Objective. During December 2002 to January 2003, a varicella outbreak occurred in an elementary school in Maine. Just 1 month before detecting the outbreak, Maine implemented varicella vaccine requirements for child care but did not require vaccination for school entry. We investigated this outbreak to examine reasons for its occurrence, including vaccine failure.
Methods. A self-administered questionnaire was sent to all students' parents to determine student disease status, medical conditions, and vaccination status, which was further confirmed by review of medical records. Parental reporting of chickenpox/varicella that occurred since September 1, 2002, in a student who attended the school was used to define a case. Parents of cases were interviewed by telephone about disease characteristics. Disease severity was classified on the basis of the number of skin lesions and the occurrence of complications. Vaccine effectiveness was calculated by comparing varicella attack rates for any disease, for moderate to severe disease, and for severe disease among vaccinated and unvaccinated students.
Results. We obtained complete information for 296 (81%) of 364 students. Varicella vaccine coverage was 74% overall and decreased by grade, from 90% in kindergarten to 60% in third grade. Attack rates increased significantly from 14% in kindergarten to 37% in third grade. Of the 53 varicella cases, 36 (68%) were unvaccinated, 12 (22%) were vaccinated, and 5 (10%) had previous disease history. Vaccine effectiveness was 89% (95% confidence interval [CI]: 79–94%) against disease of any severity, 96% (95% CI: 88–99%) against moderate to severe disease, and 100% (95% CI: undefined) against severe disease. Twenty-two percent of unvaccinated students had severe disease and 1 was hospitalized for a skin infection, whereas none of the vaccinated cases reported severe disease.
Conclusion. This outbreak was attributable primarily to failure to vaccinate, especially among children in grades 1 through 3. Catch-up vaccination of susceptible older children and adolescents is especially important to prevent accumulation of susceptibility in these groups, in which the natural disease is more severe. School entry requirements will contribute to a more rapid implementation of the existing recommendations for vaccination.
- breakthrough varicella
- catch-up vaccination
- school requirements
- vaccine effectiveness
- vaccine failure
Varicella vaccine is recommended for routine immunization of children aged 12 to 18 months and for older susceptible children and adults.1–4 In the United States, before the implementation of the varicella vaccination program, children bore most of the burden of disease, with >90% of cases, two thirds of hospitalizations, and almost half of deaths.5–7 Since the introduction of the vaccine in 1995, varicella vaccine coverage has increased continuously, and there has been a dramatic reduction in disease incidence in all age groups.8,9 Currently, estimates of vaccine coverage are available only for preschool children (aged 19–35 months) using results from the National Immunization Survey; no assessment of coverage is done for older children and adolescents. In 2002, state-specific varicella vaccine coverage in the United States among children aged 19 to 35 months was high (81%) but varied greatly by state (from 59% to 91%, with 73% in Maine).10 However, some children escape both disease and vaccination, and with the decrease in varicella-zoster virus (VZV) circulation, they may enter adulthood without immunity to varicella. Because the likelihood of severe varicella infection increases with age, these susceptible children need to be targeted for catch-up vaccination. The American Academy of Pediatrics (AAP) recommends that pediatricians seek a reliable history of varicella at every childhood visit and children aged ≥12 months and who do not have a history of disease or vaccination should be immunized.4 In 1996, the Advisory Committee on Immunization Practices (ACIP) recommended that through age 13 years all children without a history of varicella disease or vaccination be vaccinated.2
Since implementation of the varicella vaccination program, a number of postlicensure investigations and studies11–19 have evaluated vaccine effectiveness and identified potential risk factors for vaccine failure. The results of some of these investigations raised concerns about the performance of varicella vaccine.12,18–20 One of them, that reported a low vaccine effectiveness (44%),12 was published at the time the Maine Bureau of Health (MBoH) was receiving reports about outbreaks of varicella in rural schools in central Maine with cases among vaccinated students. Officials and the medical community were particularly concerned that these outbreaks may also be attributable to low vaccine effectiveness. The Centers for Disease Control and Prevention (CDC) was invited to assist with the investigation of the outbreak in 1 elementary school (school A) that had a high number of cases and a high proportion of vaccinated students among them. This investigation offered the opportunity to describe the epidemiologic features of the outbreak, assess vaccine coverage among school-aged children, compare disease presentation among vaccinated and unvaccinated students, evaluate vaccine effectiveness, and examine risk factors for vaccine failure.
School A is a public elementary school with grades K through 3 located in a rural community (population 7400) in central Maine. Classes do not regularly mix for lunch or other activities, but most students ride buses together to and from school.
Two standard questionnaires were used for data collection. Parents of all students in the school were sent a self-administered questionnaire to determine demographic characteristics, students' disease status, underlying medical conditions, vaccination history, and contact information for the student's health care provider. Parents of varicella cases were interviewed by telephone about sources of exposure outside the school and clinical characteristics of the disease (eg, date of rash onset, duration of rash, clinical presentation, severity, complications).
Because varicella is easily recognized by laypeople, we relied on parental reporting of disease and defined a varicella case as any student whose parents responded “yes” to both questions, “Has your child ever had chickenpox disease?” and, “Did the case of chickenpox occur since September 1, 2002?” September 1 was 2 maximum incubation periods before the date of rash onset of the first reported case. By using this as a reference date, we hoped to include any case that might have occurred before the first reported case. We classified cases according to vaccination and disease status. Cases were considered vaccinated (breakthrough) when they had been vaccinated >42 days before the onset of symptoms. To confirm the outbreak, polymerase chain reaction (PCR) analysis of clinical specimens collected from 4 reported cases was performed at the National VZV Laboratory at CDC.
The severity of varicella was determined clinically by the reported number of skin lesions and occurrence of complications or hospitalizations. Disease was classified as mild when there were <50 lesions, moderate when there were 50 to 500 lesions, and severe when there were >500 lesions or disease of any severity that resulted in complications (eg, skin/soft tissue infections, lower respiratory infections, fluid/electrolyte disturbances, encephalitis) or hospitalization.
Vaccination Status Ascertainment
We confirmed the vaccination status for all students using 1 of 3 methods. First, we checked Immpact, the state-wide Maine computerized vaccination registry. For students with no record of varicella vaccination in Immpact, we reviewed providers' immunization records. Finally, if the provider did not have a record of varicella vaccination, then school immunization records were used. Vaccine coverage was defined as the proportion of students who did not have a history of varicella and had received the vaccine before September 1. An unvaccinated student was defined as any student without a history of disease and without documentation of vaccination before September 1, 2002.
Vaccine Effectiveness and Effectiveness of Varicella Disease History
Vaccine effectiveness (VE) represents the percentage reduction in disease incidence attributable to vaccination. We calculated the attack rates among unvaccinated students (ARU) and among vaccinated students (ARV) and then assessed VE using the following equation: VE% = [(ARU − ARV)/ARU] × 100.21 We report VE for prevention of disease of any severity, for prevention of moderate to severe disease (where mild cases are classified with noncases), and for prevention of severe disease (where mild and moderate cases are classified with noncases). The following students were excluded from the calculation of attack rates and VE: those with varicella disease history, uncertain disease history, or unconfirmed vaccination status; those who were vaccinated during the outbreak; those with both disease and vaccination history; and those whose parents did not consent for provider record review or refused the telephone interview. We calculated effectiveness of a positive disease history in preventing varicella using the following equation: effectiveness of varicella disease history = [(ARU − AR among students with disease history)/ARU] × 100. For effectiveness of varicella history, students with history of both disease and vaccination (n = 5) were classified as having disease history only.
Risk Factors for Vaccine Failure
We examined the following factors identified in other studies as risk factors for vaccine failure: age at vaccination, time since vaccination, and history of asthma or eczema. To compare our results with previous reports, we analyzed age at vaccination as a continuous variable and then categorized it as age <14 months and ≥14 months, and ≤15 months and >15 months. Similarly, time since vaccination was analyzed as a continuous variable and then categorized as <3 years and ≥3 years before the start of the outbreak (the date of onset of the first case). To identify students with a history of asthma or eczema, we reviewed providers' medical charts of vaccinated students and varicella cases.
Data were entered into Epi Info (version 6.04; CDC, Atlanta, GA) and analyzed with SAS version 8 (SAS Institute, Cary, NC). For continuous variables, groups were compared using Wilcoxon rank-sum test. For categorical variables, risk ratios (RRs) were assessed using univariate analysis and results are presented with 95% confidence intervals (CIs). For proportions, Fisher exact test and χ2 test for trend were used. All P values were calculated by a 2-sided test.
Between September 1, 2002, and January 7, 2003 (the date of onset of the last case), 364 students attended school A; 312 of them (86%) returned a completed questionnaire. The median age of responders was 7 years (range: 5–9), and 154 (49%) were male. We obtained complete data on disease and vaccination status for 296 (81%) students; 146 were vaccinated, 51 were unvaccinated, and 99 had previous disease history (of these, 5 had history of both disease and vaccination).
The outbreak lasted 12 weeks, from October 12, 2002, to January 7, 2003; it peaked on December 6 and 7, with 7 and 8 cases per day, respectively (Fig 1). A total of 55 varicella cases were identified; complete information was obtained from 53 (96%) cases: 36 (68%) were unvaccinated, 12 (22%) were vaccinated, and 5 (10%) had previous disease history. The median age of cases was 7 years (range: 5–8); 31 (59%) were male. Cases occurred in all grades and in 20 of 21 classes. We did not identify a single index case; 5 cases (3 unvaccinated and 2 vaccinated, grades K–2) had rash onset within the first 10 days of the outbreak (the minimum incubation period). These 5 students all were exposed to their siblings who had varicella and attended other schools in town.
At the start of the outbreak, vaccine coverage in the school was 74% (146 of 197) and decreased by grade, from 90% in kindergarten to 60% in third grade. The varicella attack rates increased significantly by grade, from 14% in kindergarten to 37% in third grade (P = .008 for linear trend; Fig 2). The evidence of immunity (disease or vaccination) was 91% in kindergarten, 78% in first grade, and 82% in second and third grades; thus, the proportion of unvaccinated students in grades 1 to 3 was twice as high as that among kindergartners. Younger students were more likely to acquire immunity through vaccination, whereas older students were more likely to acquire it through disease. The proportion of students who were vaccinated ranged from 79% among kindergartners to 27% among third graders, whereas the proportion of students with disease history ranged from 12% among kindergartners to 55% among third graders (Fig 3).
A health care provider diagnosed varicella for 20 (38%) of the cases, either at an office visit or by telephone consultation. Specimens for laboratory confirmation were collected from 4 cases, 2 vaccinated and 2 unvaccinated. In all cases, wild-type VZV was identified.
Compared with unvaccinated cases, vaccinated cases were significantly less likely to have moderate or severe disease (25% vs 78%). None of the vaccinated cases had severe disease compared with 8 (22%) unvaccinated cases. Vaccinated cases also had a significantly shorter duration of rash (5 days vs 7 days) and missed fewer days from school (3 vs 5). Two unvaccinated cases had bacterial superinfection of their skin lesions, 1 of whom required hospitalization for 4 days (Table 1).
VE and Effectiveness of Varicella Disease History
VE against disease of any severity was 89% (95% CI: 79–94%), corresponding to attack rates of 8% (12 of 146) among vaccinated students and 71% (36 of 51) among unvaccinated students. VE against moderate to severe disease was 96% (95% CI: 88–99%), corresponding to attack rates of 2% (3 of 146) among vaccinated students and 55% (28 of 51) among unvaccinated students. VE against severe disease was 100% (95% CI: undefined), corresponding to attack rates of 0% (0 of 146) among vaccinated students and 16% (8 of 51) among unvaccinated students. Among students with varicella disease history, the attack rate was 5% (5 of 99). Thus, varicella disease history was highly effective (93%; 95% CI: 83–97%) in preventing a second infection.
Risk Factors for Vaccine Failure
None of the risk factors examined (age at vaccination, time since vaccination, and history of asthma or eczema) was associated with vaccine failure. However, although it did not achieve statistical significance, history of eczema was associated with an increased risk for vaccine failure (RR: 4.3; 95% CI: 0.8–23.5).
This outbreak was attributable primarily to failure to vaccinate. The high proportion of unvaccinated cases (68%) highlights that, despite the recommendations of both the AAP and the ACIP, eligible children remain unvaccinated. Although children should routinely receive varicella vaccine at age 12 to 18 months, all susceptible older children and adolescents should also be vaccinated.1–4 As exposures to varicella continue to decline as a result of the success of the US vaccination program and decline in varicella disease, catch-up vaccination of susceptible older children and adolescents is especially important to prevent increasing susceptibility and outbreaks in these groups.
Varicella may result in serious complications and death. One of the 36 unvaccinated cases in this outbreak required hospitalization, and another case required a physician visit and antibiotic treatment, both for bacterial superinfection of the skin lesions. Both children were sick for 7 days and missed 10 and 5 days of school, respectively, as a result of their illness. In the 5 to 7 years preceding implementation of the varicella vaccination program in the United States, on average, 48 (46%) of all varicella-related deaths and 7200 (68%) of all varicella-related hospitalizations occurred in children and adolescents aged <20 years every year.6,7 These severe outcomes, which now are preventable by a safe and effective vaccine, underscore the importance of adhering to current varicella vaccination recommendations.
In this outbreak setting, with the opportunity for multiple exposures to unvaccinated cases that are highly infectious, varicella vaccine was very effective, preventing 100% of severe disease, almost 100% of moderate or severe disease, and nearly 90% of disease of any severity. This is reassuring, especially because mathematical models have shown that VE tends to be underestimated in outbreak investigations.22 Our findings support previous reports11–19,23–27 that the vaccine provides excellent protection against serious varicella disease and, most commonly, 70% to 90% protection against disease of any severity. With increasing vaccine coverage in the population, the number of varicella cases will decrease and the proportion of cases who have been vaccinated will increase28; this is a function of both VE and vaccine coverage.
To help with the implementation of catch-up vaccination, varicella vaccination requirements for child care and school entry were recommended by the ACIP in 1999.3 In 2000, the AAP encouraged pediatricians to support public health officials in the development and implementation of these requirements.4 In the United States, school requirements have proved to be an extremely effective strategy for achieving and maintaining high vaccine coverage among school-aged children.29 However, the majority of states, including Maine, allow religious and/or philosophical exemptions that may be invoked by parents who choose not to vaccinate their children. MBoH indicates that this is an increasing trend in Maine (MBoH, unpublished data). This emphasizes the critical role that practicing pediatricians have to convey useful information to parents, enabling them to understand not only individual benefits and risks of vaccines but also societal risks and benefits. Maine implemented varicella vaccination requirements for child care in November 2002, 1 month before detecting the outbreak, and planned progressive implementation of school requirements starting with kindergarten in September 2003. In the United States, by September 2003, 36 states had implemented child care entry requirements and 34 states had implemented school entry requirements for varicella vaccine (CDC, unpublished data, 2003). However, these requirements do not cover cohorts of children who were enrolled in school before institution of the requirement, who are less likely to be vaccinated with increasing grades. To ensure high immunity in all groups and prevent outbreaks among older children and adolescents in the future, it is important for states to implement a policy that requires evidence of immunity for children who enter middle and/or high school.3 By September 2003, only 11 states had implemented middle or high school entry requirements for varicella vaccine.
In our investigation, we found no statistically significant risk factors for breakthrough varicella. However, among vaccinated cases, 1 of 3 students with a history of eczema and 11 of 142 students without such a history developed breakthrough varicella (RR: 4.3; 95% CI: 0.8–23.5). Although this suggests that breakthrough may be more common among students with eczema, the numbers are too small to draw a general conclusion. The role of history of eczema as a risk factor for vaccine failure needs to be examined; whether eczema is a marker for poorer immune response to vaccination as a result of either the disease itself or steroid therapy merits additional investigation. Only 1 study controlled simultaneously for the effect of multiple risk factors and found an increased risk for breakthrough varicella in the 3 months immediately after prescription of oral steroids.30
As the incidence of varicella continues to decline, laboratory confirmation of varicella will become more important. Breakthrough cases usually have mild or atypical clinical presentations. This poses particular challenges for clinical diagnosis and emphasizes the need for laboratory confirmation. Furthermore, in many situations, clinical specimens are difficult to obtain from these patients who present with a maculopapular rash only. Mild breakthrough cases may not be diagnosed if they do not present to their physician or are not recognized by school authorities. In office settings with experience in diagnosing and collecting specimens from varicella cases, a proportion as high as 69% of clinically suspected varicella cases was confirmed by PCR.18 This proportion is likely to be lower in settings without such experience, in which mild, atypical varicella cases can be misclassified easily. Moreover, the specificity of the diagnosis of breakthrough disease is expected to be lower when cases occur sporadically than in an outbreak setting. Therefore, primary care physicians should maintain a high level of suspicion especially for mild and atypical presentations, make every effort to establish epidemiologic links to cases, and obtain clinical specimens for laboratory testing. All state public health laboratories have the capability to diagnose VZV infection either by PCR or by direct fluorescent antibody assays.
The following strengths should be considered for this investigation. First, we were able to confirm the presence of wild-type VZV in both vaccinated and unvaccinated students tested. Second, we confirmed immunization status by provider record review, and we restricted analysis only to children with record-documented vaccination. Our study has some limitations. First, we relied on parental reporting of disease, a method that may have resulted in an under- or overestimation of the number of cases, especially of breakthrough and, in consequence, of the effectiveness of the vaccine. However, it is unlikely that this was a major limitation as there was increased awareness among parents and health care providers regarding the occurrence of a varicella outbreak in the community and very mild cases were reported. Also, the outbreak occurred during cold months, when conditions that most commonly are mistaken for varicella (eg, insect bites, enteroviral infections) are uncommon. Second, only 81% of the students who attended the school were included in VE analysis. However, we believe that we captured all suspected cases because we checked the school attendance and called parents whose children were absent from school and did not return the questionnaire to find the reason for their absence. Last, the small number of breakthrough cases limited our ability to explore potential risk factors for vaccine failure, although in this outbreak, vaccine performance was excellent.
Our findings suggest that this varicella outbreak was attributable primarily to failure to vaccinate rather than vaccine failure and underscore the need for implementing existing policy recommendations to ensure vaccination. Special emphasis should be placed on catch-up vaccination of susceptible older children and adolescents because the likelihood of severe infection increases with age. Requirements for elementary, middle, and high school entry will contribute to a more rapid implementation of the recommendations and prevent varicella outbreaks. Pediatricians play a critical role in ensuring that their patients are vaccinated appropriately.
Our special thanks to the principal, teachers, and attendees of school A and their parents for participation; to health care providers for timely medical information and collection of clinical specimens; to Jiancheng Huang, MD, MPH, Michael Wenzel, MD, and the staff of the Maine Bureau of Health for help in the field; to Scott Schmid, PhD, and Vladimir Loparev, PhD, of the National VZV Laboratory, CDC, for specimen testing; and to Mary McCauley, MTSC, for editorial assistance.
- Accepted August 9, 2004.
- Reprint requests to (M.M.) National Immunization Program, Centers for Disease Control and Prevention, 1600 Clifton Rd NE, MS E-61, Atlanta, GA 30333. E-mail:
No conflict of interest declared.
Dr Hayes's current affiliation is: Division of Vector-Borne Infectious Diseases, Centers for Disease Control and Prevention, Fort Collins, Colorado.
PEDIATRICS (ISSN 0031 4005). Published in the public domain by the American Academy of Pediatrics.
- ↵Committee on Infectious Diseases, American Academy of Pediatrics. Recommendations for the use of live attenuated varicella vaccine. Pediatrics.1995;95 :791– 796
- ↵Committee on Infectious Diseases, American Academy of Pediatrics. Varicella vaccine update. Pediatrics.2000;105 :136– 141
- ↵Meyer PA, Seward JF, Jumaan AO, Wharton M. Varicella mortality: trends before vaccine licensure in the United States, 1970–1994. J Infect Dis.2000;182 :383– 390
- ↵CDC. Estimated vaccination coverage with individual vaccines and selected vaccination series among children 19–35 months of age by state and immunization action plan area, National Immunization Survey, Q1/2002–Q4/2002. Available at: www.cdc.gov/nip/coverage/NIS/02/tab3_antigen_state.xls
- ↵Galil K, Fair E, Mountcastle N, Britz P, Seward J. Younger age at vaccination may increase risk of varicella vaccine failure. J Infect Dis.2002;186 :102– 105
- Dworkin MS, Jennings CE, Roth-Thomas J, Lang JE, Stukenberg C, Lumpkin JR. An outbreak of varicella among children attending preschool and elementary school in Illinois. Clin Infect Dis.2002;35 :102– 104
- Buchholz U, Moolenaar R, Peterson C, Mascola L. Varicella outbreak after vaccine licensure: should they make you chicken? Pediatrics.1999;104 :561– 563
- ↵Tugwell BD, Lee LE, Gillette H, Lorber EM, Hedberg K, Cieslak PR. Chickenpox outbreak in a highly vaccinated school population. Pediatrics.2004;113 :455– 459
- ↵Fine PEM, Zell ER. Outbreaks in highly vaccinated populations: implications for studies of vaccine performance. Am J Epidemiol.1994;139 :77– 90
- ↵Bernstein HH, Rothstein EP, Watson BM, et al. Clinical survey of natural varicella compared with breakthrough varicella after immunization with live attenuated Oka/Merck varicella vaccine. Pediatrics.1993;92 :833– 837
- Watson BM, Piercy SA, Plotkin SA, Starr SE. Modified chickenpox in children immunized with the Oka/Merck varicella vaccine. Pediatrics.1993;91 :17– 22
- ↵Ampofo K, Saiman L, LaRussa P, Steinberg S, Annunziato P, Gershon A. Persistence of immunity to live attenuated varicella vaccine in healthy adults. Clin Infect Dis.2002;34 :774– 779
- ↵Chen RT, Orenstein WA. Epidemiologic methods in immunization programs. Epidemiol Rev.1996;18 :99– 117
- ↵Kolasa MS, Klemperer-Johnson S, Papania MJ. Progress toward implementation of a second-dose measles immunization requirement for all schoolchildren in the United States. J Infect Dis.2004;189(suppl 1) :S98– S103
- ↵Verstraeten T, Jumaan AO, Mullooly JP, et al. A retrospective study of the association of varicella vaccine failure with asthma, steroid use, age at vaccination, and measles-mumps-rubella vaccination. Pediatrics.2003;112(2) . Available at: www.pediatrics.org/cgi/content/full/112/2/e98
- Copyright © 2005 by the American Academy of Pediatrics