Significant reductions in varicella incidence were reported from 1995 to 2000 in the varicella active surveillance sites of Antelope Valley (AV), California, and West Philadelphia (WP), Pennsylvania. We examined incidence rates, median age, and vaccination status of case patients for 1995–2005. Coverage data were from the National Immunization Survey. By 2005, coverage among children 19–35 months of age reached 92% (AV) and 94% (WP); 57% and 64% of case patients in AV and WP, respectively, were vaccinated; and varicella incidence declined by 89.8% in AV and 90.4% in WP. Incidence declined in all age groups, especially among children <10 years of age in both sites and among adolescents 10–14 years of age in WP. In AV, since 2000, the incidence among adolescents 10–14 and 15–19 years of age increased. Implementation of school requirements through 10th grade in WP may explain the differences in the decline in incidence among adolescents. Continued surveillance will be important to monitor the impact that the 2-dose vaccine policy in children has on varicella epidemiology.
The varicella vaccine was licensed in United States in 1995. At the start of the national varicella vaccination program, the Advisory Committee on Immunization Practices (ACIP) recommended 1 dose of varicella vaccine to be administered routinely to children at 12–18 months of age, to older susceptible children 19 months of age through 12 years of age, and to designated highrisk groups . Substantial declines in the incidence of varicella and related hospitalizations were documented within the first 5 years of the vaccination program . As the varicella vaccination program matures, with higher levels of and more expanded vaccine coverage, continued monitoring of varicella epidemiology and assessment of the need for further enhancements of the program are important. For example, recent data indicate that varicella outbreaks are occurring among highly vaccinated populations of schoolchildren [3, 4]. Reviewing all relevant information, in June 2006, the ACIP recommended a second dose of varicella vaccine as part of routine childhood immunization . Active surveillance for varicella has been ongoing in 2 project sites since 1995. We used data from these sites to describe the changing epidemiology of varicella during the 11 years of implementation of the 1-dose vaccination program (1995–2005).
The Varicella Active Surveillance Project (VASP) was established in 1995 as a cooperative agreement between the Centers for Disease Control and Prevention (CDC) and 2 health departments: the Los Angeles County Department of Health Services and the Philadelphia Department of Public Health. Respective surveillance areas within the jurisdiction of these Health Departments are Antelope Valley (AV) and West Philadelphia (WP). The methods have been described elsewhere . Briefly, AV is a geographically defined health services district in a valley ∼40 miles outside of the city of Los Angeles. It had a population of ∼350,000 in 2004, of whom 80% were white; 46% of the white population described themselves as being of Hispanic ethnicity. WP is a densely populated inner-city area in Philadelphia and had a population of ∼280,000 in 2004, of whom 76% were African American. We used data from AV and WP reported during 1995–2005.
Case definition. For surveillance purposes, varicella is defined as an illness with acute onset of diffuse (generalized) maculopapulovesicular rash without other apparent cause . Case patients were considered to be vaccinated if they had received varicella vaccine, regardless of the interval between vaccination and disease onset.
Data collection. In 2005, there were 306 reporting sites in AV and 325 reporting sites in WP. In both areas, reporting sites were asked to submit case reports biweekly, even when no cases occurred; if no reports were received, reporting sites were contacted. Surveillance staff conduct case investigations to collect data on demographics, clinical details of illness, and vaccination status. Vaccination status was collected from parental reports (using the child's vaccination record when available). From 1995 to 1999, for case patients reporting that they were vaccinated, vaccination status was verified from school or provider medical records or a shot card that documented the date of vaccination. From 2000 to 2005, vaccination status was verified from school or provider records for all cases. Interviews were conducted by telephone or home visit for households without telephones or where telephone contact is not successful.
National Immunization Survey (NIS) results among children 19–35 months of age from Los Angeles County and Philadelphia County were used to estimate vaccination coverage in AV and WP, respectively, during 1 January 1997 to 31 December 2005 .
Data analysis. SAS (version 9.1; SAS Institute) was used for data analysis. We calculated annual and age-specific incidence rates of reported varicella per 1000 population, using population estimates from the US Census Bureau. We used Poisson regression to estimate the annual trend in age-specific incidence rates over the periods 1995 (introduction of vaccine), 2000 (last published), and 2005, the last year for which complete surveillance data were available.
In AV, from 1 January 1995 to 31 December 2005, a total of 13,315 varicella cases were reported. Varicella incidence declined by 89.8%, from 10.3 cases/1000 population (2934 cases) in 1995 to 1.1 cases/1000 population (362 cases) in 2005 (P < .001). However, during this period, fluctuations in cases and incidence occurred (figure 1A and table 1). In WP, between 1 January 1995 and 31 December 2005, a total of 3921 cases were reported. Varicella incidence declined by 90.4%, from 4.1 cases/1000 population (1197 cases) in 1995 to 0.4 cases/1000 population (108 cases) in 2005 (P < .001). The lowest number of reported cases was in 2004 (incidence, 0.3 case/1000 population; 92 cases) (figure 1B and table 1).
Age-specific incidence. In both areas, age-specific incidence rates for all age groups in 2005 were significantly lower than the rates reported in 1995 (P < .001). The declines in incidence were greatest (90%–95%) among children 1–9 years of age in both sites and among adolescents 10–14 years of age in WP. In other age groups, declines in incidence varied from 57.3% (adults in WP) to 87.8% (adolescents 15–19 years of age in WP) (table 1).
In AV, during the 11-year study period, there were some fluctuations in incidence. In 2000, an increase in varicella incidence occurred in comparison with 1999, which was due to an increase in incidence among children <5 years of age and children 5–9 years of age (table 1). Another increase was observed in 2004, which was due to increases in varicella incidence in all age groups except children 1–4 years of age. The highest increase occurred among children 5–9 and 10–14 years of age. Among those 10–14 years of age and 15–19 years of age, incidence fluctuated over the last 5–6 years, with increases in incidence of 51.9% and 23.6%, respectively, between 2000 and 2005. In contrast, in WP, varicella incidence declined progressively in all age groups from 1995 to 2004. In 2005, compared with 2004, varicella incidence increased primarily among adolescents 10–14 years of age and adults. Moreover, the incidence increased among adults from 2000 to 2005.
Vaccination status of case patients. Information on vaccination status was available for 97% of the case patients ⩾1 year of age. Through the study period, for age groups eligible for vaccination, the proportion of vaccinated case patients decreased with increasing age, and WP had a higher proportion of case patients vaccinated than did AV. In 2005, the proportion of vaccinated case patients ⩾1 year of age was 64% in WP and 57% in AV; 87% (AV) and 97% (WP) of case patients 5–9 years of age of were vaccinated, and 38% (AV) and 45% (WP) of case patients 10–14 years of age were vaccinated. Among older adolescents and adults in AV and WP, respectively, 31% and 17% of case patients 15–19 years of age and 7% and 9% of case patients ⩾20 years of age were vaccinated (figure 2).
Over the years, a shift in median age of case patients was observed among both vaccinated and unvaccinated case patients (figure 3). For vaccinated case patients, between 1995 and 2005, the median age increased from 5 to 8 years in AV and from 3 to 6 years in WP. For unvaccinated case patients, the increase was more pronounced: from 5 to 13 years in AV and from 6 to 19 years in WP.
Vaccination coverage and school entry requirements. In Los Angeles County, estimated varicella vaccination coverage among children 19–35 months of age increased from 40% in 1997 to 92% in 2005 (figure 1A). Similarly, the coverage rate in Philadelphia increased from 41% in 1997 to 94% in 2005 (figure 1B). During this period, varicella vaccination requirements for child care and school entry were implemented in both areas. In AV, starting in 2001, children attending child care centers and kindergarten (K) were required to have received varicella vaccine unless they had other evidence of immunity. By school year 2005–2006, children in grades K-4 were covered by this requirement. In WP, entry requirements for varicella vaccination were implemented for child care in 1995, for school entry for kindergarten students in 2000 and for sixthgrade students in 2001. By school year 2005–2006, children in grades K—10 were covered by the school entry requirements.
The varicella vaccination program has resulted in substantial reductions in varicella incidence since its implementation 11 years ago. In varicella active surveillance sites, compared with 1995, by 2005, overall varicella incidence declined by ∼90%, with declines in various age groups ranging from 57% to 95%. This decline in incidence has occurred in the context of rapidly increasing vaccination coverage among young children, supplemented with catch-up vaccination among older children facilitated through school entry vaccination requirements. By 2005, the majority of case patients were vaccinated, consistent with the high vaccine coverage in the sites and with the effectiveness of varicella vaccine being ∼80%–85%. The beneficial effects of high population immunity are apparent in the sustained decline in incidence in infants (ineligible for vaccination) and adults where rates of vaccination appear to be low.
Our results confirm and extend findings from our previous study and from other studies that have reported reductions in varicella-related morbidity and mortality since implementation of the vaccination program [2, 8–12]. In 4 states (Michigan, Illinois, Texas, and West Virginia) with passive reporting of the aggregate number of cases, varicella disease incidence in 2001 declined by 67%–82%, compared with the average incidence during 1990–1994 . According to the Behavioral Risk Factor Surveillance System in Massachusetts, between 1998 and 2003, overall varicella incidence had declined by 79%, from 16.5 to 3.5 cases/1000 population . For deaths in which varicella was listed as the underlying cause, rates of varicella mortality decreased by 67%, from an annual average of 105 deaths during 1990–1994 to 35 deaths during 1999–2001. The decline in mortality was 92% among children 1–4 years of age and ⩾74% for all age groups <50 years . From 1994–1995 to 2002, the reduction in varicella hospitalizations and ambulatory visits was 88% and 59%, respectively . The varicella vaccination program has also resulted in an estimated reduction of $62.8 million in direct inpatient and outpatient medical expenditures .
Differences in the epidemiology of varicella have been apparent in the 2 surveillance sites, especially since 1999. Despite similar levels of vaccination coverage among children 19–35 months of age, as measured by the National Immunization Survey, the higher proportion of cases in vaccine recipients among children 1–14 years of age in WP point to higher vaccine coverage and more complete implementation of the vaccination program throughout this age range, compared with AV.
The increase in vaccination coverage resulted in a reduction in disease, especially among preschool-aged children, and shifted disease incidence to the school-aged group. Therefore, vaccination of school-aged children through implementation of school entry laws is important in raising immunity among children who missed vaccination and disease at a younger age . The impact of school entry requirements can be better seen in WP, because child care and school entry laws were implemented earlier and, by school year 2005–2006, covered a much wider age range (6 additional grades in upper elementary and middle school and some grades in high school). This resulted in an ∼70% decline in incidence from 1999 to 2005 among children 10–14 years of age, who were covered by middle-school entry requirements (in contrast to AV), as well as in a greater increase in the age of unvaccinated case patients and, therefore, in a higher median age (19 years) of unvaccinated case patients in WP in 2005. School entry requirements are also important in capturing children that come into the population from areas of lower vaccine coverage.
Varicella disease transmission continues to occur, although at a very low rate, even in areas like WP, which has high vaccination coverage rates among young children and expanded coverage among school students. Because varicella vaccine effectiveness is ∼80%–85%, even at high levels of vaccine coverage, it might not provide sufficient levels of population immunity to prevent spread of disease [14–23]. Without implementation of catch-up vaccination and administration of the second dose of varicella vaccine, in low-incidence areas there is likely to be accumulation of susceptible children and young adults, which has implications for the future. Numbers of unvaccinated persons, as well as 15%–20% of vaccinated persons who are completely or partially susceptible to varicella, may accumulate rapidly. In the future, outbreaks might be reported in age groups even older than we see today. According to national seroprevalence data from prevaccine era, >95% of persons acquired varicella before 20 years of age, and <2% of adults ⩾20 years of age were susceptible to varicella .
Considering data from VASP sites, as well as from other reporting sites in the United States, in 2005 and 2006 the ACIP expanded its recommendations to a routine 2-dose vaccination schedule for all children 4–6 years of age and catch-up seconddose vaccination for older children, adolescents, and adults who had received only 1 dose . States are encouraged to gradually implement catch-up vaccination through school entry requirements, including middle school, high school, and college entry requirements.
The following limitations should be considered when interpreting our data. Vaccination coverage data from age groups >3 years of age are not available; school entry requirements are used as a proxy to estimate coverage among older children. In WP, incidence rates were underestimated during 1995–1999 because they did not include all reporting sites. However, this would bias toward a lower impact of the vaccination program when comparisons are made from 1995 through 2000, when all sites were included in the surveillance system. Laboratory confirmation of cases has been limited and, because varicella in vaccinated persons is atypical, some cases may be missed; alternatively, other nonvaricella cases that have atypical presentation may be diagnosed as varicella. In 2005, completeness of reporting was estimated to be 74% in AV and 75% in WP. Small numbers of cases in some age groups made rates in these groups unstable, and comparisons across years in these age groups (e.g., adolescents 15–19 years of age in both sites and adults in WP) could have been affected substantially by slight changes in the number of reported cases.
States continue to establish and improve case-based varicella surveillance as recommended by the Council of State and Territorial Epidemiologists [6, 24]. At this stage of the vaccination program, collecting data on age, vaccination status, and severity of each case, outbreaks, and transmission settings is essential to monitor the changing epidemiology of varicella. In the mean time, VASP continues to play a significant role in monitoring the varicella vaccination program with population-wide and case-based data starting from the prevaccine era. Implementation and the impact of the expanded recommendations by the ACIP, particularly the routine 2 doses in children, on varicella epidemiology will continue to be monitored through surveillance.
We thank the health care providers, school nurses, child care center directors, and other professionals and community members who reported varicella cases in Antelope Valley and West Philadelphia to the Varicella Active Surveillance Project.
Supplement sponsorship. This article was published as part of a supplement entitled “Varicella Vaccine in the United States: A Decade of Prevention and the Way Forward,” sponsored by the Research Foundation for Microbial Diseases of Osaka University, GlaxoSmithKline Biologicals, the Sabin Vaccine Institute, the Centers for Disease Control and Prevention, and the March of Dimes.