bleed preceding intracranial bleed). Assisted ventilation was necessary in 9 cases. Eight patients, including 5 who were discharged against medical advice, died (57%); 6 patients were discharged, of which 5 (36%) had neuro-logical sequelae and only one was well. Table I gives the clinical summary of the 14 patients.

Table I__Summary of Patients with Late HDN

Case
no.

Vitamin K
at Birth

Age

Bleeding site

Duration 
of symptoms

PT

APTT

Outcome

1

Nil

2 months

Intracranial

3 days

1'14"

2'

Died

2

Nil

2 months

Intracranial*

2 days

>1'

>2'

DAMA

3.

Nil

1 month

Intracranial

2 days

>1'

>2'

DWS

4.

Nil

1 m 15 d

Intracranial

6 days

52"

40.8"

DWS

5.

Nil

2 months

Intracranial

2 days

>1'

>1'

DWS

6.

Nil

2 months

Intracranial

2 days

>1'

93"

Died

7.

Nil

2 months

Intracranial*

6 days

>1'

>2'

Died

8.

Nil

2 months

Skin bleeds

1 day

>1'

>2'

DW

9.

Nil

2 months

Intracranial

1 day

>1'

108"

DWS

10.

Nil

1 m 15 d

Intracranial

2 days

>1'

>2'

DWS

11.

Yes

1 month

Intracranial &

3 days

>1'

>2'

DAMA

12.

Yes

3 months

Intracranial

1 day

>1'

>2'

DAMA

13.

Yes

1 month

Intracranial

1 day

>1'

>2'

DAMA

14.

Yes

1 m 15 d

Intracranial

3 days

>1'

>2'

DAMA

DAMA = discharged against medical advice; DWS = discharged with 
sequelae; DW = discharged well;

PT = prothrombin time, APTT = activated partial thromboplastin time.

Age referes to that at presentation; patients 13 and 14 had raised 
aminotransferases.

* Clinical features of intracranial bleeding.

Discussion

Late HDN may be primary or secondary to diarrhea, cystic fibrosis, biliary atresia, alpha-1-antitrypsin deficiency, hepatitis, abetalipoproteinemia, celiac disease or chronic warfarin exposure(2,3). There is little doubt that infants are vitamin K deficient at birth or develop vitamin K deficiency in the first few days of life(1-3). This deficiency becomes marked in infants who are exclusively breast-fed. However, only a minority of infants present with bleeding manifestations. This led to the questioning of the need of vitamin K prophylaxis in all term newborns(1). Intracranial hemorrhage can be presenting manifestation of late HDN in children leading to death in some cases and neurological sequelae in the survivors(1). In a study of 15 cases from Turkey(5), convulsions and poor feeding were seen in 47% patients and pallor in 20%. This is in contrast to our findings of convulsions in 71% and pallor in all patients. Bulging fontanel was seen in a comparable proportion of patients. Neurologic findings were seen in 72% patients with 33% succumbing to illness(5), whereas 93% of our subjects had neurologic signs with CT scan confirming intracranial bleeds in 78% and a mortality of 57%. In another review(6) of 108 cases of late HDN from Germany, 58% had intracranial hemorrhage with a mortality of 19% and neurological sequelae in 21%. Thirty six percent of our children had neurological sequelae. A prospective study(7) conducted over 2 years in the British Isles, found that 10 out of the 27 infants had intracranial hemorrhage; of these with intracranial bleed, 2 died.

Vitamin K prophylaxis reduces the incidence of late HDN from 5.1 cases per 100,000 births by 90%. A single parenteral dose reduces the risk by a factor of 14.3(6). The practice of parenteral vitamin K administration is not followed universally since Golding et al.(8) reported an association between intramuscular vitamin K use and doubling of the incidence of leukemia. The limiting factor of that study was that most information came from one hospital where 80% of the infants were not delivered spontaneously and only 4% of the spontaneously delivered infants received vitamin K, indicating that the mode of delivery may be a confounding variable(9). Later studies have proved no significant association between intramuscular vitamin K and childhood leukemia(1,9). Many hospitals have adopted a policy of selective prophy-laxis, where high risk infants are given vitamin K prophylaxis. There was a resur-gence of late HDN after the practice of selective prophylaxis(10), which led to the introduction of oral vitamin K. The results of the meta-analysis on vitamin K administration have shown that intramuscular vitamin K1 is more effective than oral vitamin K in the prevention of late HDN(1,9,11). Single oral dose of 1 mg vitamin K is not effective, and the efficacy is increased with 3 doses of 2 mg (at birth, 1-2 weeks and 4 weeks), rather than 1 mg dose(11). Studies have shown that a daily dose of 25 µg of vitamin K1 given orally following an initial dose of 1 mg after birth may be as effective as intramuscular vitamin K(12). The recommendation of the American Academy of Pediatrics is to give vitamin K1 to all newborns as a single intramuscular dose of 0.5 to 1 mg(4). Since breast fed infants are at highest risk of late HDN, it is imperative that vitamin K prophylaxis be promoted. Warning bleeds like umbilical bleed (which may precede intracranial bleeds), epistaxis or skin bleeds should be taken seriously in any breast-fed infant. Our study highlights the morbidity and mortality associated with a potentially preventable condition. Vitamin K prophylaxis should be offered to all newborns who are exclusively breast-fed.