Averhoff F, Shapiro CN, Bell BP, Hyams I, Burd L, Deladisma A, Simard EP, Nalin D, Kuter B, Ward C, Lundberg M, Smith N, Margolis HS. Control of Hepatitis A Through Routine Vaccination of Children. JAMA. 2001;286(23):2968-2973. doi:10.1001/jama.286.23.2968
Author Affiliations: National Immunization Program (Drs Averhoff and Ms Deladisma) and Division of Viral Hepatitis, National Center for Infectious Diseases (Drs Shapiro, Bell, and Margolis and Mr Simard), Centers for Disease Control and Prevention, Atlanta, Ga; Butte County Department of Public Health, Oroville, Calif (Drs Ward and Lundberg and Ms Hyams); California Department of Health Services, Immunization Branch, Berkeley (Mr Burd and Dr Smith); and Merck Vaccine Division (Dr Nalin) and Merck Research Laboratories (Dr Kuter), West Point, Pa.
Context The impact of routine hepatitis A vaccination of children living in
large communities with elevated disease rates has not been evaluated.
Objective To determine the effect of routine vaccination of children on disease
incidence in a community with recurrent hepatitis A epidemics.
Design, Setting, and Participants Community-based demonstration project conducted from January 12, 1995,
through December 31, 2000, in Butte County, California, among children aged
2 to 17 years.
Intervention In 1995, vaccination was offered to children aged 2 to 12 years during
vaccination clinics conducted on 2 occasions 6 to 12 months apart at most
schools in the county. In 1996-2000, vaccine was distributed to community
health care clinicians, who vaccinated eligible children without charge. Vaccine
was also available at health department clinics, selected child care centers,
and other sites.
Main Outcome Measures Hepatitis A vaccination coverage, hepatitis A incidence, and vaccine
Results During the study period, 29 789 (66.2%) of an estimated 44 982
eligible children received at least 1 vaccine dose; 17 681 (39.3%) received
a second dose. The number of hepatitis A cases among the entire county population
declined 93.5% during the study period, from 57 cases in 1995 to 4 in 2000,
the lowest number of cases reported in the county since hepatitis A surveillance
began in 1966. The 2000 incidence rate of 1.9 per 100 000 population
was the lowest of any county in the state. Of the 245 cases reported during
the 6-year period, 40 (16.3%) occurred among children 17 years of age or younger,
of which 16 (40%) occurred in 1995 and only 1 in 2000. One of the 27 case
patients eligible for vaccination had been vaccinated, having received the
first dose 3 days before symptom onset. The estimated protective vaccine efficacy
was 98% (95% confidence interval, 86%-100%).
Conclusions In this population, hepatitis A vaccine was highly effective in preventing
disease among recipients. Childhood vaccination appears to have decreased
hepatitis A incidence among children and adults and controlled the disease
in a community with recurrent epidemics.
Hepatitis A continues to be one of the most frequently reported vaccine-preventable
diseases in the United States. Hepatitis A incidence displays a cyclic pattern,
and most disease occurs in the context of community-wide outbreaks during
which a large proportion of patients do not have a recognized risk factor.1- 6
Available data suggest that young children, frequently asymptomatic when infected,
play an important role in hepatitis A virus (HAV) transmission.7- 12
Until recently, immunoglobulin and improved hygiene were the only measures
available to prevent and control hepatitis A.13
Although immunoglobulin has been shown to be effective in preventing hepatitis
A among persons with a recognized exposure, it has limited effectiveness in
controlling the spread of HAV infection in the community because many people
are unaware of their exposure.3,14- 17
Hepatitis A vaccines, available in the United States since 1995, are
highly effective in preventing disease among immunized persons.18- 21
Routine vaccination of children living in small communities that had experienced
recurrent hepatitis A epidemics and in which most adults are immune has been
shown to be effective in helping to interrupt disease transmission.22,23 However, the impact of routine childhood
vaccination on disease incidence in large communities that experience recurrent
epidemics, where most of the population is thought to be susceptible to the
disease, has not been determined.5 We report
the results of routine hepatitis A vaccination of children in one such community.
Butte County, with an estimated year 2000 population of 207 158,
is located in the San Joaquin Valley of California. During an outbreak of
hepatitis A from 1985 to 1988, 495 cases (average annual incidence, 112.7
per 100 000 population) were reported, and between 1988 and 1992, the
incidence fluctuated between 20 and 45 per 100 000 population. The number
of reported cases increased to 128 (65.2 per 100 000 population) in 1993
and then to 230 (117.3 per 100 000 population) in 1994. The 1994 rate
was highest among children younger than 15 years (193.0 per 100 000 population).
From January 12, 1995, through December 31, 2000, hepatitis A vaccine
was offered free of charge to children residing in the county by means of
a demonstration project conducted in 2 phases. In 1995, the project focused
primarily on controlling the ongoing outbreak by vaccination of children under
an investigational new drug (IND) protocol before the vaccine was licensed
by the US Food and Drug Administration (FDA). From 1996 through 2000, following
FDA licensure of the vaccine, the project evaluated the effectiveness of routine
vaccination of children. The study was approved by the institutional review
board of the Centers for Disease Control and Prevention (CDC) and the Butte
County Board of Supervisors.
During the first phase, in 1995, vaccination was recommended and made
available to children aged 2 to 12 years. The vaccine was offered to eligible
children during vaccination clinics conducted on 2 occasions 6 to 12 months
apart at most schools in the county. School-aged children up to 12 years (or
in the seventh grade) were encouraged to receive the vaccine during the school
vaccination clinics. Preschool-aged children also could be vaccinated at the
school clinics. In addition, vaccine was offered to all eligible children
aged 2 to 12 years at routine vaccination clinics conducted at Butte County
Health Department (BCHD) sites.
Each child was given a vaccine adverse event report card, thermometer,
and instructions for parents on how to record temperature and adverse events.
Parents were asked to return the cards to the school or health department
and to report serious adverse events to the health department immediately
The second phase of the study, 1996-2000, focused on establishment of
sustained childhood hepatitis A vaccination in the county through community
health care practitioners. Vaccine was distributed to clinicians who agreed
to vaccinate eligible children in their practice without charge and provide
information on each vaccinated child to a registry maintained by the BCHD.
Vaccine was also made available at health department clinics, selected child
care centers, and other sites, such as the Department of Agriculture's Special
Supplemental Program for Women, Infants, and Children (WIC) clinics. The target
population increased annually to include those additional birth cohorts of
children who had reached their second birthday and children who were older
than 12 years. By the end of the study, children aged 2 to 17 years were eligible
to receive vaccine and were included in the analysis.
The vaccination registry maintained by the BCHD recorded date and place
of vaccination, demographic data, and adverse events. First- and second-dose
vaccination coverage was calculated using registry data and population estimates.
Population estimates, including race and ethnicity, were obtained from the
State of California, Department of Finance, Demographic Research Unit. The
estimated number of unvaccinated children was determined by subtracting the
number of children vaccinated (from the registry) from the estimated population.
These estimates were also used for calculating vaccine effectiveness.
Inactivated hepatitis A vaccine (VAQTA, Merck & Co Inc, West Point,
Pa) was administered as the pediatric formulation (25 U, approximately 25
ng of purified viral protein in 0.5-mL volume) on a 2-dose (0- and 6- to 12-month)
schedule. Because the first phase of the study was conducted under an IND
protocol, the written consent reflected that the vaccine had not been licensed
by the FDA. When the vaccine was licensed in 1996, a standard, simplified
consent form, similar to the vaccine information statement recommended by
the CDC, was used. The vaccine is licensed for persons aged 2 years or older.
Although hepatitis A reporting is required by California law, enhanced
surveillance was established in the county in 1995, coincident with implementation
of the vaccination program. Reporting was stimulated in sites most likely
to encounter cases and included weekly contact with emergency department and
walk-in clinic staff to ascertain the number of suspected cases or serologic
test results indicative of hepatitis A. In addition, laboratories that conducted
diagnostic testing for hepatitis A were contacted and asked to report all
positive test results that originated from Butte County. Surveillance was
further stimulated by mailing letters to primary care practitioners (including
pediatricians, family physicians, and general practitioners) that encouraged
reporting of suspected cases. In addition, information about the study was
presented at regularly scheduled continuing education meetings for physicians
A case patient was defined as a patient with a positive serologic test
result for IgM antibody to HAV (anti-HAV) or the clinical diagnosis of hepatitis
A without serologic confirmation. All suspected case patients were interviewed
by a study nurse (I.H.). A detailed standard case investigation form, the
Viral Hepatitis Surveillance Program form, was used to collect additional
demographic and risk factor information on reported cases. Case patients were
asked about receipt of hepatitis A vaccine, and the vaccination registry was
also searched for verification of receipt of vaccine.
To assess the impact of the vaccination program, age-specific hepatitis
A incidence in the county during the 6-year vaccination program was compared
with the incidence during 1990-1994, before the program. The mean of the 5
or 6 annual rates, using the estimated population denominator for the relevant
year for each rate, was calculated. In addition, statewide, age-specific hepatitis
A incidence was compared with that in Butte County for 1990-1994 and 1995-2000,
using surveillance data reported to the National Notifiable Diseases Surveillance
System (NNDSS) of the CDC.24 The NNDSS data
for 2000 are provisional.
Vaccine effectiveness was estimated by comparing the incidence of reported
hepatitis A among children who had received at least 1 dose of vaccine to
the incidence among unvaccinated children of the same age. Vaccine effectiveness
was calculated using the following formula: 1 − (observed attack rate
in vaccine recipients/observed attack rate in unvaccinated children). The
hepatitis A attack rate in vaccine recipients was calculated as follows: number
of reported cases among vaccinated children (ie, receiving ≥1 dose of hepatitis
A vaccine)/number of vaccinated children in the vaccine registry. The attack
rate for unvaccinated children was calculated as follows: number of reported
cases among unvaccinated children/estimated number of unvaccinated children
in the target age group. Vaccine effectiveness estimates were based on number
of children eligible at the end of the study period (ie, aged 2-17 years).
Data were analyzed using Epi Info software,
version 6.04b (CDC) and SAS statistical software, version 8e (SAS Institute
Inc, Cary, NC). The Taylor series 95% confidence interval (CI) for the estimate
of vaccine effectiveness was calculated.
From January 12, 1995, through December 31, 2000, of an estimated 44 982
children in the age groups eligible for hepatitis A vaccination, 29 789
(66.2%) received at least 1 dose of vaccine and 17 681 (39.3%) received
2 doses. First-dose vaccination coverage increased annually from 35.2% to
66.2%, and second dose coverage increased from 14.5% to 39.3% (Figure 1). Cumulative first dose vaccination coverage was similar
among children aged 5 to 10 years (69.0%, 11 352/16 460) and 11
to 17 years (67.8%, 14 047/20 730) and lower among children aged
2 to 4 years (56.3%, 4390/7792).
The number and ages of children vaccinated varied during the 2 phases
of the project. In 1995, when school-based vaccination was the primary strategy
used, 10 754 (35.2%) of the 30 575 children aged 2 to 12 years received
at least 1 dose of hepatitis A vaccine. Of these, 9805 (91.2%) were school-aged
(>5 years), and 10 205 (94.8%) received their first dose of vaccine during
school-based vaccination clinics. In 1996-2000 (the second phase of the program),
an additional 19 035 children received 1 or more vaccine doses; of these,
10 668 (56.0%) were aged 5 years or older and 8367 (44.0%) were 2 to
4 years. After 1995, most children (14 484 [76%]) were vaccinated in
The average annual hepatitis A incidence in Butte County during the
5 years before the demonstration project (1990-1994) was 47.9 per 100 000
population (range, 122.5-11.8 per 100 000 population) and declined by
56.8% to 20.7 per 100 000 population (range, 48.7-0.97 per 100 000
population) during 1995-2000 (Figure 2).
The age-specific incidence decreased among all age groups during 1995-2000
compared with that in 1990-1994, but the effect was most pronounced among
the younger age groups (Figure 3).
Incidence decreased 78.9% among children aged 17 years or younger (67.3 to
13.5 per 100 000 population) compared with 44.3% among persons older
than 17 years (41.3 to 23.0 per 100 000 population).
Of the 245 cases reported to the BCHD between 1995 and 2000, 243 (99.2%)
were serologically confirmed. Thirteen patients required hospitalization and
4 died (aged 39, 41, 56, and 79 years). During this period, the reported number
of cases declined 93.5%, from 57 in 1995 to 4 in 2000 (Figure 2). Of the 40 (16.3%) cases that occurred among children
17 years or younger during 1995-2000, 16 (40%) occurred in 1995 and only 1
occurred in 2000.
During 1995-2000, 60.8% (n = 149; cumulative incidence, 147.3 per 100 000
population) of cases occurred among males. Non-Hispanic whites accounted for
87.3% (n = 214; cumulative incidence, 123.2 per 100 000 population) of
the cases, whereas Native Americans (n = 15; cumulative incidence, 449.1 per
100 000 population), Hispanics (n = 10; cumulative incidence, 53.9 per
100 000 population), Asian and Pacific Islanders (n = 5; cumulative incidence,
60.8 per 100 000 population), and blacks (n = 1; cumulative incidence,
36.2 per 100 000 population) accounted for the remainder.
Recognized potential sources of infection reported by case patients
included contact with a hepatitis A case (n = 89, 36.3%) and illegal drug
use (n = 25, 10.2%). Antecedent international travel to a country with endemic
hepatitis A (n = 9), association with child care (n = 7), homosexual activity
(n = 5), and association with a common source outbreak of hepatitis A (n =
2) were each reported by less than 3% of case patients. For 108 case patients
(44.1%), no risk factor could be identified. All 25 case patients who identified
drug use as a risk factor were reported during 1996-1997.
During 1990-1997, annual hepatitis A incidence in Butte County was at
least equal to, and up to 5.7 times higher, than the overall California rate
(Figure 4). In 1998, the Butte County
hepatitis A rate fell below the overall California rate and stayed lower through
2000. In 2000, the Butte County rate (1.9 per 100 000 population) was
the lowest reported in any county in the state (data not shown). Compared
with the previous 5 years, age-specific incidence during the study period
declined 80% among children aged 17 years or younger in the county, compared
with 21% in the state (Figure 5).
During the study period, 27 hepatitis A cases occurred among children
eligible for the vaccination program, including 26 among the 15 193 unvaccinated
children and 1 in a 12-year-old boy who had received his first dose of vaccine
3 days before onset of illness. The estimated protective efficacy of 1 or
more doses of vaccine was 98% (95% CI, 86%-100%).
No serious adverse events were reported among the 29 789 children
who received 1 or more doses during the study period. Among the 5471 vaccine
doses for which report cards were available (accounting for 30.6% of prelicensure
vaccine doses administered), adverse reactions were reported on 1983 (36.2%),
were generally described as mild, and included injection site reactions, fever,
This 6-year project demonstrated the feasibility and effectiveness of
routine childhood hepatitis A vaccination and its impact on community-wide
rates of disease. Hepatitis A incidence declined to historic lows following
the introduction of hepatitis A vaccination, and since 1998, the number of
reported cases has been lower than in any year since hepatitis A surveillance
began in 1966. The decline appears to be sustained; only 2 cases were reported
to the CDC during the first half of 2001 (CDC, unpublished data, 2001). Furthermore,
the 2000 Butte County disease rate was the lowest of any county in California,
suggesting that community-wide childhood vaccination was responsible for a
sustained reduction in hepatitis A incidence in Butte County. Finally, no
serious adverse events were reported among the nearly 30 000 vaccine
recipients, supporting the safety of hepatitis A vaccination.
Although the overall purpose of the demonstration project was to evaluate
whether childhood hepatitis A vaccination could be sustained, the purpose
of the first phase was to evaluate whether rapid vaccination of multiple age
cohorts could control the ongoing outbreak.25
However, the impact of the first phase is uncertain. As has been the experience
in other large communities, the vaccination program in Butte County was not
begun until well into the second year of the outbreak, when hepatitis A incidence
may have already been waning.2,5
Other California counties with similar historical patterns of hepatitis A
incidence and without vaccination programs also reported declines in incidence
during this time. However, vaccination may have shortened the epidemic, which
lasted 2 years compared with the previous 3-year epidemic. In addition, sustained
community-wide vaccination may prevent future outbreaks, as has been observed
in smaller communities.5,23
Before the demonstration project, children had the highest disease incidence
in Butte County, and incidence declined to levels below that of adults following
the implementation of the vaccination program. Of the cases that were reported
during the vaccination program, including during the peak year of 1997, a
substantial proportion occurred among adult users of illegal drugs; this appears
to have mitigated the impact of childhood vaccination on overall disease incidence.
Other studies have suggested that HAV transmission can be sustained among
adult users of illegal drugs with little transmission to or from children.26 However, the long-term impact of vaccination of children
and adolescents on transmission among adult drug users is unknown.
This study has some potential limitations. Because the incidence of
hepatitis A varies from year to year, it is not possible to know what the
incidence would have been without the vaccination program and we did not have
the benefit of a "control" community. However, we compared hepatitis A incidence
in Butte County to that in the rest of California and also before and after
implementation of the vaccination program, with Butte County serving as its
own control. Both analyses suggested the effectiveness of the vaccination
Alternative explanations for the observed decline in hepatitis A incidence
could be postulated. "Surveillance fatigue" could have occurred if cases continued
to occur but were not reported to the surveillance system. However, hepatitis
A reporting did not begin with the advent of the vaccination program but has
been ongoing in Butte County and the rest of the state for more than 20 years.
Although it is possible that the initial enthusiasm for the enhanced surveillance
implemented during the program waned, it is extremely unlikely that passive
routine surveillance would be affected. Furthermore, physicians were not the
only source of reports; positive reports were also received from laboratories.
Depletion of the pool of susceptible persons in the population would also
lead to a decline in incidence. However, although the age-specific prevalence
of immunity in Butte County is unknown, serologic surveys indicate that most
of the US population remains susceptible to HAV infection.5
Although reasons for the spontaneous waning of community-wide outbreaks are
not well understood, taking into consideration the relatively low baseline
prevalence of immunity and the number of cases typically reported during outbreaks,
a considerable proportion of the population remains susceptible to HAV infection
even after outbreaks.
We calculated vaccine effectiveness in preventing symptomatic, clinically
recognizable hepatitis A, not asymptomatic HAV infection. This allows for
comparison between our results and those of published efficacy studies.18,19 We did not verify the vaccination
status of children not listed in the vaccination registry. However, because
of the widespread availability of the study vaccine, we believe that very
few, if any, children in Butte County received hepatitis A vaccine outside
Childhood hepatitis A vaccination in Butte County posed some unique
challenges, including the lack of a routinely scheduled immunization visit
for children aged 2 years or older and the need to initially administer vaccine
under an IND protocol, which necessitated more involved consent procedures.
School-based vaccination clinics were used at the outset as the primary vaccination
sites and achieved only modest vaccination coverage rates in schoolchildren
and low rates in preschool children. When clinician-based vaccination was
introduced as the primary vaccination strategy for all children following
vaccine licensure, vaccination coverage increased, indicating that it is possible
to achieve fairly high levels of hepatitis A vaccination coverage among children.
To our knowledge, no other community of this size in the United States
has delivered hepatitis A vaccine to such a large proportion of its children
and sustained a vaccination program for 6 years. The Butte County experience
suggests that, over time, routine vaccination of children can reduce overall
disease rates in the community. This previews the potential impact of routine
childhood hepatitis A vaccination, as recently recommended by the Advisory
Committee on Immunization Practices for areas of the United States with consistently
high hepatitis A infection rates.5