- Email: [email protected]
Varicella vaccination coverage inverse correlation with varicella hospitalizations in Spain
Vaccine 32 (2014) 7043–7046
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
Short communication
Varicella vaccination coverage inverse correlation with varicella hospitalizations in Spain Ruth Gil-Prieto a,∗ , Laura Garcia-Garcia a , María San-Martin b , Angel Gil-de-Miguel a a b
Area of Preventive Medicine & Public Health, Rey Juan Carlos University, Madrid, Spain Medical Department, Sanofi Pasteur, MSD, Madrid, Spain
a r t i c l e
i n f o
Article history: Received 28 July 2014 Received in revised form 24 October 2014 Accepted 28 October 2014 Available online 6 November 2014 Keywords: Varicella Vaccine Coverage Spain Hospitalizations
a b s t r a c t Varicella vaccines available in Spain were marketed in 1998 and 2003 for non-routine use. Since 2006 some regions included universal varicella vaccination in their regional routine vaccination programs at 15–18 months of age. Regions without universal vaccination in toddlers, but instead with the strategy of vaccinating susceptible adolescents, reached different varicella vaccination coverage through private market. This study shows the correlation between severe varicella zoster virus infections requiring hospitalization and the varicella vaccination coverage by region. A total of 3009 hospital discharges related to varicella were reported in 2009–2010. The overall annual rate of hospitalization was 3.27 cases per 100,000. In children younger than 5 years old varicella hospitalization rate was 30.73 cases per 100,000. Varicella related hospitalizations were significantly lower in the regions with universal vaccination. In those regions without universal vaccination at 15–18 months of age, those with higher coverage in private market showed lower hospitalization rates. © 2014 Elsevier Ltd. All rights reserved.
In the absence of vaccination, varicella affects almost all children in the course of their childhood [1]. Varicella is usually mild in immunocompetent children, but can cause severe complications and even fatalities [2]. In the United States, a major reduction in the varicella incidence was observed after the introduction of varicella vaccination, not only of cases of varicella but also of hospitalizations due to its complications [3]. In other countries like Australia [4], Germany [5], Canada [6], or Italy (Region of Veneto) [7] universal childhood vaccination against varicella has reached similar effects. Varicella is a notifiable disease in Spain. 141,339 and 157,914 cases were reported to the National System of Notifiable Diseases (EDO) in 2009 and 2010, respectively (http://www.isciii.es/ISCIII/ es/contenidos/fd-servicios-cientifico-tecnicos/fd-vigilanciasalertas/fd-enfermedades/enfermedades-declaracion-obligatoriaseries-temporales.shtml). The Advisory Committee on Vaccines (CAV) of the Spanish Association of Paediatrics (AEP), taking into account epidemiological data as well as evidence on the safety, effectiveness and efficiency of vaccines, recommends the administration of the first dose of the
∗ Corresponding author. Tel.: +34 914888625; fax: +34 914888955. E-mail address: [email protected] (R. Gil-Prieto). http://dx.doi.org/10.1016/j.vaccine.2014.10.076 0264-410X/© 2014 Elsevier Ltd. All rights reserved.
varicella vaccine at age 12 months and the second dose at age 2–3 years [8]. Starting in late 2006, four regions of Spain – Madrid (1-dose) and Navarra, Ceuta and Melilla (2 doses since 2007) – included varicella vaccination in their routine childhood schedule at 15–18 months of age [9]. The other 15 autonomous regions vaccinate susceptible children at 10–14 years with one or two doses but, in these 15 regions, infant vaccination is neither financed nor reimbursed by the public health care system. Following the recommendations of the AEP, paediatricians prescribe varicella vaccination at 12–18 months of age with an important number of children being vaccinated with the parents paying the full cost of vaccination. The estimated vaccination coverage in these regions was 32% in 2009–2010, ranging from 13% to 64% (vaccine distribution data provided by IMS Health). This epidemiological study aimed to describe population-based data on the burden of hospitalization for varicella in 2009–2010 according to the varicella vaccination coverage in the different Regions in Spain. The Spanish centralized hospital discharge database (CMBD: Conjunto Mínimo Básico de Datos) includes more than 98% of hospitalizations in the Spanish health care system, which covers 99.5% of the population [10]. It has been used for research purposes including epidemiological studies on varicella [11].
7044
R. Gil-Prieto et al. / Vaccine 32 (2014) 7043–7046
All varicella related hospitalizations reported from January 1st, 2009 through December 31st, 2010 were analyzed. The 9th International Classification of Diseases ICD-9-CM codes for varicella (052.X) were selected. The annual incidences of hospitalizations both in the general population and in children under 5 years old were calculated. Data were also obtained by Autonomous Region. Spearman correlation was used for assessing the association between the hospitalization rates and the vaccination coverage. Data on vaccination coverage by region were obtained from distribution data provided by IMS Health. In all test the significance level used was p < 0.05. Statistical analyses were performed using the Statistical Package for Social Sciences (SPSS/PASW, version 19.0; Chicago, IL) and R, version 2.15.3 (R Core Team, 2013, Vienna, Austria). A total of 3009 hospital discharges related to varicella were reported in 2009–2010. The overall annual rate of hospitalization was 3.27 (CI 95%: 3.16–3.39) cases per 100,000. Of those, 1472 hospitalizations occurred in children up to 5 years old with a rate of hospitalization of 30.73 (CI 95%: 29.16–32.30) cases per 100,000. Six percent of the patients hospitalized were immunocompromised. Varicella was coded as principal diagnosis in 66% of the hospitalizations, of those, 45.8% were coded as varicella, 26.8% as hemorrhagic pneumonitis, 22.6% complications of varicella, 4.5% encephalitis and 0.3 myelitis. There were 42 deaths related to varicella infection during the study period, all of them occurred in immunocompromised patients. In Table 1, the 17 Autonomous Regions and the 2 Autonomous Cities are grouped regarding the varicella vaccination coverage and their hospitalization rates during the period. There is an inverse linear correlation between vaccination coverage and hospitalization rates during the 2009–2010 period, both for all age population (Rho = −0.850, p < 0.001) and for children up to 5 years old (Rho = −0.496, p = 0.036). If we only consider non-routine
vaccination regions (Madrid, Navarra, Ceuta and Melilla excluded) the inverse linear correlation for all age population remains significant (Rho = −0.717, p < 0.001). It is not the case for children up to 5years old (Rho = −0.148, p = 0.597). Fig. 1 shows the hospitalization rates related to varicella in the different autonomous regions. Navarra, Madrid, Ceuta, Melilla, Castilla León and Canary Islands showed lower hospitalization rates in the 2009–2010 period (1.37, 1.99, 1.29, 2.73, 1.91 and 1.93 per 100,000, respectively). The first four regions had universal coverage with vaccination coverage of nearly 90%, Castilla-León and Canary Islands do not have universal infant vaccination programs, but reached the highest coverage in the non-routine vaccination (64% and 46%, respectively). Among the rest of the regions, Cantabria and Murcia registered the highest hospitalizations rate (5.70 and 5.20 per 100,000, respectively (Fig. 1). These regions were among those with the lowest vaccination coverage (16% and 21% in the period 2009–2010, respectively). This retrospective epidemiology study shows the hospitalization rates in the different regions of Spain. Importantly it highlights the inverse correlation between the varicella vaccination coverage in Spain and the incidence of hospitalizations. The main finding of this study is that the hospitalization rates related to VZV infection in Spain are lower in the regions with universal vaccination and among those with non-routine vaccination but reaching high coverage compared to the regions with non-routine vaccination and low coverage. This pattern can also be observed in 2010 for hospitalizations occurred in children <5 years old, where more than 50% of the hospitalization occurred and which is the age group directly benefited by the vaccination. Our results show a potential beneficial effect from herd protection in older children and adolescents due to some reduction in the varicella circulation in the community. Important point to consider as the global prevalence of antibodies for varicella-zoster virus among young adults in a recent study in Spain was 92.8% [12].
Table 1 Hospitalization rates by Autonomous Regions in Spain regarding the vaccination coverage. Vaccination coveragea
Regions
<25% 25–34% 35–44% 45–54% >55% Universal vaccination
Asturias, Baleares, Cantabria, Murcia ˜ Galicia, P. Vasco Andalucía, Aragón, Castilla la Mancha, Cataluna, Valencia, Extremadura, La Rioja Canarias C. León**** Madrid, Navarra, Ceuta and Melilla
Vaccination coverage*
Regions
<25% 25–34% 35–44% 45–54% >55% Universal vaccination
Asturias, Baleares, Cantabria, Murcia ˜ Galicia, P. Vasco Andalucía, Aragón, Castilla la Mancha, Cataluna, Valencia, Extremadura, La Rioja Canarias C. León**** Madridb , Navarra, Ceuta and Melilla
Hospitalization rates (per 100,000) 2009**
2010***
2009–2010
4.84 3.89 3.66 2.91 2.67 2.41
4.83 3.55 2.48 0.96 1.16 1.46
4.84 3.72 3.07 1.93 1.91 1.93
Hospitalization rates in children <5 (per 100,000) 2009**
2010***
2009–2010
50.81 26.99 33.11 42.90 36.87 17.30
41.97 31.84 22.87 60.02 53.55 7.04
46.35 29.44 27.94 51.55 45.33 12.08
a Distribution data provided by IMS Health. Spearman correlation was used for assessing the association between the hospitalization rates and the vaccination coverage. b ˜ A vaccine coverage of 92.7% has been considered for Madrid [García Comas L, Ordobás Gavín M, Canellas Llabrés S, Gutiérrez Rodríguez A, Servicio de Epidemiología. Subdirección General de Promoción de la Salud y Prevención. Dirección General de Atención Primaria. Comunidad de Madrid. Boletín Epidemiológico de la Comunidad de Madrid. 2010; 7 (16) ISSN 1695-7059 http://www.madrid.org/cs/Satellite?blobcol=urldata&blobheader=application%2Fpdf&blobheadername1=Content-disposition &blobheadername2=cadena&blobheadervalue1=filename%3DJulio2010.pdf&blobheadervalue2=language%3Des%26site%3DPortalSalud&blobkey=id&blobtable=MungoBlobs &blobwhere=1271908141371&ssbinary=true]. * A vaccine coverage of 90% has been considered for Ceuta. Melilla and Navarra. ** Coverage groups 2009: <25% (Asturias, Baleares, Cantabria, Murcia). 25–34% (Andalucía, Aragón, Castilla la Mancha, Catalonia, Galicia, Basque Country). 35–44% (Valencia, Extremadura, La Rioja). 45–54% (Canary Islands). >54% (Castilla Leon) and Systematic vaccination (Madrid, Navarra, Ceuta, Melilla). *** 2010: <25% (Asturias, Baleares, Cantabria, Murcia). 25–34% (Andalucía, Castilla la Mancha, Catalonia, Galicia, Basque Country). 35–44% (Aragón, Valencia, Extremadura, Canary Islands). 45–54% (La Rioja). >54% (Castilla Leon) and Systematic vaccination (Madrid, Navarra, Ceuta, Melilla). **** Vaccination coverage in Castilla y León for the study period (2009–2010) was 64%.
R. Gil-Prieto et al. / Vaccine 32 (2014) 7043–7046
7045
Fig. 1. Hospitalization rates (per 100,000) for varicella infection and varicella vaccination coverage by Autonomous region in Spain (2009–2010). * A vaccine coverage ˜ Llabrés S, Gutiérrez Rodríguez A, Servicio of 92.7% has been considered for Madrid with a 1-dose vaccination program [García Comas L, Ordobás Gavín M, Canellas de Epidemiología. Subdirección General de Promoción de la Salud y Prevención. Dirección General de Atención Primaria. Comunidad de Madrid. Boletín Epidemiológico de la Comunidad de Madrid. 2010; 7 (16) ISSN 1695-7059 http://www.madrid.org/cs/Satellite?blobcol=urldata&blobheader=application%2Fpdf&blobheadername1=Content -disposition&blobheadername2=cadena&blobheadervalue1=filename%3DJulio2010.pdf&blobheadervalue2=language%3Des%26site%3DPortalSalud&blobkey=id&blobtable= MungoBlobs&blobwhere=1271908141371&ssbinary=true]. * A vaccine coverage of 90% has been considered for Ceuta. Melilla and Navarra with 2-doses vaccination programmes.
Our data are in line with those published in other studies following varicella vaccine introduction with high coverage in Australia—>80% coverage [13], Spain—[14], and in poor coverage countries or regions, such as Japan or Turkey [15,16]. Meanwhile universal vaccination programs in childhood drastically reduce the varicella morbidity and mortality reducing the circulation of the virus, vaccination of susceptible adolescents reduces the number of severe cases more likely to occur in adolescents and young adults [17]. Some countries have implemented different programs for the successful control of varicella infections. Australia implemented a single dose program with the monovalent varicella vaccine at 18 months of age and a catch-up program at 11–13 years of age [18], the United States, recommended a second dose of varicella vaccine at 4 to 6 years in 2006 after outbreaks of breakthrough disease were noted in the early to mid-2000s despite estimates of 1 dose effectiveness of 80% to 90% [3]. Canada also implemented a two-dose immunization strategy in 2007 [6]. In Navarra, one of the Spanish regions with routine varicella vaccine included in this study, the incidence of varicella decreased by 98.5% in the vaccinated cohorts (children aged 1 to 8) between 2006 and 2012. In unvaccinated age groups, an important impact was also observed: 90.5% in infants under 1 year of age, and 89.4% in children aged nine years. Hospitalizations rates due for varicella or its complications decreased by 89% [14]. Our results for the regions where varicella vaccination is not included in the routine childhood vaccination schedule at 15–18 months, but an important percentage of parents buy the vaccine in private market and vaccinate their children before they reach the
age when the vaccination is publicly funded, show that the pediatrician and medical societies recommendations are impacting in the incidence of severe cases of varicella that require hospitalization thanks to the families economic effort. Changes in public health policies in Spain, such as the inclusion of varicella vaccination in the vaccination routine schedule of the different regions would contribute to the control of severe varicella episodes as a consequence of the drastic decrease in disease incidence, both in vaccinated and unvaccinated individuals, as it has been observed in regions with such vaccination programs. Furthermore, routine varicella vaccination is an example of successful reduction in racial and ethnic disparities in infectious disease incidence in the USA [19]. Varicella causes a significant burden of severe disease requiring hospital admission in Spain. Both community and nosocomial infections in susceptible and especially immunocompromised patients may increase in the absence of a varicella vaccine in the Spanish market. Further studies considering a cost effective approach, with rigorously applied economic evaluation methods [20] that can provide to the health authorities and policy makers with arguments towards the preventive strategies are strongly needed. This study has some limitations derived from the use of a National-wide hospital data base. The reliability of hospital-based surveillance depends on the quality of the discharge report, the clinical history and the codification process [10]. Potential biases associated to coding or the assumption of causality between disease and potential complications could result in misclassification. CMBD does not give information regarding laboratory confirmation.
7046
R. Gil-Prieto et al. / Vaccine 32 (2014) 7043–7046
However, we assumed that a differential diagnosis of severe varicella has been confirmed, as it is a notifiable disease in Spain. Varicella is a common and easily recognizable disease, so this misclassification may be lower than for other diagnoses. Conflict of interest statement MSM is employee of Sanofi Pasteur, MSD, Spain. RGP and AGM have received support for grants and travel grants for Scientific Congress from Sanofi Pasteur, MSD and GSK. Acknowledgement The Subdirección General del Instituto de Información Sanitaria provided with the information in which this study is based. References [1] Bonanni P, Breuer J, Gershon A, Gershon M, Hryniewicz W, Papaevangelou V, et al. Varicella vaccination in Europe-taking the practical approach. BMC Med 2009;7(May):26, http://dx.doi.org/10.1186/1741-7015-7-26. [2] Streng A, Grote V, Carr D, Hagemann C, Liese JG. Varicella routine vaccination and the effects on varicella epidemiology—results from the Bavarian Varicella Surveillance Project (BaVariPro), 2006–2011. BMC Infect Dis 2013;13:303. Available at: http://biomedcentral.com/1471-2334/13/303 [3] Baxter R, Tran NT, Ray P, Lewis E, Fireman B, Black S, et al. Impact of vaccination on the epidemiology of varicella: 1995–2009. Pediatrics 2014;134:24. Accessible at: http://pediatrics.aappublications.org/content/134/24.full.html [4] Carville KS, Riddell MA, Kelly HA. A decline in varicella but an uncertain impact on zoster following varicella vaccination in Victoria, Australia. Vaccine 2010;16(28):2532–8. [5] Siedler A, Arndt U. Impact of the routine varicella vaccination programme on varicella epidemiology in Germany. Eurosurveillance 2010;15, pii=19530. [6] Tan B, Bettinger J, McConnell A, Scheifele D, Halperin S, Vaudry W, et al. Members of the Canadian Immunization Monitoring Program, Active (IMPACT). The effect of funded varicella immunization programs on varicella-related hospitalizations in IMPACT centers, Canada, 2000–2008. Pediatr Infect Dis J 2012;31:956–63.
[7] Pozza F, Piovesan C, Russo F, Bella A, Pezzotti P, Emberti Gialloreti L. Impact of universal vaccination on the epidemiology of varicella in Veneto, Italy. Vaccine 2011;29:9480–7. [8] Moreno-Pérez D, Álvarez García FJ, Arístegui Fernández J, Barrio Corrales F, Cilleruelo Ortega MJ, Corretger Rauet JM, et al. Comité Asesor de Vacunas de ˜ ˜ Immunisation schedule of the de Pediatría, Espana. la Asociación Espanola Spanish Association of Paediatrics: 2013 recommendations. An Pediatr (Barc) 2013;78(Jan (1)):59e1–627e. ˜ [9] Pena-Rey I, Martínez de Aragón MV, Villaverde Hueso A, Terres Arellano M, Alcalde Cabero E, Suárez Rodríguez B. Epidemiology of varicela in Spain in pre and post vaccination periods. Rev Esp Salud Públ 2009;83:711–24. [10] Peiro S, Librero J. Evaluación de la calidad a partir del conjunto mínimo de datos básicos al alta hospitalaria. Rev Neurol 1999;29:651–61. [11] Guillen JM, Gil-Prieto R, Alvaro A, Gil A. Burden of adult varicella hospitalizations in Spain (2001–2007). Hum Vaccin 2010;6(Aug (8)):659–63. [12] González-Escalada A, García-García L, Viguera-Ester P, Marín-García P, García J, Gil de Miguel A, et al. Seroprevalence of antibodies against measles, rubella, mumps, varicella-zoster, and B. Pertussis in young adults of Madrid, Spain. Hum Vaccin Immunother 2013;9(Jun (9)):1918–25. [13] Marshall HS, McIntyre P, Richmond P, Buttery JP, Royle JA, Gold MS, et al. Changes in patterns of hospitalized children with varicella and of associated varicella genotypes following introduction of varicella vaccine in Australia. Pediatr Infect Dis J 2012;(Dec (17)):530–7. [14] García Cenoz M, Castilla J, Chamorro J, Martínez-Baz I, Martínez-Artola V, Irisarri F, et al. Impact of universal two-dose vaccination on varicella epidemiology in Navarre, Spain, 2006 to 2012. Eurosurveillance 2013;18, pii=20552. [15] Ozaki T. Varicella vaccination in Japan: necessity of implementing a routine vaccination program. J Infect Chemother 2013;19(Apr (2)):188–95. [16] Kurugol Z, Halicioglu O, Koc F, Koturoglu G, Aksit S. Varicella rates among unvaccinated and one-dose vaccinated healthy children in Izmir, Turkey. Int J Infect Dis 2011;15(Jul (7)):e475–80. [17] Van Hoek AJ, Melegaro A, Zagheni E, Edmunds WJ, Gay N. Modelling the impact of a combined varicella and zoster vaccination programme on the epidemiology of varicella zoster virus in England. Vaccine 2011;29(Mar (13)):2411–20. [18] Ward K, Dey A, Hull B, Quinn HE, Macartney K, Menzies R. Evaluation of Australia’s varicella vaccination program for children and adolescents. Vaccine 2013;31(Feb (10)):1413–9, http://dx.doi.org/10.1016/j.vaccine.2012.12.052 (Epub 2013 Jan 2). [19] Luman ET, Ching PL, Jumaan AO, Seward JF. Uptake of varicella vaccination among young children in the United States: a success story in eliminating racial and ethnic disparities. Pediatrics 2006;117(Apr (4)):999–1008. [20] García-Altés A. Systematic review of economic evaluation studies: are vaccination programs efficient in Spain? Vaccine 2013;31(Mar (13)):1656–65.
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
Short communication
Varicella vaccination coverage inverse correlation with varicella hospitalizations in Spain Ruth Gil-Prieto a,∗ , Laura Garcia-Garcia a , María San-Martin b , Angel Gil-de-Miguel a a b
Area of Preventive Medicine & Public Health, Rey Juan Carlos University, Madrid, Spain Medical Department, Sanofi Pasteur, MSD, Madrid, Spain
a r t i c l e
i n f o
Article history: Received 28 July 2014 Received in revised form 24 October 2014 Accepted 28 October 2014 Available online 6 November 2014 Keywords: Varicella Vaccine Coverage Spain Hospitalizations
a b s t r a c t Varicella vaccines available in Spain were marketed in 1998 and 2003 for non-routine use. Since 2006 some regions included universal varicella vaccination in their regional routine vaccination programs at 15–18 months of age. Regions without universal vaccination in toddlers, but instead with the strategy of vaccinating susceptible adolescents, reached different varicella vaccination coverage through private market. This study shows the correlation between severe varicella zoster virus infections requiring hospitalization and the varicella vaccination coverage by region. A total of 3009 hospital discharges related to varicella were reported in 2009–2010. The overall annual rate of hospitalization was 3.27 cases per 100,000. In children younger than 5 years old varicella hospitalization rate was 30.73 cases per 100,000. Varicella related hospitalizations were significantly lower in the regions with universal vaccination. In those regions without universal vaccination at 15–18 months of age, those with higher coverage in private market showed lower hospitalization rates. © 2014 Elsevier Ltd. All rights reserved.
In the absence of vaccination, varicella affects almost all children in the course of their childhood [1]. Varicella is usually mild in immunocompetent children, but can cause severe complications and even fatalities [2]. In the United States, a major reduction in the varicella incidence was observed after the introduction of varicella vaccination, not only of cases of varicella but also of hospitalizations due to its complications [3]. In other countries like Australia [4], Germany [5], Canada [6], or Italy (Region of Veneto) [7] universal childhood vaccination against varicella has reached similar effects. Varicella is a notifiable disease in Spain. 141,339 and 157,914 cases were reported to the National System of Notifiable Diseases (EDO) in 2009 and 2010, respectively (http://www.isciii.es/ISCIII/ es/contenidos/fd-servicios-cientifico-tecnicos/fd-vigilanciasalertas/fd-enfermedades/enfermedades-declaracion-obligatoriaseries-temporales.shtml). The Advisory Committee on Vaccines (CAV) of the Spanish Association of Paediatrics (AEP), taking into account epidemiological data as well as evidence on the safety, effectiveness and efficiency of vaccines, recommends the administration of the first dose of the
∗ Corresponding author. Tel.: +34 914888625; fax: +34 914888955. E-mail address: [email protected] (R. Gil-Prieto). http://dx.doi.org/10.1016/j.vaccine.2014.10.076 0264-410X/© 2014 Elsevier Ltd. All rights reserved.
varicella vaccine at age 12 months and the second dose at age 2–3 years [8]. Starting in late 2006, four regions of Spain – Madrid (1-dose) and Navarra, Ceuta and Melilla (2 doses since 2007) – included varicella vaccination in their routine childhood schedule at 15–18 months of age [9]. The other 15 autonomous regions vaccinate susceptible children at 10–14 years with one or two doses but, in these 15 regions, infant vaccination is neither financed nor reimbursed by the public health care system. Following the recommendations of the AEP, paediatricians prescribe varicella vaccination at 12–18 months of age with an important number of children being vaccinated with the parents paying the full cost of vaccination. The estimated vaccination coverage in these regions was 32% in 2009–2010, ranging from 13% to 64% (vaccine distribution data provided by IMS Health). This epidemiological study aimed to describe population-based data on the burden of hospitalization for varicella in 2009–2010 according to the varicella vaccination coverage in the different Regions in Spain. The Spanish centralized hospital discharge database (CMBD: Conjunto Mínimo Básico de Datos) includes more than 98% of hospitalizations in the Spanish health care system, which covers 99.5% of the population [10]. It has been used for research purposes including epidemiological studies on varicella [11].
7044
R. Gil-Prieto et al. / Vaccine 32 (2014) 7043–7046
All varicella related hospitalizations reported from January 1st, 2009 through December 31st, 2010 were analyzed. The 9th International Classification of Diseases ICD-9-CM codes for varicella (052.X) were selected. The annual incidences of hospitalizations both in the general population and in children under 5 years old were calculated. Data were also obtained by Autonomous Region. Spearman correlation was used for assessing the association between the hospitalization rates and the vaccination coverage. Data on vaccination coverage by region were obtained from distribution data provided by IMS Health. In all test the significance level used was p < 0.05. Statistical analyses were performed using the Statistical Package for Social Sciences (SPSS/PASW, version 19.0; Chicago, IL) and R, version 2.15.3 (R Core Team, 2013, Vienna, Austria). A total of 3009 hospital discharges related to varicella were reported in 2009–2010. The overall annual rate of hospitalization was 3.27 (CI 95%: 3.16–3.39) cases per 100,000. Of those, 1472 hospitalizations occurred in children up to 5 years old with a rate of hospitalization of 30.73 (CI 95%: 29.16–32.30) cases per 100,000. Six percent of the patients hospitalized were immunocompromised. Varicella was coded as principal diagnosis in 66% of the hospitalizations, of those, 45.8% were coded as varicella, 26.8% as hemorrhagic pneumonitis, 22.6% complications of varicella, 4.5% encephalitis and 0.3 myelitis. There were 42 deaths related to varicella infection during the study period, all of them occurred in immunocompromised patients. In Table 1, the 17 Autonomous Regions and the 2 Autonomous Cities are grouped regarding the varicella vaccination coverage and their hospitalization rates during the period. There is an inverse linear correlation between vaccination coverage and hospitalization rates during the 2009–2010 period, both for all age population (Rho = −0.850, p < 0.001) and for children up to 5 years old (Rho = −0.496, p = 0.036). If we only consider non-routine
vaccination regions (Madrid, Navarra, Ceuta and Melilla excluded) the inverse linear correlation for all age population remains significant (Rho = −0.717, p < 0.001). It is not the case for children up to 5years old (Rho = −0.148, p = 0.597). Fig. 1 shows the hospitalization rates related to varicella in the different autonomous regions. Navarra, Madrid, Ceuta, Melilla, Castilla León and Canary Islands showed lower hospitalization rates in the 2009–2010 period (1.37, 1.99, 1.29, 2.73, 1.91 and 1.93 per 100,000, respectively). The first four regions had universal coverage with vaccination coverage of nearly 90%, Castilla-León and Canary Islands do not have universal infant vaccination programs, but reached the highest coverage in the non-routine vaccination (64% and 46%, respectively). Among the rest of the regions, Cantabria and Murcia registered the highest hospitalizations rate (5.70 and 5.20 per 100,000, respectively (Fig. 1). These regions were among those with the lowest vaccination coverage (16% and 21% in the period 2009–2010, respectively). This retrospective epidemiology study shows the hospitalization rates in the different regions of Spain. Importantly it highlights the inverse correlation between the varicella vaccination coverage in Spain and the incidence of hospitalizations. The main finding of this study is that the hospitalization rates related to VZV infection in Spain are lower in the regions with universal vaccination and among those with non-routine vaccination but reaching high coverage compared to the regions with non-routine vaccination and low coverage. This pattern can also be observed in 2010 for hospitalizations occurred in children <5 years old, where more than 50% of the hospitalization occurred and which is the age group directly benefited by the vaccination. Our results show a potential beneficial effect from herd protection in older children and adolescents due to some reduction in the varicella circulation in the community. Important point to consider as the global prevalence of antibodies for varicella-zoster virus among young adults in a recent study in Spain was 92.8% [12].
Table 1 Hospitalization rates by Autonomous Regions in Spain regarding the vaccination coverage. Vaccination coveragea
Regions
<25% 25–34% 35–44% 45–54% >55% Universal vaccination
Asturias, Baleares, Cantabria, Murcia ˜ Galicia, P. Vasco Andalucía, Aragón, Castilla la Mancha, Cataluna, Valencia, Extremadura, La Rioja Canarias C. León**** Madrid, Navarra, Ceuta and Melilla
Vaccination coverage*
Regions
<25% 25–34% 35–44% 45–54% >55% Universal vaccination
Asturias, Baleares, Cantabria, Murcia ˜ Galicia, P. Vasco Andalucía, Aragón, Castilla la Mancha, Cataluna, Valencia, Extremadura, La Rioja Canarias C. León**** Madridb , Navarra, Ceuta and Melilla
Hospitalization rates (per 100,000) 2009**
2010***
2009–2010
4.84 3.89 3.66 2.91 2.67 2.41
4.83 3.55 2.48 0.96 1.16 1.46
4.84 3.72 3.07 1.93 1.91 1.93
Hospitalization rates in children <5 (per 100,000) 2009**
2010***
2009–2010
50.81 26.99 33.11 42.90 36.87 17.30
41.97 31.84 22.87 60.02 53.55 7.04
46.35 29.44 27.94 51.55 45.33 12.08
a Distribution data provided by IMS Health. Spearman correlation was used for assessing the association between the hospitalization rates and the vaccination coverage. b ˜ A vaccine coverage of 92.7% has been considered for Madrid [García Comas L, Ordobás Gavín M, Canellas Llabrés S, Gutiérrez Rodríguez A, Servicio de Epidemiología. Subdirección General de Promoción de la Salud y Prevención. Dirección General de Atención Primaria. Comunidad de Madrid. Boletín Epidemiológico de la Comunidad de Madrid. 2010; 7 (16) ISSN 1695-7059 http://www.madrid.org/cs/Satellite?blobcol=urldata&blobheader=application%2Fpdf&blobheadername1=Content-disposition &blobheadername2=cadena&blobheadervalue1=filename%3DJulio2010.pdf&blobheadervalue2=language%3Des%26site%3DPortalSalud&blobkey=id&blobtable=MungoBlobs &blobwhere=1271908141371&ssbinary=true]. * A vaccine coverage of 90% has been considered for Ceuta. Melilla and Navarra. ** Coverage groups 2009: <25% (Asturias, Baleares, Cantabria, Murcia). 25–34% (Andalucía, Aragón, Castilla la Mancha, Catalonia, Galicia, Basque Country). 35–44% (Valencia, Extremadura, La Rioja). 45–54% (Canary Islands). >54% (Castilla Leon) and Systematic vaccination (Madrid, Navarra, Ceuta, Melilla). *** 2010: <25% (Asturias, Baleares, Cantabria, Murcia). 25–34% (Andalucía, Castilla la Mancha, Catalonia, Galicia, Basque Country). 35–44% (Aragón, Valencia, Extremadura, Canary Islands). 45–54% (La Rioja). >54% (Castilla Leon) and Systematic vaccination (Madrid, Navarra, Ceuta, Melilla). **** Vaccination coverage in Castilla y León for the study period (2009–2010) was 64%.
R. Gil-Prieto et al. / Vaccine 32 (2014) 7043–7046
7045
Fig. 1. Hospitalization rates (per 100,000) for varicella infection and varicella vaccination coverage by Autonomous region in Spain (2009–2010). * A vaccine coverage ˜ Llabrés S, Gutiérrez Rodríguez A, Servicio of 92.7% has been considered for Madrid with a 1-dose vaccination program [García Comas L, Ordobás Gavín M, Canellas de Epidemiología. Subdirección General de Promoción de la Salud y Prevención. Dirección General de Atención Primaria. Comunidad de Madrid. Boletín Epidemiológico de la Comunidad de Madrid. 2010; 7 (16) ISSN 1695-7059 http://www.madrid.org/cs/Satellite?blobcol=urldata&blobheader=application%2Fpdf&blobheadername1=Content -disposition&blobheadername2=cadena&blobheadervalue1=filename%3DJulio2010.pdf&blobheadervalue2=language%3Des%26site%3DPortalSalud&blobkey=id&blobtable= MungoBlobs&blobwhere=1271908141371&ssbinary=true]. * A vaccine coverage of 90% has been considered for Ceuta. Melilla and Navarra with 2-doses vaccination programmes.
Our data are in line with those published in other studies following varicella vaccine introduction with high coverage in Australia—>80% coverage [13], Spain—[14], and in poor coverage countries or regions, such as Japan or Turkey [15,16]. Meanwhile universal vaccination programs in childhood drastically reduce the varicella morbidity and mortality reducing the circulation of the virus, vaccination of susceptible adolescents reduces the number of severe cases more likely to occur in adolescents and young adults [17]. Some countries have implemented different programs for the successful control of varicella infections. Australia implemented a single dose program with the monovalent varicella vaccine at 18 months of age and a catch-up program at 11–13 years of age [18], the United States, recommended a second dose of varicella vaccine at 4 to 6 years in 2006 after outbreaks of breakthrough disease were noted in the early to mid-2000s despite estimates of 1 dose effectiveness of 80% to 90% [3]. Canada also implemented a two-dose immunization strategy in 2007 [6]. In Navarra, one of the Spanish regions with routine varicella vaccine included in this study, the incidence of varicella decreased by 98.5% in the vaccinated cohorts (children aged 1 to 8) between 2006 and 2012. In unvaccinated age groups, an important impact was also observed: 90.5% in infants under 1 year of age, and 89.4% in children aged nine years. Hospitalizations rates due for varicella or its complications decreased by 89% [14]. Our results for the regions where varicella vaccination is not included in the routine childhood vaccination schedule at 15–18 months, but an important percentage of parents buy the vaccine in private market and vaccinate their children before they reach the
age when the vaccination is publicly funded, show that the pediatrician and medical societies recommendations are impacting in the incidence of severe cases of varicella that require hospitalization thanks to the families economic effort. Changes in public health policies in Spain, such as the inclusion of varicella vaccination in the vaccination routine schedule of the different regions would contribute to the control of severe varicella episodes as a consequence of the drastic decrease in disease incidence, both in vaccinated and unvaccinated individuals, as it has been observed in regions with such vaccination programs. Furthermore, routine varicella vaccination is an example of successful reduction in racial and ethnic disparities in infectious disease incidence in the USA [19]. Varicella causes a significant burden of severe disease requiring hospital admission in Spain. Both community and nosocomial infections in susceptible and especially immunocompromised patients may increase in the absence of a varicella vaccine in the Spanish market. Further studies considering a cost effective approach, with rigorously applied economic evaluation methods [20] that can provide to the health authorities and policy makers with arguments towards the preventive strategies are strongly needed. This study has some limitations derived from the use of a National-wide hospital data base. The reliability of hospital-based surveillance depends on the quality of the discharge report, the clinical history and the codification process [10]. Potential biases associated to coding or the assumption of causality between disease and potential complications could result in misclassification. CMBD does not give information regarding laboratory confirmation.
7046
R. Gil-Prieto et al. / Vaccine 32 (2014) 7043–7046
However, we assumed that a differential diagnosis of severe varicella has been confirmed, as it is a notifiable disease in Spain. Varicella is a common and easily recognizable disease, so this misclassification may be lower than for other diagnoses. Conflict of interest statement MSM is employee of Sanofi Pasteur, MSD, Spain. RGP and AGM have received support for grants and travel grants for Scientific Congress from Sanofi Pasteur, MSD and GSK. Acknowledgement The Subdirección General del Instituto de Información Sanitaria provided with the information in which this study is based. References [1] Bonanni P, Breuer J, Gershon A, Gershon M, Hryniewicz W, Papaevangelou V, et al. Varicella vaccination in Europe-taking the practical approach. BMC Med 2009;7(May):26, http://dx.doi.org/10.1186/1741-7015-7-26. [2] Streng A, Grote V, Carr D, Hagemann C, Liese JG. Varicella routine vaccination and the effects on varicella epidemiology—results from the Bavarian Varicella Surveillance Project (BaVariPro), 2006–2011. BMC Infect Dis 2013;13:303. Available at: http://biomedcentral.com/1471-2334/13/303 [3] Baxter R, Tran NT, Ray P, Lewis E, Fireman B, Black S, et al. Impact of vaccination on the epidemiology of varicella: 1995–2009. Pediatrics 2014;134:24. Accessible at: http://pediatrics.aappublications.org/content/134/24.full.html [4] Carville KS, Riddell MA, Kelly HA. A decline in varicella but an uncertain impact on zoster following varicella vaccination in Victoria, Australia. Vaccine 2010;16(28):2532–8. [5] Siedler A, Arndt U. Impact of the routine varicella vaccination programme on varicella epidemiology in Germany. Eurosurveillance 2010;15, pii=19530. [6] Tan B, Bettinger J, McConnell A, Scheifele D, Halperin S, Vaudry W, et al. Members of the Canadian Immunization Monitoring Program, Active (IMPACT). The effect of funded varicella immunization programs on varicella-related hospitalizations in IMPACT centers, Canada, 2000–2008. Pediatr Infect Dis J 2012;31:956–63.
[7] Pozza F, Piovesan C, Russo F, Bella A, Pezzotti P, Emberti Gialloreti L. Impact of universal vaccination on the epidemiology of varicella in Veneto, Italy. Vaccine 2011;29:9480–7. [8] Moreno-Pérez D, Álvarez García FJ, Arístegui Fernández J, Barrio Corrales F, Cilleruelo Ortega MJ, Corretger Rauet JM, et al. Comité Asesor de Vacunas de ˜ ˜ Immunisation schedule of the de Pediatría, Espana. la Asociación Espanola Spanish Association of Paediatrics: 2013 recommendations. An Pediatr (Barc) 2013;78(Jan (1)):59e1–627e. ˜ [9] Pena-Rey I, Martínez de Aragón MV, Villaverde Hueso A, Terres Arellano M, Alcalde Cabero E, Suárez Rodríguez B. Epidemiology of varicela in Spain in pre and post vaccination periods. Rev Esp Salud Públ 2009;83:711–24. [10] Peiro S, Librero J. Evaluación de la calidad a partir del conjunto mínimo de datos básicos al alta hospitalaria. Rev Neurol 1999;29:651–61. [11] Guillen JM, Gil-Prieto R, Alvaro A, Gil A. Burden of adult varicella hospitalizations in Spain (2001–2007). Hum Vaccin 2010;6(Aug (8)):659–63. [12] González-Escalada A, García-García L, Viguera-Ester P, Marín-García P, García J, Gil de Miguel A, et al. Seroprevalence of antibodies against measles, rubella, mumps, varicella-zoster, and B. Pertussis in young adults of Madrid, Spain. Hum Vaccin Immunother 2013;9(Jun (9)):1918–25. [13] Marshall HS, McIntyre P, Richmond P, Buttery JP, Royle JA, Gold MS, et al. Changes in patterns of hospitalized children with varicella and of associated varicella genotypes following introduction of varicella vaccine in Australia. Pediatr Infect Dis J 2012;(Dec (17)):530–7. [14] García Cenoz M, Castilla J, Chamorro J, Martínez-Baz I, Martínez-Artola V, Irisarri F, et al. Impact of universal two-dose vaccination on varicella epidemiology in Navarre, Spain, 2006 to 2012. Eurosurveillance 2013;18, pii=20552. [15] Ozaki T. Varicella vaccination in Japan: necessity of implementing a routine vaccination program. J Infect Chemother 2013;19(Apr (2)):188–95. [16] Kurugol Z, Halicioglu O, Koc F, Koturoglu G, Aksit S. Varicella rates among unvaccinated and one-dose vaccinated healthy children in Izmir, Turkey. Int J Infect Dis 2011;15(Jul (7)):e475–80. [17] Van Hoek AJ, Melegaro A, Zagheni E, Edmunds WJ, Gay N. Modelling the impact of a combined varicella and zoster vaccination programme on the epidemiology of varicella zoster virus in England. Vaccine 2011;29(Mar (13)):2411–20. [18] Ward K, Dey A, Hull B, Quinn HE, Macartney K, Menzies R. Evaluation of Australia’s varicella vaccination program for children and adolescents. Vaccine 2013;31(Feb (10)):1413–9, http://dx.doi.org/10.1016/j.vaccine.2012.12.052 (Epub 2013 Jan 2). [19] Luman ET, Ching PL, Jumaan AO, Seward JF. Uptake of varicella vaccination among young children in the United States: a success story in eliminating racial and ethnic disparities. Pediatrics 2006;117(Apr (4)):999–1008. [20] García-Altés A. Systematic review of economic evaluation studies: are vaccination programs efficient in Spain? Vaccine 2013;31(Mar (13)):1656–65.