Rotavirus is the leading cause of severe diarrhea in

GASTROENTEROLOGY 2009;137:1970 1975 Incidence of Rotavirus and All-Cause Diarrhea in Northeast Brazil Following the Introduction of a National Vaccination Program RICARDO G. GURGEL,*, ANNA K. BOHLAND,* SARAH C. F. VIEIRA,* DÉBORA M. P. OLIVEIRA,*, PAULA B. FONTES,* VIVIANE F. BARROS,* MARCELA F. RAMOS,* WINIFRED DOVE, TOYOKO NAKAGOMI, OSAMU NAKAGOMI, JAILSON B. CORREIA, NIGEL CUNLIFFE, and LUIS E. CUEVAS,# *Federal University of Sergipe, Aracaju, Brazil; University of Liverpool, Liverpool, United Kingdom; Expanded Programme of Immunization, Department of Health, Aracaju, Sergipe, Brazil; Nagasaki University, Nagasaki, Japan; Instituto de Medicina Integral Professor Fernando Figueira, IMIP, Recife, Brazil; and # Liverpool School of Tropical Medicine, Liverpool, United Kingdom BACKGROUND & AIMS: Rotavirus vaccines were introduced in Brazil in 2006; we evaluated their effects in the state of Sergipe, Brazil. METHODS: We performed a cross-sectional survey of children with diarrhea attending emergency services in Aracaju, Brazil, between October 2006 and April 2008 and a cluster sampling survey to assess vaccination coverage. Vaccine efficacy was assessed using the screening method. Diarrhea consultation and hospitalization data (2003 2007) were obtained from state and national surveillance systems. RESULTS: Rotavirus was detected in 59 of 534 stool samples (11%) from children attending emergency services. The number of rotavirus-positive samples decreased from 18 of 74 (24%) in 2006 to 31 of 321 (9.5%) in 2007 and 10 of 136 (7.4%) in 2008 (P.01). Diarrhea severity was greater in children with rotavirus (P.01) but decreased over time (P.001). Of the rotaviruses detected, 56 of 59 (95%) were P[4]G2 genotype, 1 was P[4]G-non-typeable (NT), 1 was P[NT]G2, and 1 was P[NT]GNT. Diarrhea consultations decreased from 3020 in 2004 to 604 in 2007; reductions were greatest among children under 5 years old. Diarrhea hospitalizations decreased from 2121 in 2003 to 1176 in 2007. Vaccine coverage was 90.3%. Vaccines were highly effective against the strain P[8]G1; efficacy against P[4]G2 genotype was 89% (95% confidence interval: 0.87 0.92) in Aracaju and 95% in Sergipe. CONCLU- SIONS: Since vaccines were introduced in 2006, there has been an overall reduction in diarrhea consultations and hospitalizations in northeast Brazil, with the greatest reductions in young children. This might have resulted from vaccination and improved sanitation. Although a single rotavirus genotype (P[4]G2) was recovered, vaccine efficacy was high against this genotype. Rotavirus is the leading cause of severe diarrhea in infants and young children, causing 600,000 deaths and 2 million hospitalizations each year. 1 Therefore, the adoption of a rotavirus vaccine (Rotarix; GlaxoSmithKline Biologicals, Rixensart, Belgium) by the National Expanded Program of Immunization of Brazil in 2006 represented a landmark in the global effort to reduce the morbidity and mortality associated with diarrheal disease. This human, monovalent serotype P[8]G1 rotavirus vaccine was introduced after its safety and efficacy was demonstrated in a large, multi-country clinical trial, 2 and nearly 4 million doses were delivered in the first 2 years of the program. Monitoring the efficacy of vaccines under operational conditions is important 3,4 to demonstrate the benefit that vaccines could have in the control of diarrhea when delivered outside a research environment 5,6 and to inform their potential introduction in other settings. 7 Initial reports after the introduction of Rotarix in Sergipe State in northeast Brazil suggested that the vaccine had started to reduce the proportion of cases of diarrhea because of rotavirus and that a single genotype (P[4]G2) was predominant. 8,9 The report in Aracaju, however, included only 4 months of surveillance and a small number of rotavirus strains. Despite the large scale of the vaccination program, however, only 1 further report from Rio de Janeiro documented similar changes in the predominant circulating genotypes, 10 and there have been no reports to document changes in the overall and rotavirusspecific incidence of diarrhea or on virus epidemiology. This study therefore aimed to describe the epidemiologic changes observed in all-cause and rotavirus-associated diarrhea episodes in the first 2 years after the vaccine introduction in the State of Sergipe, Brazil, and to describe its efficacy under operational conditions. Patients and Methods The study was based in Aracaju, the capital of Sergipe State, where Rotarix was introduced in March Abbreviations used in this paper: CI, confidence interval; CONEP, Brazilian National Commission for Ethics and Research; NT, nontypeable; PCV, proportion of cases with rotavirus diarrhea that have been vaccinated; PPV, proportion of the population that has been vaccinated; SIVEP-DDA, Epidemiological Surveillance System for Acute Diarrheal Diseases; VE, vaccine efficacy. 2009 by the AGA Institute 0016-5085/09/$36.00 doi:10.1053/j.gastro.2009.07.046

December 2009 ROTAVIRUS DIARRHEA IN NORTHEAST BRAZIL 1971 2006. Aracaju has a population of 0.5 million, representing approximately 25% of the State s population. 11 Information for this study was obtained from several sources. First, the proportion of cases of diarrhea because of rotavirus was ascertained through a prospective survey of children 10 years old presenting with acute diarrhea 14 days duration to the emergency services of Hospital de Urgências de Sergipe in Aracaju between October 2006 and April 2008. This is the largest emergency center in the State and provides 24-hour, free medical care. The study period encompassed the peak season for rotavirus in 2007. For logistical reasons, only patients attending from Monday to Friday between 8 AM and 4 PM who could provide a stool specimen while in the hospital were included. Approximately 40% of the patients eligible to participate were enrolled, and it is thus likely that children with severe diarrhea are overrepresented. In total, 534 children were recruited, with an average of 30 children per month. Children were prospectively assessed to describe the medical history, clinical presentation, and severity of the episodes using a standardized score as previously described. 12 Because most parents carry their children s vaccination cards when attending health services and the rotavirus vaccine could be confused with the oral polio vaccine, rotavirus vaccination status was only accepted if verified against the vaccination card. Stool samples collected were stored frozen at 70 C until tested for the presence of rotavirus by enzymelinked immunosorbent assay (Rotaclone; Meridian Diagnostics, Cincinnati, OH). Rotavirus genotypes were determined by hemi-nested reverse transcription-polymerase chain reaction using consensus and type-specific primers as previously described. 9 The number of diarrhea consultations to the health services of Aracaju was obtained from Sergipe s diarrhea surveillance system (Epidemiological Surveillance System for Acute Diarrheal Diseases [SIVEP-DDA]) from January 2004 to December 2007. SIVEP-DDA obtains weekly information on the number of patients attending the health centers of Aracaju by age. Although the system only receives information from approximately 10 sentinel sites, it was strengthened and digitized in preparation for the introduction of the vaccine, and it is likely that its sensitivity has increased in the last few years. The data reported by SIVEP- DDA were combined with census population figures to calculate incidence by age group per 1000 population. In addition, the number of diarrhea hospitalizations in Brazil from 2003 to 2007 was obtained from the Ministry of Health web site (www.datasus.gov.br). Rotavirus vaccination coverage for Aracaju was ascertained through a stratified household cluster sampling survey 6 of 770 children 6 weeks and 18 months old in June 2007. These age cutoffs were selected to include children eligible for rotavirus vaccination because the vaccine introduction in March 2006 and a larger sample size ( 3) were used to obtain smaller confidence limits than routine cluster sampling surveys. A total of 110 clusters were selected randomly from a list of all Agentes de Saude (community health workers) in Aracaju city. These workers belong to the Program Saude da Familia (Family Health Program) and regularly visit families residing in their assigned geographic area to undertake preventive and curative health activities. The Program Saude da Familia has approximately 94% population coverage, with poorest areas being overrepresented because they receive priority for program implementation. Each community health worker is responsible for an average of 150 families (approximately 600 residents). The households of the clusters selected were visited consecutively until 7 children had been enrolled as required for cluster sampling surveys. 6 The immunization cards of the children were reviewed, and children with 2 doses of the vaccine were considered fully vaccinated. In addition, rotavirus vaccination coverage in Aracaju, Sergipe, and Brazil in 2006 and 2007 was obtained from the National Expanded Program of Immunizations web site (http:// pni.datasus.gov.br/inf_estatistica_cobertura.asp). Data were analyzed using descriptive statistics, with 95% confidence intervals (CIs), and 2 and Student t tests, as appropriate. Vaccine efficacy (VE) was assessed using the screening method. This method requires ascertaining the proportion of cases with rotavirus diarrhea that have been vaccinated (PCV) and the proportion of the population that has been vaccinated (PPV). The PCV was obtained from the cross-sectional survey in Hospital de Urgências de Sergipe, and the PPV was ascertained Table 1. Number of Children Attending the Emergency Hospital in 2006 2008 With Rotavirus by Age and Vaccination Status Age, y Rotavirus ELISA Vaccination status, n (%) Vaccinated Not vaccinated Not known All 1 Positive 12 (10%) 14 (11%) 3 (13) 29 (11%) Negative 113 (90) 110 (89) 24 (87) 247 (89) 1 2 Positive 4 (13) 12 (23) 3 (14) 19 (18) Negative 28 (88) 41 (77) 18 (86) 87 (88) 2 Positive 0 (0) 11 (8) 0 (0) 11 (7) Negative 7 (100) 125 (92) 9 (100) 141 (93) All Positive 16 (10) 37 (12) 6 (11) 59 (11) Negative 148 (90) 276 (88) 51 (89) 475 (89)

1972 GURGEL ET AL GASTROENTEROLOGY Vol. 137, No. 6 through the cluster sampling survey (June 2007) and the Expanded Program of Immunizations. VE was calculated using the formula VE (PPV PCV)/PPV (1 PCV). With this formula, 95% confidence limits of the VE are calculated using the upper and lower 95% confidence limits of the vaccine coverage estimate. 6 The study protocol was approved by the research ethics committees of the Liverpool School of Tropical Medicine, Sergipe s Federal University, and the Brazilian National Commission for Ethics and Research (CONEP). Results Rotavirus was detected in 59 (11%) of 534 stool samples obtained from children attending the Hospital de Urgências de Sergipe between October 2006 and April 2008 as described in Table 1. Twenty-one of the children with rotavirus had been identified between October 2006 and January 2007 as previously reported 9 but are included here for completeness of the analysis. Fifty-six (95%) of rotaviruses identified were genotype P[4]G2, 1 was P[4]G-non-typeable (NT), 1 was P[NT]G4, and 1 was P[NT]GNT. The proportion of stool specimens with rotavirus decreased from 18 (24%) of 74 samples in 2006 to 31 (9.5%) of 325 and 10 (7.4%) of 136 samples in 2007 and 2008, respectively (P.01). The proportion of rotavirus-positive samples was higher from October to December 2006 and from June to November 2007 (Figure 1). Children 1 and 2 years old were less likely to have rotavirus (29 of 276 [11%] children 1 and 11 of 152 [7%] children 2 years old) than 1 2 year olds (19 of 106 [18%], P.03 and P.01, respectively). Nearly all children (533, 99%) had data available for the assessment of diarrhea severity, 12 and their scores are shown in Table 2, stratified for children with and without rotavirus. Children with rotavirus were more likely to have higher scores than children without rotavirus with Table 2. Severity Scores of Children With and Without Rotavirus Severity score Rotavirus negative, n (Cumulative %) Rotavirus positive, n (Cumulative %) 2 6 (1.3) 0 (0.0) 3 16 (4.6) 1 (1.7) 4 28 (10.5) 1 (3.4) 5 22 (15.2) 2 (6.9) 6 42 (24.0) 1 (8.6) 7 51 (34.7) 8 (22.4) 8 32 (41.5) 5 (31.0) 9 39 (49.7) 3 (36.2) 10 39 (57.9) 4 (43.1) 11 40 (66.3) 3 (48.3) 12 30 (72.6) 3 (53.4) 13 35 (80.0) 3 (58.6) 14 35 (87.4) 5 (67.2) 15 18 (91.2) 3 (72.4) 16 17 (94.7) 3 (77.6) 17 10 (96.8) 6 (87.9) 18 9 (98.7) 6 (98.3) 19 6 (100.0) 1 (100.0) medians of 12 and 10, respectively (P.01). Of these, 190 (35.6%) had scores 12 (severe) and 343 (64.4%) scores 12 (mild/moderate diarrhea). Thirty (51%) of 59 children with rotavirus had a score 12 compared with 160 (34%) of 474 children without rotavirus (P.01), with mean (SD) scores of 11.8 (4.4) and 9.7 (4.0), respectively (P.01). The severity of the rotavirus diarrhea episodes varied over the study period, with children in 2006 being more likely to have scores 12 than children with rotavirus in 2007 and 2008 (Figure 2) ( 2 for trend, P.001). One hundred sixty-four (31%) children attending the hospital had been vaccinated, 313 (59%) had not been vaccinated, and 57 (10%) did not bring their vaccination cards. Children 2 years old were more likely to be vaccinated (157/382 [41%]) than older children (7/152 [5%], P.001). Among the 53 children with rotavirus and vaccination information, 7 (44%) of 16 vaccinated, Figure 1. Number of stool samples collected each month and positive for rotavirus in Aracaju. The superimposed line describes the percentage of samples positive for rotavirus. 2 For trend, P.003. Figure 2. Number of children with mild/moderate and severe rotavirus diarrhea in Aracaju, 2006 2008.

December 2009 ROTAVIRUS DIARRHEA IN NORTHEAST BRAZIL 1973 Table 4. Rotavirus Vaccine Coverage and Number of Doses Administered in Aracaju, Sergipe, and Brazil Aracaju Sergipe Brazil Year Number of doses Vaccine coverage (%) Number of doses Vaccine coverage (%) Number of doses Vaccine coverage (%) 2006 3472 37.0 19,256 51.5 1,373,246 45.3 2007 7965 84.6 33,159 90.3 2,306,181 80.1 NOTE. Source: http://pni.datasus.gov.br/inf_estatistica_cobertura. asp. Figure 3. Number of cases of diarrhea reported to Sergipe s Diarrhoea Surveillance System (SIVEP-DDA) from 2004 to 2007, by age. and 19 (51%) of 37 unvaccinated children had scores 12 (P.5). The number of diarrhea consultations in Aracaju (Sergipe s SIVEP-DDA) decreased year on year from 3020 (139/1000 population) in 2004 to 604 (14/1000 population) in 2007 (Figure 3). Children 5 years old experienced larger reductions than older children, representing 69% (2036/2952) of all consultations in 2004 and 44% (268/604) in 2007 (P.001). Remarkably, the highest number of consultations in 2007 was reported in individuals 10 years old. The number of diarrhea hospitalizations between 2003 and 2007 is shown in Table 3, and the number of hospitalizations because of acute respiratory infections is included for comparison. A total of 712,333 diarrhea hospitalizations were reported for the country in the 5-year period. In agreement with the reduction in consultations, the number of hospitalizations decreased over the same period, with larger reductions in Aracaju and Sergipe than in the country as a whole. This decline preceded the introduction of the vaccine, and, in 2007, the number of hospitalizations in Aracaju and Sergipe only reached 56% and 45%, respectively, of the numbers reported in 2003. In comparison, the number of hospitalizations because of acute respiratory infections has declined by a lower percentage from 2003 to 2007 ( 7% in Aracaju and 16% nationally), and the decline has not accelerated in the last 2 years, as observed with diarrhea incidence. The household cluster sampling survey in Aracaju established that the rotavirus vaccination coverage was 81% (95% CI, 78% 84%) in 2007. The number of vaccine doses administered and vaccine coverage reported by the Ministry of Health in 2006 and 2007 are shown in Table 4. Vaccination coverage in Aracaju reached 37% in 2006 and 84.6% in 2007. Vaccine coverage for Sergipe State was higher than for Aracaju City, and both estimates were higher than the national average in 2007. The estimation of VE needs to consider the low number of rotaviruses identified, the unusual presence of a single genotype, and the varying vaccination coverage during the study period. Only 1 child had an infection with a rotavirus other than the G2 or P[4], and the absence of other genotypes was observed in both vaccinated and unvaccinated children. VE for genotypes other than G2P[4], therefore, could be close to 100%, provided these changes are not due to natural fluctuation of serotypes. Using the screening method for VE against the G2P[4] genotype, 53 of 59 children with rotavirus had vaccination information. Of these, 16 (30%) had received the vaccine (PCV 0.3). The PPV was 81% (95% CI: 78% 84%) in Aracaju (cluster sampling survey) and 90.3% for Sergipe State (Ministry of Health). Replacing these Table 3. Number of Hospitalizations Because of Diarrhea and Acute Respiratory Infections Reported by the National Health Service in Aracaju, Sergipe State, and Brazil Year Diarrhea, n (%) Acute respiratory infections, n (%) Aracaju Sergipe Brazil Aracaju Sergipe Brazil 2003 605 (100) 2121 (100) 128,529 (100) 1537 (100) 4268 (100) 387,674 (100) 2004 494 (82) 2111 (100) 121,521 (95) 1348 (88) 3594 (84) 355,072 (92) 2005 536 (89) 2317 (109) 123,647 (96) 1319 (86) 2900 (67) 325,317 (84) 2006 465 (78) 1909 (90) 123,805 (94) 1245 (81) 2686 (43) 325,855 (84) 2007 280 (47) 1176 (55) 92,701 (72) 1434 (93) 3088 (72) 325,901 (84) Total 2380 9634 712,333 6883 16,536 1,719,819 NOTE. Baseline 2003 (100%).

1974 GURGEL ET AL GASTROENTEROLOGY Vol. 137, No. 6 values in the formula (VE (PPV PCV)/PPV (1 PCV) results in a VE against G2P[4] of 89% (95% CI: 0.87% 0.92%) in Aracaju and 95% in Sergipe State. If the calculation of VE only includes children eligible to receive the vaccine because of age, then 15 (48%) of 31 children with rotavirus were vaccinated and VE would be 79%. Discussion This is the most comprehensive report of changes in the epidemiology of rotavirus and non-rotavirus diarrhea since the introduction of the vaccine in northeast Brazil and the first attempt to assess VE under operational conditions in Brazil. The data collected indicate that there has been an overall significant reduction in the number of rotavirus and non-rotavirus diarrhea consultations and hospitalizations. These reductions were larger in Sergipe than at the national level and seem to have been more pronounced in children under 2 years old. The discrepant falling rates between Aracaju and Sergipe and the rest of Brazil are difficult to explain. The northeast of Brazil is the most economically deprived area of the country and has received priority for development by recent Federal governments. Diarrhea incidence (as is the case for infant mortality and other health indicators) is higher in this region, and infectious diseases remain prevalent, whereas morbidity/mortality patterns in the South resemble those of industrialized countries. Thus, although apparently paradoxical, it is possible that areas with high background incidence of diarrhea are experiencing higher reductions than areas with very low rates. However, this needs to be confirmed with longer data series. The highest number of diarrhea consultations in 2007 surprisingly occurred in individuals 10 years old. These changes follow a long-term decline in the incidence and mortality of diarrhea, which was first reported from Sergipe in 1992 13 when the city of Aracaju was undertaking a rapid improvement of its water and sanitation services. At that time, the declining rates were attributed to improved environmental hygiene and increased access to oral rehydration solutions, 13 and it is not possible to solely attribute the reductions in rotavirus and nonrotavirus diarrhea consultations and hospitalizations in the last 2 years to the vaccine. The proportion of children with rotavirus diarrhea and the severity of these episodes have decreased substantially in the last 2 years. Rotavirus causes severe diarrhea and is usually overrepresented in hospital-based studies worldwide, 14,15 and Brazil was no exception because most studies before the introduction of the vaccine reported that rotavirus was the most frequent pathogen recovered in hospitalized children. 16 It is thus remarkable that only 7% of the children attending the reference hospital in 2008 had rotavirus. The virus seems to continue having seasonal fluctuations, with peaks occurring in the second half of the year and a poor association with rain and temperature (data not shown). Although the period reported is too short to exclude interannual and seasonal variation, rotavirus incidence seems to have decreased with increasing vaccination coverage, especially among rotaviruses possessing the G1 or P[8] types shared with the vaccine. The same predominance of P[4]G2 and virtual disappearance of all other genotypes were reported from a similar study in a pediatric reference hospital in Recife, about 500 km northeast of Aracaju. 8 The reduction in rotavirus disease burden also seems to be higher than would have been predicted from vaccine performance in clinical trials. 2 It has earlier been suggested that the true burden of rotavirus is underestimated 17 and that a high vaccination coverage may lead to a reduction of circulating rotaviruses in the environment, reduce vulnerability through a reduction in exposure, and result in larger reductions than expected, as recently reported from the United States. 5 VE for genotypes other than G2P[4] therefore could be higher than expected, but, unless the vaccine has effected a high degree of herd immunity or an unusual seasonal variation, it is difficult to explain the complete absence of other rotavirus genotypes in unvaccinated children. The continued identification of predominantly G2P[4] strains over the 18 months period since vaccine introduction is also unprecedented. This genotype has become predominant in Central and South America in recent years and constituted the majority of the strains recovered in the north of Brazil. 18 Thus, it remains possible that our observations in Aracaju 9 and Recife 8 may be due to natural variation of genotypes over time. 18,19 An alternative possibility, however, is that Rotarix is relatively less efficacious against P[4]G2 strains. Rotarix is a monovalent vaccine derived from a P[8]G1 strain with an efficacy of 91% against severe rotavirus episodes caused by P[8]G1 strains and 87% against strains sharing only the P[8] antigen. 2 Cross protection against other strains has been documented, and its efficacy against P[4]G2 was reported to be 45% in Latin American children (although the numbers of G2 strains were very small), and efficacy reached 75% against severe rotavirus diarrhea in European children. 2,20 Although the recent predominance of P[4]G2 in northeastern Brazil is striking, the number of P[4]G2 strains in Aracaju has continued to decline in the most recent months of the study, and continued surveillance is necessary to elucidate whether they might eventually disappear. The reduced number, together with the reduced severity of the rotavirus episodes over the 18 months period (because a higher proportion of the cases in 2006 had severe diarrhea than in 2007 and 2008), and the high VE against P[4]G2 strains documented by the screening method suggest that the burden of rotavirus diarrhea is decreasing in Sergipe. This study thus reports substantial reductions in the burden of rotavirus and non-rotavirus diarrhea in northeast Brazil. These are likely to be the result of long-term

December 2009 ROTAVIRUS DIARRHEA IN NORTHEAST BRAZIL 1975 improvements in the water and sewage systems 13 and the synergistic effect of the rotavirus vaccine in the last 2 years. This study also highlights the difficulty of interpreting these epidemiologic changes without good background data of the incidence of diarrhea before vaccine introduction and during postmarketing surveillance. Clinical trials assessing the efficacy of rotavirus vaccines are near completion in Africa and Asia, and the vaccines are likely to be introduced in other immunization programs in these continents. Brazil s pioneering adoption of the Rotarix vaccine and the high vaccination coverage attained over a short period could demonstrate the high potential of these vaccines to control rotavirus diarrhea in the future. This information, however, will only emerge if continuous postmarketing surveillance is effectively maintained over the following years. References 1. Parashar UD, Gibson CJ, Bresse JS, et al. Rotavirus and severe childhood diarrhea. Emerg Infect Dis 2006;12:304 306. 2. Ruiz-Palacios GM, Perez-Schael I, Velazquez FR, et al. Safety and efficacy of an attenuated vaccine against severe rotavirus gastroenteritis. N Engl J Med 2006;354:11 22. 3. de Oliveira LH, Danovaro-Holliday MC, Matus CR, et al. Rotavirus vaccine introduction in the Americas: progress and lessons learned. Expert Rev Vaccines 2008;7:345 353. 4. Grimwood K, Buttery JP. Clinical update: rotavirus gastroenteritis and its prevention. Lancet 2007;370:302 304. 5. CDC. Delayed onset and diminished magnitude of rotavirus activity United States, November 2007-May 2008. MMWR Morb Mortal Wkly Rep 2008;57:697 700. 6. Orenstein WA, Bernier RH, Dondero TJ, et al. Field evaluation of vaccine efficacy. Bull World Health Organ 1985;63:1055 1068. 7. Rotavirus vaccines. Wkly Epidemiol Rec 2007;82:285 295. 8. Nakagomi T, Cuevas LE, Gurgel RG, et al. Apparent extinction of non-g2 rotavirus strains from circulation in Recife, Brazil, after the introduction of rotavirus vaccine. Arch Virol 2008;153:591 593. 9. Gurgel R, Cuevas L, Vieira S, et al. Predominance of P[4]G2 rotavirus in a vaccinated population in Brazil. Emerg Infect Dis 2007;13:3. 10. Carvalho-Costa FA, Araujo IT, Santos de Assis RM, et al. Rotavirus genotype distribution after vaccine introduction, Rio de Janeiro, Brazil. Emerg Infect Dis 2009;15:95 97. 11. IBGE IBdGeE. Censo Demográfico 2002. In: Estatística IBdGe, editor. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2008. 12. Nakagomi T, Nakagomi O, Takahashi Y, et al. Incidence and burden of rotavirus gastroenteritis in Japan, as estimated from a prospective sentinel hospital study. J Infect Dis 2005;192(Suppl 1):S106 S110. 13. Gurgel RQ, Andrade JM, Machado-Neto P, et al. Diarrhea mortality in Aracaju, Brazil. Ann Trop Paediatr 1997;17:361 365. 14. Vernacchio L, Vezina RM, Mitchell AA, et al. Diarrhea in American infants and young children in the community setting: incidence, clinical presentation and microbiology. Pediatr Infect Dis J 2006; 25:2 7. 15. Banerjee I, Ramani S, Primrose B, et al. Comparative study of the epidemiology of rotavirus in children from a community-based birth cohort and a hospital in South India. J Clin Microbiol 2006; 44:2468 2474. 16. Gurgel RQ, Cunliffe NA, Nakagomi O, et al. Rotavirus genotypes circulating in Brazil before national rotavirus vaccination: a review. J Clin Virol 2008;43:1 8. 17. Angel J, Franco MA, Greenberg HB. Rotavirus vaccines: recent developments and future considerations. Nat Rev Microbiol 2007;5:529 539. 18. Oliveira A, Mascarenhas JDP, Soares LS, et al. Re-emergence of G2 Rotavirus in Northern Brazil reflects a natural changing pattern over time; 2008 3-4/05/08. Istanbul, Turkey. Rotavirus Symposium;2008:60 61. 19. Patel MM, de Oliveira LH, Bispo AM, et al. Rotavirus P[4]G2 in a vaccinated population, Brazil. Emerg Infect Dis 2008;14:863 865. 20. Vesikari T, Itzler R, Matson DO, et al. Efficacy of a pentavalent rotavirus vaccine in reducing rotavirus-associated health care utilization across three regions (11 countries). Int J Infect Dis 2007;11(Suppl 2):S29 S35. Received January 6, 2009. Accepted July 10, 2009. Reprint requests Address requests for reprints to: Luis E. Cuevas, MD, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA, United Kingdom. e-mail: lcuevas@liv.ac.uk; fax: (44) 1517 053 219. Conflicts of interest The authors disclose the following: N. Cunliffe is the principal investigator on a clinical trial of Rotarix in Malawi. The remaining authors disclose no conflicts. Funding R.Q. Gurgel received a traveling scholarship from the Ministry of Education of Brazil (CAPES) to travel to the University of Liverpool, United Kingdom, in 2007. L.E. Cuevas is co-investigator on a research grant to fund the studies awarded by the Brazilian CNPq (grant 476841/2006-2) to Prof Gurgel. The study sponsor did not play any role in the study design, collection, analysis, or interpretation of data.