 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
National Immunization Program, Centers for Disease Control and Prevention, Atlanta, Georgia
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS. Data were for 178616 children (19–35 months of age) and were collected in the 1997 to 2004 National Immunization Survey. The main outcome measures were estimated varicella vaccine coverage from 1997 to 2004, coverage among susceptible children (ie, those without a history of varicella disease), racial/ethnic disparities, risk factors for nonvaccination, missed opportunities to vaccinate simultaneously with other recommended vaccines, and projected increases in coverage after elimination of missed opportunities for simultaneous vaccination.
RESULTS. Varicella vaccine coverage rates increased from 26% in 1997 to 87% in 2004. State-specific coverage rates increased 44 to 80 percentage points and were >80% in 42 states and >90% in 13 states by 2004. Coverage among susceptible children increased from 62% in 1999 to 88% in 2004. From 1998 onward, no statistically significant differences in coverage were found between white and black children, whereas Hispanic children had higher coverage rates than white children in 1998 to 2001 and 2004. Risk factors for undervaccination included living in the Midwest region, living in a household with >1 child, living in nonmetropolitan areas, living below the poverty level, having a mother who did not have a college degree, and having public providers. If missed opportunities for simultaneous vaccination had been eliminated, then coverage rates would have increased from 58% to 94% in 1999 and from 87% to 96% in 2004.
CONCLUSIONS. Uptake of varicella vaccine has been steady and is an example of successful elimination of racial and ethnic disparities. Additional focus should be placed on reducing missed opportunities for simultaneous vaccination, improving coverage in rural areas and the Midwest region, and closing remaining gaps related to maternal education, provider type, and multiple-children households.
Key Words: varicella vaccine • racial disparities • vaccination coverage • immunization
Abbreviations: CDC—Centers for Disease Control and Prevention • NIS—National Immunization Survey • MSA—metropolitan statistical area • MMR—measles-mumps-rubella • Hib—Haemophilus influenzae type b • DTaP—diphtheria-tetanus-acellular pertussis • VFC—Vaccines for Children • SCHIP—State Children's Health Insurance Program
Before the availability of vaccine, almost everyone developed varicella disease (ie, chickenpox) during their lifetimes, with 
4 million cases occurring annually in the United States and leading to an estimated 11000 to 13500 hospitalizations and 100 to 150 deaths per year.1–5 A live attenuated varicella vaccine was licensed in March 1995 and was made available to private providers in May 1995. It became available to the public sector in May 1996; in July 1996, it was recommended for routine vaccination of US children 12 to 18 months of age (1 dose) and for susceptible older children, adolescents, and adults6,7 by the Advisory Committee for Immunization Practices. The United States was the first country to implement a universal vaccination program for childhood varicella. By 2004, >40 million doses of varicella vaccine had been distributed in the United States (Centers for Disease Control and Prevention [CDC], unpublished data, 2004), and disease incidence, hospitalizations, and deaths had decreased dramatically.4,5,8,9
Increasing varicella vaccination coverage to >90% is a national Healthy People 2010 goal, and eliminating racial/ethnic health disparities is an overarching objective.10 In the United States, varicella vaccination coverage among children 19 to 35 months of age has been measured since 1997 through the National Immunization Survey (NIS) and reported annually by the CDC.11 In this study, we evaluate the uptake of varicella vaccine from 1997 through 2004, estimate vaccine coverage after adjustment for history of varicella disease, examine racial/ethnic disparities in coverage, determine risk factors associated with lack of vaccination, and assess missed opportunities to vaccinate simultaneously with other recommended vaccines. We also project increases in vaccination coverage that could result from elimination of missed opportunities for simultaneous vaccination.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Beginning in 1997, varicella vaccination coverage has been estimated by the NIS. The current study was based on data collected from 1997 to 2004, during which period >30000 household interviews were completed each year (89–94% of households identified with age-eligible children). Analysis was restricted to children whose immunization history was verified by their health care providers (65–71%). In total, we analyzed data for 178616 children (21310–23642 children per year). These data are weighted to account for nonresponding households and vaccination providers, as well as lower vaccination coverage among children in households without telephones, to ensure that estimations of coverage are representative of all children 19 to 35 months of age for each of the 8 survey years.12,13
Outcomes Measured
The primary outcome for this study was valid varicella vaccination coverage, defined as receipt of varicella vaccine at or after 12 months of age. We assessed annual coverage rates for the nation and for each census region and state (including the District of Columbia). In addition, we examined the percentage of children who received varicella vaccine on time, as recommended by the Advisory Committee on Immunization Practices (ie, 12–18 months of age).14
In addition to vaccination, children can acquire immunity to varicella through occurrence of disease and thus are not required to receive vaccine.7 Beginning in 1999, respondents were queried regarding the occurrence of varicella disease among their children. For 1999 to 2004, we evaluated vaccination coverage among susceptible children by excluding children with reported disease.
Some children in the 1997 and 1998 NIS were born before July 1995 and thus were 
12 months of age when the recommendation to vaccinate all children 12 to 18 months of age was published. Because those children had less opportunity to be vaccinated than did those who were <12 months of age at the time of the recommendation, we also evaluated coverage rates excluding them.
To identify risk factors associated with undervaccination or slow vaccine uptake, we evaluated vaccine coverage according to various characteristics of the child, household, and immunization providers. In the household portion of the survey, parents or caregivers reported race/ethnicity of the child (Hispanic, non-Hispanic black or white, or mixed/other) and maternal education level. Poverty status was determined on the basis of household size, composition, and income reported by the survey respondents, as defined by the US Census Bureau.15 Urbanicity (metropolitan statistical area [MSA] central city, MSA non-central city, or non-MSA) was determined on the basis of respondents' telephone area code/exchange. Provider facility type (public, private, or other/mixed) was reported by the providers.
Validated vaccination dates are available for analysis in the 1999 to 2004 NIS. We used these dates to examine simultaneous vaccination with other vaccines. We defined simultaneous vaccination as administration of varicella vaccine on the same day as other routinely recommended vaccines. Recommended periods for administration of several other vaccines overlap that of varicella, including measles-mumps-rubella (MMR) vaccine at 12 to 15 months of age, the third dose of poliovirus vaccine at 6 to 18 months of age, the fourth dose of diphtheria-tetanus-acellular pertussis (DTaP) vaccine at 15 to 18 months of age, and the booster dose of Haemophilus influenzae type b (Hib) vaccine at 12 to 15 months of age.14 For children who had not received varicella vaccine, we calculated the proportion who had received another vaccine at 12 months of age, which we defined as a missed opportunity for simultaneous vaccination. Finally, we projected improvements in varicella vaccination coverage levels that would be achieved if these missed opportunities for simultaneous vaccination were eliminated.
Statistical Analyses
Percentage estimates, odds ratios, adjusted odds ratios, and 95% confidence intervals were calculated with SUDAAN software, version 8.0.16 Bivariate analyses and factor-level 
2 tests were conducted to assess risk factors for lack of varicella vaccination. Multivariate logistic regression analysis was used to evaluate the effect of race/ethnicity before and after controlling for other demographic factors (poverty status, maternal education level, type of vaccination provider, number of children in the household, and urbanicity). The level of significance was set a priori at .05. All analyses were weighted and accounted for the complex sampling design of the NIS.12
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
In the 1997 and 1998 NIS, 7373 (31%) and 98 (<1%) of the sampled children, respectively, were born before July 1995 and thus were 12 months of age at the time routine varicella vaccination was recommended officially by the Advisory Committee on Immunization Practices. When these children were excluded from the analysis, coverage rates increased from 26% to 29% in 1997 but remained unchanged (43%) in 1998.
Racial/Ethnic Disparities and Other Risk Factors
After lower uptake the first year among Hispanic and black children, coverage rates increased substantially for both of these groups (Fig 2). From 1998 onward, there were no statistically significant differences in coverage rates between white and black children, whereas Hispanic children had significantly higher coverage rates than white children in 1998 to 2001 and 2004 (Figs 2 and 3). After controlling for poverty status, maternal education level, type of vaccination provider, number of children in the household, and urbanicity, initial uptake values were statistically equivalent between white and black children (adjusted odds ratio: 0.9; P = .3), and black children were significantly more likely than white children to have received varicella vaccine in 1999, 2000, and 2002 (adjusted odds ratio: 1.2–1.3; P < .05) (Fig 3). Similarly, controlling for the aforementioned factors increased the odds ratios for Hispanic versus white children, and Hispanic children were significantly more likely than white children to have received varicella vaccine from 1998 onward (adjusted odds ratio: 1.2–1.6; P < .05).
|
|
Simultaneous Vaccination (1999–2004)
Among children who received varicella vaccine, most received a dose of another vaccine on the same date. This proportion increased steadily, from 71% in 1999 to 86% in 2004. Most commonly, children received varicella vaccine simultaneously with MMR vaccine; this proportion increased from 55% in 1999 to 70% in 2004. A substantial percentage of children also received varicella vaccine with Hib vaccine (29–37%), DTaP vaccine (19–26%), poliovirus vaccine (16–31%), and hepatitis B vaccine (7–23%).
Most children who had not received varicella vaccine had a medical visit at age 12 months during which 1 other vaccination was given. These missed opportunities for simultaneous vaccination decreased consistently between 1999 and 2004 (from 86% to 64% among children who had not received varicella vaccine and from 37% to 8% among all children) (Fig 4). White children were consistently more likely than black or Hispanic children to have had a missed opportunity for simultaneous vaccination (2–5 percentage points more than black children; P < .05 for each year; 3–7 percentage points more than Hispanic children; P < .001 for each year). If missed opportunities for simultaneous vaccination had been eliminated, then varicella vaccination coverage would have increased from 58% to 94% in 1999 and from 85% to 96% in 2004 (reaching 93–95% among Hispanic and black children). Simultaneous administration with MMR vaccine alone would have increased varicella vaccination coverage rates to 91% in 1999 and 94% in 2004.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our data suggest that one of the greatest opportunities for improving varicella vaccine coverage is simultaneous administration with other vaccines at existing health care visits. An MMR-varicella combination vaccine was licensed in the United States in September 2005. With national MMR vaccine coverage at 93% in 2004,11 widespread use of an MMR-varicella combination vaccine could boost varicella vaccine coverage immediately. In addition, missed opportunities for vaccination can be minimized by encouraging parents and providers to deliver varicella vaccine to children at their first health care visit at 12 months of age.
Disparities in vaccination coverage between white children and black or Hispanic children have been described for most universally recommended childhood vaccinations.17–21 In 2004, coverage among black children 19 to 35 months of age continued to lag behind that among white children by an estimated 3 percentage points for 3 doses of Hib vaccine and for 1 dose of MMR vaccine and by 7 percentage points for 4 doses of DTaP vaccine (P < .01 for each comparison). Differences for 3 doses of hepatitis B vaccine (2 percentage points; P = .06) and poliovirus vaccine (1 percentage point; P = .14) were smaller. The black/white disparity was 8 percentage points for the standard combined series of these vaccines (P < .001), and Chu et al22 found that this gap widened from 1996 to 2001 by an average of 1.1 percentage points per year. Differences between white and Hispanic children were smaller but were statistically significant for 4 doses of DTaP, 3 doses of Hib, and the combined series of vaccines (P < .01 for each comparison). In contrast, we found that, although initial uptake of varicella vaccine was higher among white children, differences between white and black children were not statistically significant after 1997 and Hispanic children had significantly higher coverage rates than did white children in 5 of the 8 years. Furthermore, after controlling for poverty status, maternal education level, type of vaccination provider, number of children in the household, and urbanicity, initial uptake rates were not statistically different between white children and either black or Hispanic children, whereas black children were significantly more likely than white children to have received varicella vaccine in 1999, 2000, and 2002 and Hispanic children were more likely than white children to have received varicella vaccine in all years except 1997.
Reasons for the lack of disparity are not completely clear. We found that white children were more likely than either black or Hispanic children to have missed an opportunity to receive varicella vaccine with other vaccines administered after 12 months of age. In addition, timing of varicella vaccine introduction may be a factor. Varicella was the first new vaccine to be recommended universally after implementation of the Vaccines for Children (VFC) program in 1994. The goal of VFC was to reduce vaccine cost barriers by providing free vaccine to uninsured and underinsured children. Because black children are twice as likely as white children to be eligible for VFC (66% vs 31%; P < .001; CDC, unpublished data, 2002), the program might have had a disproportionately positive effect on black children. Similarly, the State Children's Health Insurance Program (SCHIP) was established in October 1997 with the goal of expanding health insurance, including well-child care and immunizations, to children whose family income is above Medicaid eligibility but not high enough for purchase of private insurance. Although all recommended childhood vaccinations are included in VFC and SCHIP benefits, these programs might have had a greater impact on administration of newly recommended vaccines and they might have helped to prevent some missed opportunities for simultaneous vaccination.
Other risk factors that we identified for slow uptake of varicella vaccine, such as living below the poverty level, having a mother who did not have a college degree, living in a household with >1 child, and having public vaccination providers, were similar to risk factors described previously for undervaccination in general.18,23–27 However, we found that gaps in varicella vaccine coverage between these groups narrowed over the study period, which suggests that the associated barriers are being overcome successfully. Unlike for other vaccines, living in a non-MSA area was a strong and persistent risk factor for lack of varicella vaccination, consistent with previous findings.18 Storage and handling are substantially more complicated for varicella vaccine than for other vaccines, because varicella vaccine must be kept frozen at less than –13°C; deep freezers may not be readily available in rural areas, and their purchase may be more cost-prohibitive for practices with relatively few young children. In addition, pediatricians, who are more likely to offer varicella vaccine to their patients than are family physicians or general practitioners,28,29 are under-represented in rural areas.30,31
Before implementation and during the early years of the program, concerns were raised regarding uptake for varicella vaccine.32,33 Identified barriers to use mirrored those described for the introduction of other childhood vaccines, including safety, availability, cost, reimbursement, storage and handling, effectiveness, duration of immunity, effects of vaccination on the epidemiologic features of the disease, absence of child care and school requirements, and the perception that the disease is not serious enough to warrant routine childhood vaccination.8,34–37 Our data indicated that uptake of varicella vaccination increased rapidly after vaccine introduction, which suggests that these barriers are being overcome.
Previously published varicella vaccination coverage rates underestimated true coverage, because they did not adjust for children with a history of varicella disease. Although information on disease history was not collected in the NIS during 1997 to 1998, it is likely that this bias was even greater for these years, because surveillance data suggested that disease incidence was still high.8,9 Because of the dramatic decreases in varicella disease incidence since vaccine introduction, vaccine coverage rates that do not adjust for the proportion of children with natural immunity are now a relatively good proxy for true vaccine coverage.
Adolescents and adults are at increased risk of morbidity and death resulting from varicella.1,2 As opportunities for exposure to varicella disease decrease, it becomes increasingly critical (1) to ensure high vaccination coverage rates among young children, to prevent them from reaching adolescence or adulthood without having immunity to varicella, and (2) to monitor vaccine coverage and disease history among older children, adolescents, and adults and to vaccinate those without evidence of immunity. Information on vaccination coverage among older children and adults is not currently available in the NIS, and mechanisms for collecting this information should be developed.
In addition to a lack of information on the vaccination status of older children and adults, this study is subject to several limitations. First, the NIS is a telephone survey; although data are weighted to adjust for households without telephones and for nonresponse, some bias may remain. Second, history of varicella disease was reported by the caregivers and was not confirmed medically. However, because varicella disease is recognized easily, we expect misclassification to be minimal. Third, we do not have information on health care visits during which no vaccinations were administered or on vaccinations given after data collection, which limits our ability to evaluate more completely missed opportunities for vaccination and catch-up doses. Finally, the NIS was not designed to provide county-level coverage estimates, which limits our ability to identify geographically localized groups of children who have not received varicella vaccine.
Achievement of the Healthy People 2010 goal of 90% coverage for varicella vaccine will require continued cooperation between providers and parents to ensure that children are vaccinated as recommended. Additional focus should be placed on reducing missed opportunities for simultaneous vaccination, improving coverage in rural areas and the Midwest region, and continuing to close gaps related to maternal education, provider type, and multiple-children households. The success of varicella vaccination in eliminating racial disparities is encouraging and should be studied more extensively, to identify factors that could be applied to other currently recommended vaccines and to new vaccines on the horizon.
|
ACKNOWLEDGMENTS |
|---|
We acknowledge Dr Abigail Shefer for review of the manuscript and Ron Nuse for graphical support.
|
FOOTNOTES |
|---|
Address correspondence to Elizabeth T. Luman, PhD, National Immunization Program, Centers for Disease Control and Prevention, 1600 Clifton Rd NE, MS E62, Atlanta, GA 30333. E-mail: ELC7{at}cdc.gov
The authors have indicated they have no financial relationships relevant to this article to disclose.
Dr Jumaan's current address is: Agency for Toxic Substances and Disease Registry, Atlanta, GA 30333.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. V. Groom, M. L. Washington, P. J. Smith, and R. T. Bryan Underimmunization of American Indian and Alaska Native Children Pediatrics, May 1, 2008; 121(5): 938 - 944. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Fishbein, K. R. Broder, L. Markowitz, and N. Messonnier New, and Some Not-so-New, Vaccines for Adolescents and Diseases They Prevent Pediatrics, January 1, 2008; 121(Supplement_1): S5 - S14. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
