 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
a Division of Infectious Diseases and
b Center for Epidemiology and Biostatistics, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio;
c University of Cincinnati College of Medicine, Cincinnati, Ohio
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
STUDY DESIGN. Discharge data for hospitalizations and ED encounters from 1990 through 2003 were queried for patients <20 years of age with varicella International Classification of Diseases, Ninth Revision, Clinical Modification codes (052.0–052.9) in any diagnostic position. Addresses were geocoded for identification of Hamilton County, Ohio, residents. Rates were calculated according to year, age, and race, with census estimates.
RESULTS. During the 14-year study period, there were 3983 incident varicella cases; 335 patients were hospitalized and 3833 were treated only in the ED. The rate of varicella-related hospitalizations decreased from 15.7 cases per 100000 population to 5.5 cases per 100000 population between the prevaccine period (1990–1995) and the postvaccine period (1996–2003); varicella-related ED use decreased from 178.2 cases per 100000 population to 61.2 cases per 100000 population. In the prevaccine period, hospitalization and ED visit rates were significantly higher for black children than for white children. In the postvaccine period, hospitalization rates did not differ according to race but ED visit rates remained significantly higher for black children, compared with white children.
CONCLUSIONS. Varicella-related hospitalization and ED visit rates decreased significantly for both white and black children in Hamilton County, Ohio, after the introduction of varicella vaccine, and the racial disparity found before licensure decreased after licensure.
Key Words: racial disparity • vaccine • varicella
Abbreviations: ED—emergency department • ICD-9-CM—International Classification of Diseases—Ninth Revision—Clinical Modification • CCHMC—Cincinnati Children's Hospital Medical Center • RR—relative risk • CI—confidence interval
In March 1995, a live attenuated varicella zoster virus vaccine (Varivax; Merck and Co, West Point, PA) was licensed in the United States for use among healthy susceptible children 
12 months of age. Two months later, the American Academy of Pediatrics, Committee on Infectious Diseases, recommended universal vaccination of all children 12 to 18 months of age and susceptible children 19 months of age.1 In July 1996, the Advisory Committee on Immunization Practices of the Centers for Disease Control and Prevention concurred with the American Academy of Pediatrics recommendations for children <13 years of age but recommended that children 13 years of age be vaccinated only if they were at risk for exposure or serious complications.2 In 1999, the Advisory Committee on Immunization Practices updated its recommendations to include immunization recommendations for child care and school entry and for postexposure prophylaxis.3
Before vaccine licensure, varicella mortality rates had been estimated for New York State4 and the United States,5,6 and several studies had provided estimates of community disease rates7–11 and hospitalization rates.12–22 Since varicella vaccine licensure, several studies designed to examine varicella disease burden have been reported,15–17,23–26 of which 3 demonstrated significant decreases in varicella incidence23–25 and another showed a decrease in the number of varicella-related deaths.26 Only the study of varicella-related mortality rates provided estimates according to race.26
Racial disparities in varicella disease rates and susceptibility were described in several studies that examined disease rates before vaccine licensure.13,14,19–21,27–30 In military populations, there were significantly greater rates of varicella among black adults, compared with white adults, before vaccine licensure.19–21 Similar findings were reported from an analysis of data for a managed care population14 and from an analysis of Connecticut state hospital discharge data.13 In the study examining varicella-related mortality rates, rates of varicella-related deaths were comparable for white and black subjects and decreases in death rates were seen for both groups after licensure of the varicella vaccine.26
The major goal of the varicella vaccination program in the United States is to reduce the rates of morbidity and death associated with varicella infections. To assess the effectiveness in achieving this goal, data are needed on prelicensure and postlicensure disease burdens. Cincinnati Children's Hospital Medical Center (CCHMC), located in Hamilton County, Ohio, is a 325-bed pediatric hospital that provides care for children in an 8-county primary service area, including >97% of children from Hamilton County. Therefore, nearly all Hamilton County emergency department (ED) and hospital cases of varicella are captured in the CCHMC discharge data system; this provides a unique opportunity to conduct a population-based study of the impact of varicella vaccine on the epidemiologic features of varicella disease. We conducted this study to examine varicella-related hospitalization and ED visit rates before and after varicella vaccine licensure and to determine the potential racial disparity in varicella disease burden.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
1% in- and out-migration.
A distinction was made between repeat visits for the same episode of varicella and a new episode of varicella. New episodes were defined as those occurring after an interval of 30 days; only the first varicella episode was included in the study. If there were repeat visits with an ED visit occurring before or after a hospitalization, then the hospitalization was chosen for analysis. Therefore, only first, incident cases were included in the analyses.
To examine whether CCHMC captures data for the Hamilton County population, we queried an Ohio Hospital Association-sponsored database that captures discharges from all Ohio hospitals and those from other states with counties that border Ohio. Data were available for inpatients from 1996 to 2003. Only 1 Hamilton County child with varicella was hospitalized at another hospital during that time period. Since late 1994, several CCHMC outpatient facilities have been opened. Databases for these outpatient facilities were reviewed to determine whether children from Hamilton County with varicella had been treated there. Only 23 Hamilton County children with varicella diagnoses were identified from these facilities in 1994 to 2003; 20 were white, 1 was black, and 2 were classified as other race.
Ages were stratified as <1 year, 1 to 4 years, 5 to 9 years, 10 to 14 years, and 15 to 19 years of age. Overall estimates included all black and white children. Because there was an insufficient sample size for examination of other racial/ethnic groups (n = 99), including Hispanic (n = 3), Asian (n = 1), "mixed" (n = 7), and "other" (n = 88), for race-specific analyses, only 2 race categories were used, namely, white and black.
US Census population estimates for each age/race stratum were used as denominators, and each year's population estimates were used as a surrogate measure for person-years of observation time. Incidence rates were calculated for all race and age groups. Trends according to year for all age and race categories were evaluated with the Cochran-Armitage test for trend.
The prevaccine period was defined as 1990 to 1995 and the postvaccine period as 1996 to 2003. Proportions before and after licensure were compared with the Mantel-Haenszel 
2 test; t tests were used to compare means. Significance was set at P < .05. Relative risks (RRs) and 95% confidence intervals (CIs) were calculated. An analysis was also performed to examine trends in rates in the postlicensure years. Microsoft Excel 2002 and Microsoft Access 2002 (Microsoft, Redmond, WA) were used for database management, and SAS version 8.02 (SAS Institute, Cary, NC) was used for data analyses. Approval was obtained from the CCHMC institutional review board with a limited data set agreement for this research.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Demographic characteristics for incident cases according to visit type are shown in Table 1. Overall, 52% of the patients were male and 31% were white, and the mean age was 5.1 years. Of hospitalized children, 54% were male, 61% were white, and 90% were <10 years of age. The mean length of stay for hospitalized children was 4.2 days, and the median length of stay was 3 days (range: 1–33 days). The length of stay did not differ between prelicensure and postlicensure periods (4.3 days vs 3.9 days; P = .50), and there was no statistical difference in the length of stay between white and black children (3.8 days vs 4.2 days; P = .37). The mean length of stay was significantly longer for 10- to 14-year-old children (7.3 days), compared with children in the other age groups (range of means: 3.3–4.6; P < .01). Of children treated in the ED, 52% were male, 28% were white, and 87% were <10 years of age. Black children were older than white children for both hospitalizations (5.0 years vs 3.9 years; P = .03) and ED visits (5.5 years vs 4.4 years; P < .0001).
|
|
Shift in Average Age
The mean age of children with varicella was examined according to year and for the entire prelicensure and postlicensure study periods. Overall, there was no significant difference in the average age of patients between the prelicensure period and the postlicensure period. Among white children, those treated in the ED were significantly older in the postlicensure period (mean age: 4.7 years), compared with the prelicensure period (mean age: 4.1 years; P = .01); this difference was not observed for hospitalized children. The age of black children with varicella, whether hospitalized or treated in the ED, did not differ between the study periods.
Race as a Risk Factor for Varicella
Figures 1 and 2 show the rates of varicella-related hospitalizations and ED visits, according to year, for all children and for both race categories. During the 14-year period, the risk difference between black children and white children decreased. In the prelicensure period, black children were significantly more likely than white children to be hospitalized (RR: 1.6; 95% CI: 1.2–2.1) or to have an ED visit (RR: 6.1; 95% CI: 5.6–6.6) because of varicella. In the postlicensure period, there was no longer a significantly increased risk for hospitalization among black children, compared with white children (RR: 1.2; 95% CI: 0.8–1.9). Black children continued to be at increased risk for an ED visit attributable to varicella (RR: 5.5; 95% CI: 4.8–6.3), compared with white children, although the risk difference was smaller than in the prelicensure period.
|
|
Prelicensure and postlicensure, varicella-related, hospitalization rates were calculated with 4 methods (Table 3), ie, (1) varicella code in the first position, (2) varicella code in the first 2 positions, (3) varicella with infectious complications, and 4) the method described by Galil et al,16 defined as varicella as the principal discharge code, postvaricella encephalitis or varicella pneumonitis in any position, varicella in any position for a person with an immunocompromising condition, or a code for a well-described complication of varicella as the principal diagnosis with varicella in any subsequent field. The rates according to race for each of the methods were compared with our method with varicella codes in any position. Regardless of which varicella case definition method was used, there were significant decreases in varicella-related hospitalization rates after introduction of the varicella vaccine. The decreases were significant for both white children and black children with any method, except for varicella with infectious complications. Similar to the results we found with the varicella code in any position, racial disparities were seen before licensure but not after licensure with all methods except for infectious complications of varicella.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
10 times more often than hospitalizations. It is unclear, however, whether these ED visits represented predominately moderate to severe cases of varicella that required ED care or included large numbers of mild cases brought to the ED, rather than to a primary care provider, for evaluation. The prelicensure rates in our study were higher than those seen in previous reports.13,14,16 These differences are likely attributable to the different methods used to identify cases. Although the age categories for the studies varied, in a comparison of prelicensure hospitalization rates among 5- to 9-year-old children our rate of 18.4 cases per 100000 population was higher than those of other studies, which ranged from 10.0 to 13.8 cases per 100000 population.13,14,16 When we analyzed our data with the different varicella case definitions used in other studies, we found that our prelicensure rates were comparable to rates in those studies. Although our methods included potential incidental cases, with the use of the same methods throughout the study period and a broad case definition our choice of methods to identify varicella cases should not affect our overall comparisons of prelicensure and postlicensure rates and rates according to race.
To examine the magnitude of potential inclusion of incidental cases, we examined carefully all discharge diagnoses for hospitalized children with varicella. Because "incidental cases" of varicella-related hospitalizations have not been clearly defined, it is difficult to compare studies and to determine the most accurate method for determining varicella disease burden in the ED and hospital settings. With inclusion of all patients with a varicella ICD-9-CM code in any position, some patients with hospital-acquired or truly incidental cases of varicella would be included. However, it is unclear which method estimates most accurately the varicella disease burden. Inclusion of only cases with codes in the first position or even the first 2 positions might be too simplistic and would exclude many hospitalized patients with varicella, generally those with underlying diseases and complications. However, if we had used the criteria described by Galil et al,16 we would have excluded children who had a varicella code in the second position (and admission attributable to varicella) but did not have a primary diagnosis that would be considered a "well-described complication of varicella." Therefore, the best method is not necessarily the simplest. Future studies should consider these findings carefully, so that the most-accurate estimate of varicella-related hospitalizations can be made.
Rates of varicella-related hospitalizations and ED visits were higher among children <5 years of age, compared with those 5 years of age, with the highest rates among children <12 months of age in both the prelicensure and postlicensure periods. Decreases in varicella-related hospitalization and ED visit rates were seen for all age groups, even groups for which immunization is not recommended or may be underutilized. The decreases in rates among children <12 months of age are likely attributable to herd immunity, whereas the decreases among older children and adolescents are probably a result of both herd immunity and immunization.
The decreases in varicella-related hospitalization and ED visit rates we observed might be attributable to implementation of the varicella immunization program; however, other factors also might have contributed to the decreases. Shifts in health care among Hamilton County might have occurred. We do not think this was the case for hospitalizations because only 1 Hamilton County child was identified in our regional discharge database as being hospitalized outside CCHMC. For ED visits, there might have been shifts in care that changed the overall and race-specific rates. For example, better access to primary care providers might have decreased the ED visit rate for black children, causing an apparent decrease in the risk difference for varicella-related ED visits. However, the opposite is more likely to be true, because the varicella cases identified at the CCHMC urgent care sites involved predominately white children. Therefore, the risk difference may be even lower than we were able to demonstrate with our data. It is also possible that changes in the medical management of varicella affected the epidemiologic features of varicella-related hospitalizations and ED visits. For example, the use of acyclovir early in the course of varicella might have contributed to the decreases in varicella-related encounters in both settings. It was demonstrated that acyclovir use could modify the course of illness,31 but it is not clear that acyclovir use would result in lower ED and hospitalization rates. We are not aware of any published data regarding temporal trends in the use of acyclovir in the United States, and we were unable to obtain any local data to examine the use of acyclovir in Hamilton County.
Racial disparities in susceptibility to varicella were reported previously for adolescents14 and adults13,19–22 but not younger children. Both national27 and military28–30 seroprevalence studies showed higher rates of varicella susceptibility among black subjects, compared with white subjects, for all age groups studied. Although the reasons for these findings are not well understood, one possible explanation may be the difference in child care center use for black children, compared with white children. Black families, especially those with lower incomes, may use large day care centers less frequently than white families. This may result in postponement of exposure opportunities and thus increased susceptibility at older ages. In the national study, black children who were below the poverty level were more likely to be seronegative than were black children who were at or above the poverty level.27
Other reasons for this disparity could be less access to primary care among black children or delays in diagnosis, resulting in more complications or increased severity of disease, among black children. However, we did not see any difference in hospital length of stay between white children and black children, which makes increased severity of disease less likely. There might be differences in prescribing patterns for antiviral agents, with decreased use of acyclovir for black children, compared with white children, which could account in part for varicella-related hospitalization and ED visit rates being higher for black children. Early recognition and diagnosis of soft-tissue infections may be more difficult among children of color. Therefore, there might have been delays in diagnosis among black children, resulting in increased rates of hospitalization among black children, compared with white children, before licensure.
Vaccine uptake might have contributed to eliminating the disparity in hospitalization rates and decreasing the risk difference in ED visits. Data from the Centers for Disease Control and Prevention National Immunization Survey indicated that varicella vaccine uptake rates among young children in the state of Ohio were comparable for white and black children,32 with overall coverage rates for 19- to 35-month-old children increasing steadily over the 8-year period from 1996 through 2003 (from 12.6% of children immunized in 1996 to 81.4% in 2003).32 There are no available data to determine the success of catch-up immunization for the cohort of older children without a history of varicella disease. Although we anticipated that decreases might not be seen among older children, we found no increase in the mean age of cases among either black or white children after licensure. It is important to promote varicella immunization for this cohort of children, especially given the national data showing black children to be at increased risk for varicella susceptibility, which could result ultimately in increased risk of varicella among older black children, compared with older white children. However, it is reassuring that our data demonstrated decreased rates for all age and race categories, there is no longer a racial disparity for varicella-related hospitalizations, and there is a decreasing risk difference for ED visits.
It is not clear whether our results are generalizable to other populations. However, varicella immunization coverage does not vary greatly among young children across the United States,32 and our results are consistent with previous studies.
The US varicella immunization program has been accompanied by decreasing rates of varicella-related hospitalizations and ED visits and a reduction in the racial disparity of varicella-related hospitalization and ED visit rates. The analysis of population-based hospital and ED discharge data for varicella disease was an effective and relatively simple method to examine the effect of the varicella immunization program on varicella epidemiologic features. Additional questions remain regarding the racial disparity seen before licensure. It will be important to determine whether this decrease in disparity will be seen in other populations. The methods used in this study can be used to assess the impact of varicella vaccine and other vaccines in the future.
|
FOOTNOTES |
|---|
Addresscorrespondence to Mary Allen Staat, MD, MPH, Division of Infectious Diseases, Cincinnati Children's Hospital Medical Center, 3333 Burnet Ave, Cincinnati, OH 45229. E-mail: mary.staat{at}cchmc.org
Data were presented in part at the Pediatric Academic Society Meeting, Society for Pediatric Research; May 5, 2002; Baltimore, MD.
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
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. |
||
|
||
