Article Text


Unexplained fever in neonates may be associated with hepatitis B vaccine


AIM To investigate whether hepatitis B vaccination has increased the number of cases of unexplained neonatal fever.

METHOD The files of all infants born from 1 January 1991 to 31 December 1992, in whom a diagnosis of “injected antibiotic” or “disease of temperature regulation” was recorded, were reviewed. Those who had unexplained fever of 38°C or higher during the first three days of life were divided into two groups: infants who did not receive the hepatitis B vaccine (1991) and infants who did (1992).

RESULTS In 1992 the incidence of unexplained fever in hepatitis B vaccinated neonates was significantly higher than in the 1991 group of pre-vaccination neonates (35 out of 5819 (0.6%) vs 14 out of 5010 neonates (0.28%) respectively, p=0.013).

CONCLUSIONS The increase in the number of cases of unexplained neonatal fever seems to be associated with the introduction of routine hepatitis B vaccination on the first day of life. The possibility that an excess number of neonates will undergo unnecessary procedures and treatment to diagnose unexplained fever justifies planning a controlled study to determine whether these preliminary findings point to a significant problem.

Statistics from

Adverse events have been reported in 1.0 to 7.3% of infants who received hepatitis vaccine,1 2 the most common in neonates being fever (37.5–39°C) with an incidence of 1.0 to 3.7%.2 Routine administration of hepatitis B vaccine to all neonates on the first day of life was introduced in Israel on 1 January 1992. This study aimed to determine whether an increased incidence of unexplained early neonatal fever occurred in the wake of vaccination.


We reviewed the charts of all full term infants (37 weeks of gestation or more and birthweight ⩾ 2.5 kg) born at the Chaim Sheba Medical Center, Israel, from 1 January 1991 to 31 December 1992, who had a discharge diagnosis of “temperature regulation disease” or who had received intravenous antibiotics as an inpatient. In our department temperature is measured using a digital rectal thermometer (IIVAC 281 model 811, San Diego, CA) in all instances of suspected fever.

“Unexplained” neonatal fever was defined as a temperature of ⩾ 38°C during the first three days of life in the absence of sepsis (positive blood and/or cerebrospinal fluid cultures), dehydration (loss of at least 10% of birthweight), maternal fever (⩾ 38°C at delivery), or respiratory distress (oxygen required for > 4 hours after birth). Chart entries concerning these specific diagnostic details for infants not vaccinated with hepatitis B, born during 1991, were compared with entries for infants born during 1992, who were vaccinated against hepatitis B. No other changes in nursery staff or routine policies had occurred during the two year period.


Twenty seven out of 5010 full term infants born in 1991 and 68 out of 5819 born in 1992 had had a diagnosis of neonatal fever. Initial review of the 1992 group yielded 18 infants with fever <38°C who were excluded from the final analysis (table 1). Further review of case files revealed identifiable causes of fever (sepsis, dehydration, maternal fever, respiratory distress) in 13 infants in the 1991 and 15 infants in the 1992 groups, respectively. These infants were excluded from analysis (table 1).

Table 1

Comparison of infants with neonatal fever before (1991) and after (1992) introduction of routine hepatitis B immunisation

Among the remaining infants with unexplained fever occurring within three days of birth, an increase of more than 100% was noted between 1991, when hepatitis B vaccine was not administered (0.28%), and 1992, when hepatitis was given routinely in the first day of life to all neonates (0.6%) (p=0.013) (table 1).

All charts of infants with unexplained fever were then reviewed for descriptive clinical characteristics (table 2). The only significant difference between the two groups was the longer duration of fever in the vaccinated group (p<0.05).


Hepatitis B vaccine is the first vaccine to be universally recommended for neonates.3 The rate of febrile reaction to it reportedly ranges from 0 to 7.3 hours.1 2 A lower rate of adverse events was reported in infants and children than in adults.1 In neonates the most common susbequent event was fever (37.5–39.0°C) and was the most common serious sign reported by the Vaccine Adverse Event Reporting System (VAERS).2 Their median reported time from vaccination to onset of fever was 1 day and mean maximum temperature was 38.9°C.

We reviewed the charts of infants with the diagnosis of temperature regulation disease, and of those who received intravenous antibiotics, to identify all infants with neonatal fever. The percentage of infants with unexplained fever during the first three days of life was significantly higher in 1992 (0.6%) when hepatitis B vaccine was given routinely on the first day of life to all neonates compared with 1991 (0.28%) when it was not given. The 0.32% difference is compatible with the 0 to 7.3% reported rate of febrile reaction to the vaccine.1 2 4-9 Notably, had we included infants with a temperature above 37.5°C, the significance would have been greater (p<0.001). Furthermore, infants born to mothers with maternal fever were excluded from our analysis even though some should probably have been classified as having unexplained neonatal fever.

The more than twofold increase in the percentage of infants with unexplained fever was not associated with the increased rate of enteroviral illnesses in 1992. The monthly distribution of cases of unexplained neonatal fever was stable, except for November 1992, when eight cases were noted. No other changes in our nursery staff or routine policies were introduced during 1992.

In the report by VAERS 24 neonates had severe neonatal events after hepatitis B vaccination; fever was the most common serious neonatal event and was reported in 13 (54%) neonates. The 13 neonates with fever reported by VAERS were admitted for a median of three days and 10 underwent evaluation for sepsis.2 In our study, all 35 neonates underwent a full sepsis evaluation, intravenous antibiotic treatment, and prolonged hospital stay.

Although Israel has a low incidence of hepatitis B carriers among the general Jewish Israeli population 0.5–0.6%,10 11 the Ministry of Health decided to provide active immunisation for every neonate on the first day of life, mainly because of the large scale immigration from countries in which the virus is hyperendemic. It was expected that since the rate of transmission of anti-hepatitis B antibodies from Israeli mothers to offspring is only 23%, there would be a good antibody response to an early first immunisation.10 11

In conclusion, we found that an increased incidence of unexplained neonatal fever, which resulted in evaluation for sepsis, administration of intravenous antibiotics, and prolonged hospital stay, may be associated with vaccination against hepatitis B on the first day of life. Although our data are significant, our numbers are small, therefore a larger controlled trial is justified to determine if the benefit conferred by universal vaccination of neonates against hepatitis B is outweighed by the risks and costs of unnecessary diagnostic procedures and treatments.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.