Population-Based Multicentric Survey of Hepatitis B Infection and Risk Factor Differences among Three Regions in Brazil

Am. J. Trop. Med. Hyg., 81(2), 2009, pp. 240 247 Copyright 2009 by The American Society of Tropical Medicine and Hygiene Population-Based Multicentric Survey of Hepatitis B Infection and Risk Factor Differences among Three Regions in Brazil Leila M. M. B. Pereira, * Celina M. T. Martelli, Edgar Merchán-Hamann, Ulisses R. Montarroyos, Maria C. Braga, Maria L. C. de Lima, Maria R. A. Cardoso, Marília D. Turchi, Marcelo A. Costa, Luiz C. A. de Alencar, Regina C. Moreira, Gerusa M. Figueiredo, and Ricardo A. A. Ximenes for the Hepatitis Study Group Faculdade de Ciências Médicas de Pernambuco, Hospital Universitário Oswaldo Cruz, Universidade de Pernambuco, Recife, Pernambuco, Brazil; Instituto de Patologia Tropical e Saúde Publica, Departamento de Saúde Coletiva, Universidade Federal de Goiás, Goiania, Goiás, Brazil; Faculdade de Ciências da Saúde, Departamento de Saúde Coletiva, Universidade de Brasília, Distrito Federal, Brasília, Brazil; Centro de Pesquisas Aggeu Magalhães, Fundação Oswaldo Cruz, Recife, Pernambuco, Brazil; Departamento Medicina Social/Núcleo de Saúde Pública, e Departamento de Medicina Tropical, Universidade Federal de Pernambuco, Recife, Pernambuco, Brazil; Faculdade de Saúde Publica, Departamento de Epidemiologia, Universidade de São Paulo, São Paulo, São Paulo, Brazil; Hospital de Base do DF, Brasília, Distrito Federal, Brasília, Brazil; Instituto Adolfo Lutz, São Paulo, São Paulo, Brazil; Programa Nacional de Prevenção e Controle das Hepatites Virais, Ministério da Saúde, Brasília, Distrito Federal, Brazil; Instituto do Fígado de Pernambuco, Universidade de Pernambuco, Recife, Pernambuco, Brazil Abstract. This multicentric population-based study in Brazil is the first national effort to estimate the prevalence of hepatitis B (HBV) and risk factors in the capital cities of the Northeast, Central-West, and Federal Districts (2004 2005). Random multistage cluster sampling was used to select persons 13 69 years of age. Markers for HBV were tested by enzyme-linked immunosorbent assay. The HBV genotypes were determined by sequencing hepatitis B surface antigen (HBsAg). Multivariate analyses and simple catalytic model were performed. Overall, 7,881 persons were included; < 70% were not vaccinated. Positivity for HBsAg was less than 1% among non-vaccinated persons and genotypes A, D, and F cocirculated. The incidence of infection increased with age with similar force of infection in all regions. Males and persons having initiated sexual activity were associated with HBV infection in the two settings; healthcare jobs and prior hospitalization were risk factors in the Federal District. Our survey classified these regions as areas with HBV endemicity and highlighted the risk factors differences among the settings. INTRODUCTION * Address correspondence to Leila M. M. B. Pereira, Faculdade de Ciências Médicas de Pernambuco, Universidade de Pernambuco, Recife, CEP 50100130, Brazil. E-mail: leilapereira@pq.cnpq.br 240 Hepatitis B virus (HBV) infection is a major health problem and causes significant levels of morbidity and mortality worldwide. The World Health Organization estimates that two billion people have at least one marker of HBV infection, and 350 millions are chronically infected and at risk of developing chronic liver disease and hepatocellular carcinoma. 1,2 Sexual activity, perinatal transmission, health-care activities, non-sexual household contact, and needle-sharing, are some of the well-documented HBV transmission routes. 2,3 In developing countries, comprising most of Asia, sub-saharan Africa, the Pacific and the Amazon region, more than 8% of persons are chronically infected in the general population, with a high rate of infection during childhood. 4 In contrast, in areas of low/intermediate endemicity, most HBV infections occur during adolescence and adult life. 2 The HBV vaccine was first licensed in the United States in 1981 and has proved to be a highly effective and safe vaccine that is widely used as part of the pediatric vaccination schedule in most countries. 2,5 Most of Brazil is classified as having intermediate endemicity for HBV infection, with the main exception being the region of high HBV endemicity in the western Amazon region where hepatitis delta virus also circulates. 6 10 In 1989, the Brazilian government first implemented immunization against hepatitis B for infants and children in the western Amazon region and gradually expanded this to other regions. Vaccination of risk groups for the entire country started in 1992, and new groups have been successively added to the original list. Since 1998, HBV vaccine has been incorporated into the immunization schedule for infants as a national policy; in 2001, this was broadened to include children and adolescents. 11 The epidemiology of HBV infection in Brazil has been assessed mainly using data from blood donors, health-care workers, hemodialysis and schistosomiasis patients, 12,13 pregnant women, and drug users, 14 16 with few studies conducted in restricted areas. 7,8,17 19 Therefore, population-based data for HBV infection are lacking for most areas of the country. The present study is part of an ongoing population-based hepatitis survey aimed at estimating the prevalence and predictive factors for HBV infection in all state capitals in the Northeast and Central-West regions and in the Federal capital of Brazil. MATERIALS AND METHODS Study populations and sample design. The Brazilian National Hepatitis A, B, and C survey is a cross-sectional household investigation in a representative sample of the population. We report on the hepatitis B survey conducted in the Northeast region (nine capital cities), the Central-West region (three capital cities), and the Federal District of Brazil, also located in the Central-West region, during September 2004 June 2005 ( Figure 1 ). The Northeast and Central-West regions contain approximately 36% of the country s total population, with 51.5 and 13.2 million inhabitants, respectively. For the target age group (13 69 years of age), the population was 32.4 millions in the Northeast region, 8.3 million the Central-West region, and 1.5 million inhabitants for the Federal District. The inclusion criterion was persons 13 69 years of age residing in the capital cities. More detailed information on sampling and data collection has been reported. 20 Briefly, the study population was stratified in two age strata (13 19 years) and (20 69 years) and by state capital in each area, using a constant sampling fraction. A random sample was obtained using

HEPATITIS B INFECTION AND RISK ASSESSMENT 241 Figure 1. Location of study areas in Brazil. a multistage cluster sampling strategy based on census tract, which is the smallest geographic unit with sociodemographic data available (approximately 300 households corresponding to 1,000 inhabitants). Census tracts were selected by systematic sampling (with probability proportional to size) to represent all socioeconomic groups measured by the mean number of schooling of the head of the household. In this second stage, blocks were systematically selected with probability proportional to size, and a systematic sampling of households was drawn in each block. All residents in the age group considered were recruited in the selected households. 20 A total of 7,881 individuals were enrolled in the study: 3,077 persons in 1,504 households living in 254 census tract in the Northeast region, 3,152 persons in 1,595 houses located in 94 census tracts in the Central-West region, and 1,652 persons in 787 households and 31 census tracts in the Federal District. Data collection. Individual data on socioeconomic status, sexual behavior, blood-related procedures, alcohol consumption and drug use were collected by trained interviewers using a structured questionnaire during household visits. Selfreported HBV vaccine status was assessed and confirmed by a vaccination card, when available. Laboratory tests and definitions. Blood samples were collected after the interview and specimens were tested for antibodies against hepatitis B core antigen (HBc) using an enzyme-linked immunosorbent assay (Axsym ; Abbott Laboratories, Abbott Park, IL) in central public health laboratories. Positive results were considered to indicate current or past HBV infection. Samples positive for HBc were tested for antibodies against hepatitis B surface antigen (HBs). Samples negative for HBs were tested for hepatitis B surface antigen (HBsAg). Both tests were performed using the same technique (Axsym ). Borderline results were retested and indeterminate results were excluded from the analysis. Chronic HBV infection was determined by the presence of HBsAg and antibody against HBc. The presence of HBV infection was determined by detection of antibody against HBc in accordance with the National Health and Nutritional examination survey. 21 Samples positive for HBsAg were tested for HBV DNA and genotyped. A nested polymerase chain reaction was carried out as described by Kaneko and others 22 with some modifications. The S region was amplified according to a standard technique 23 for identification of the HBV genotype. Sequencing reactions for characterization of viral strains were performed using a polymerase chain reaction according to a method previously reported. 24 Genotyping were conducted by comparison of sequences obtained with other known sequences from different HBV genotypes deposited in the GeneBank by using the Edit Seq and MegAlign programs in Lasergene software (DNAStar, Madison, WI). Statistical analysis. Data analysis was performed separately according to region (Northeast, Central-West, and Federal District). Prevalence and 95% confidence intervals (CIs) were calculated, stratified according to vaccination status, and corrected by design effect. The force of infection in the three regions was estimated from HBV age-specific seroprevalence by catalytic modeling. The simple catalytic model was used to calculate incidence, immunoconversion, and/or reversion rates from cross-sectional data. The underline assumption is that the current age-specific patterns reflect the effect of age-independence incidence. In such a context, the incidence is equivalent to an incidence rate calculated with person-time denominator and denominated force of infection. 25 Risk factor analysis included the population at risk for HBV infection (non-vaccinated individuals). The association between positivity for antibodies against HBc and individual variables was assessed by calculating odds ratios (ORs) with a 95% CI and P value. Variables associated with the outcome in univariate analyses with P < 0.20 were successively included in a multivariate logistic regression model. In univariate and multivariate analysis, the ORs were corrected using the random design effect caused by the study design and weighting caused by the difference in sampling fraction between age groups. Stata version 9.2 statistical software (Statacorp, College Station, TX) was used for statistical analysis. Sample size. The sample size was calculated on the basis of an expected prevalence of seropositivity for HBsAg for each region based on previous studies (Northeast = 0.2%, Central- West = 1.04%, and Federal District = 0.09%). 26,27 The variance was estimated as 20% of the prevalence value for the Northeast and Central-West regions and 30% of the prevalence value for the Federal District. Ethical issues. Interviews and blood sample collection were performed after persons signed a written consent form. For minors, the legal guardian s consent was obtained. The project was reviewed and approved by the National Research Ethics Committee of the Brazilian National Health Council and the local research ethics committees. Each positive result was delivered by the medical coordinator, and all negative results were sent by mail to participants. Individuals who were positive for HBsAg were referred to

242 PEREIRA AND OTHERS hepatology outpatient clinics in public referral hospitals for clinical evaluation. Persons negative for HBV were referred to public health services for vaccination. RESULTS Prevalence. Of 7,881 individuals investigated, approximately 30% reported HBV vaccination. Residents of the Central- West region, including the Federal District, reported higher frequencies of HBV vaccination (35%) compared with persons in the Northeast region (25%). Vaccination status was reported by approximately 80% of the population in all regions. The prevalence of HBV infection among non-vaccinated individuals was approximately 10%, and the prevalence of infection among vaccinated individuals was 4.3%. Table 1 shows the prevalence of antibodies against HBc according to vaccination status, stratified according to region. Among non-vaccinated groups, the prevalence of infection varied from 8.1% (95% CI = 6.2 9.9%) to 10.3% (95% CI = 8.4 12.0%) in the Federal District and Central-West region; differences were not statistically significant. Higher frequencies of infection were consistently found among nonvaccinated groups than among vaccinated groups in all areas. Higher prevalence of HBV infection was detected among males and among older age groups in all regions. Estimated prevalence of hepatitis B infection was higher for males than for females in all regions. Vaccinated individuals had lower median values for age than non-vaccinated individuals in all regions: 19 and 28 years for the Northeast region, 18 and 32 years for the Central-West region, and 13 and 32 years for the Federal District. Of the 2,453 participants who presented a vaccination card, the positive and negative predictive values of self-reported HBV vaccination status were 93.7% (95% CI = 92.5 94.7%) and 95.0% (95% CI = 92.7 96.6%), respectively, with similar figures for all regions. The simple catalytic model for non-vaccinated individuals shows that the incidence of infection increases with age, with similar force of infection in all regions ( Figure 2 ). The prevalence of HBV infection increased with age for vaccinated and Figure 2. Prevalence of hepatitis B viral infection data using a simple catalytic model, Northeast, Central-West, and Federal District, Brazil, 2005. non-vaccinated groups, showing a similar trend among the regions. The prevalence of HBsAg for the non-vaccinated population 13 69 years of age was 0.19% (95% CI = 0.02 0.36%) for the Northeast region, 0.47% (95% CI = 0.13 0.81%) for the Central-West region, and 0.60% (95% CI = 0.01 1.19%) for the Federal District. There was no statistically significant difference among the settings. Genotypes A and F were identified in the Northeast region and genotypes A and D were identified in the Federal District; genotypes A, D, and F co-circulate in the Central-West region. Genotype A was predominant (approximately 50%), followed by genotypes D and F. Bivariate analysis. Table 2 shows individual risk factors associated with hepatitis B infection adjusted for age. Male sex was consistently associated with infection across the three regions (OR ~ 2.0). In the Northeast region, social and economic status, as expressed by schooling and literacy, was a predictor of HBV infection. Tattooing was also related to the antibody marker (OR = 2.2, 95% CI = 1.2 4.3). In the Central-West region, a low level of education and having been employed in the past week were associated with HBV infection, as well as hospitalization in the past 12 months (OR = 1.95, 95% CI = 1.0 3.7). In the Federal District, Table 1 Prevalence of hepatitis B viral infection (antibody to hepatitis C core antigen) by region, age, and sex according to vaccination status, Brazilian Viral Hepatitis National Survey, 2005* No. Prevalence (95% CI) No. Prevalence (95% CI) No. Prevalence (95% CI) Total 3,077 5.5 (4.6 6.5) 3,152 5.3 (4.6 6.1) 1,652 3.8 (2.7 4.9) Non-vaccinated 1,614 9.8 (8.1 11.4) 1,519 10.3 (8.4 12.0) 790 8.1 (6.2 9.9) Age group, years 13 19 567 3.5 (1.9 5.1) 407 2.2 (0.8 3.5) 188 1.6 (0.0 3.4) 20 69 1,047 13.1 (10.9 15.4) 1,112 13.2 (10.9 15.5) 602 10.1 (7.8 12.4) Sex Female 849 8.1 (6.30 9.9) 798 8.3 (6.4 10.1) 429 4.7 (2.4 6.9) Male 765 11.7 (9.2 14.1) 721 12.5 (10.0 15.0) 361 12.2 (8.9 15.4) Vaccinated 787 4.7 (3.3 6.1) 1,088 5.0 (3.7 6.3) 590 2.5 (0.9 4.2) Age groups 13 19 444 1.3 (0.1 2.6) 405 1.5 (0.6 2.3) 372 1.3 (0.0 2.7) 20 69 343 9.0 (6.2 11.9) 683 11.1 (7.9 14.3) 218 4.6 (1.4 7.7) Sex Female 489 4.0 (2.2 5.9) 634 4.6 (3.2 6.0) 341 2.1 (0.3 3.8) Male 298 5.7 (3.2 8.2) 454 5.7 (3.2 8.3) 249 3.2 (0.9 5.6) Unknown vaccine status 676 6.6 (4.7 8.6) 545 8.8 (6.4 11.1) 272 5.1 (2.6 7.6) * CI = confidence interval. Total population by regions with weighted prevalence adjusted for random effect and age.

HEPATITIS B INFECTION AND RISK ASSESSMENT 243 Table 2 Relative odds of hepatitis B infection for individual factors among non-vaccinated persons 13 69 years of age in three regions in Brazil, 2005* No. OR (95% CI) P No. OR (95% CI) P No. OR (95% CI) P Sex Female 855 1.00 803 1.00 429 1.00 Male 765 1.86 (1.32 2.61) < 0.0001 721 2.01 (1.45 2.79) < 0.0001 361 2.69 (1.41 5.14) < 0.0001 Literacy Yes 1,498 1.00 1,473 1.00 763 1.00 No 122 2.29 (1.20 4.36) 0.012 51 1.47 (0.79 2.74) NS 27 1.06 (0.43 2.58) NS Schooling Illiterate 195 1.00 153 1.00 63 1.00 Basic level 725 0.55 (0.28 1.05) 0.071 650 0.76 (0.21 1.18) NS 330 0.94 (0.35 2.49) NS Secondary level 560 0.27 (0.13 0.57) 0.001 485 0.59 (0.30 1.16) NS 275 0.51 (0.19 1.34) NS University 113 0.24 (0.09 0.67) 0.007 234 0.64 (0.30 1.38) NS 121 0.49 (0.19 1.25) NS Paid work past week Yes 667 1.00 828 1.00 444 1.00 No 930 0.73 (0.51 1.03) 0.076 688 0.63 (0.42 0.94) 0.024 346 0.29 (0.13 0.60) 0.001 Hospitalization Never 1,105 1.00 990 1.00 553 1.00 Past 12 months 141 0.95 (0.45 1.99) NS 113 1.95 (1.03 3.67) 0.039 46 1.02 (0.26 3.91) NS Ever 370 1.07 (0.69 1.69) NS 406 1.00 (0.67 1.50) NS 189 1.83 (0.95 3.55) 0.070 Blood transfusion Never 1,513 1.00 1,382 1.00 726 1.00 Past 12 months 16 0.97 (0.21 4.54) NS 36 1.05 (0.48 2.29) NS 8 Ever 68 0.90 (0.42 1.94) NS 68 0.82 (0.39 1.76) NS 45 3.24 (1.48 7.09) 0.001 Surgery Never 898 1.00 761 1.00 416 1.00 Past 12 months 168 0.73 (0.35 1.50) NS 147 1.18 (0.56 2.50) NS 63 1.37 (0.41 4.59) NS Ever 547 0.83 (0.54 1.27) NS 615 0.77 (0.49 1.21) NS 311 0.72 (0.38 1.37) NS Dental treatment Never 214 1.00 111 1.00 60 1.00 Past 12 months 586 0.84 (0.43 1.65) NS 637 0.48 (0.23 1.00) 0.05 321 0.45 (0.16 1.30) NS Ever 817 0.99 (0.52 1.89) NS 775 0.52 (0.24 1.10) NS 409 0.52 (0.18 1.51) NS Health-care related job No 1,548 1.00 1,431 1.00 741 1.00 Yes 69 0.79 (0.37 1.69) NS 93 1.59 (0.90 2.79) NS 49 2.05 (1.03 4.07) 0.038 Tatoo No 1,529 1.00 2,727 1.00 728 1.00 Yes 88 2.24 (1.17 4.30) 0.015 85 1.36 (0.54 3.45) NS 62 1.63 (0.69 3.88) NS Body piercing No 1,550 1.00 1,457 1.00 728 Yes 69 1.02 (0.31 3.30) NS 66 62 Sharing of cutting objects No 472 1.00 709 1.00 237 1.00 Yes 1,145 0.88 (0.61 1.29) NS 812 1.00 (0.66 1.49) NS 553 0.51 (0.28 0.93) 0.028 * OR = odds ratio; CI = confidence interval; NS = not significant. Weighted ORs adjusted for random effect and age. individuals reporting paid employment in the past week had higher levels of HBV infection compared with those who reported that they had no paid work. Individuals who reported a history of blood transfusion showed a three-fold increase in prevalence of infection compared with those with no such history (OR = 3.24, 95% CI = 1.48 7.09). Healthcare related jobs (OR = 2.05, 95% CI = 1.03 4.05) was a predictor of HBV infection and ever being hospitalized showed an association with a borderline P value (OR = 1.83, 95% CI = 0.95 3.55, P = 0.07). Approximately 80% of the study population (13 69 years of age) in the three regions reported having initiated sexual activity; this was a risk factor for infection in the Northeast and Central-West regions. In the Northeast region, factors related to HBV infection were having a bisexual partner or not knowing (OR = 4.49, 95% CI = 2.06 9.81) and a history of a sexually transmitted disease (OR = 1.74, 95% CI = 1.16 2.60). In the Central-West region, a sexual partner with a past or current history of hepatitis was a predictor of infection. A life history of more than one sexual partner was associated with HBV infection in the Central-West region and Federal District (Table 3 ). Smoked drugs had frequencies varying from 6.3% (Central- West region) and 11.1% (Federal District). Heavy alcohol consumption ranged from 5.8% (Central-West region) to 10.8% (Northeast region). Reporting of sniffed drugs varied from 5.2% in the Northeast region to 3.2% in the Central- West region. Self-acknowledge of inhaled drugs ranged from 1.7% in the Northeast region to 0.8% in the Federal District. Only a few individuals reported current or past use of injected drugs in all regions. Factors associated with infections were use of smoked and sniffed drugs, and heavy alcohol consumption in the Northeast region. In the Central-West region, use of smoked drugs was marginally significant in relation to hepatitis B infection. In the Federal District, having ever used inhaled drugs and having ever drunk alcohol was related to seropositivity ( Table 4 ).

244 PEREIRA AND OTHERS Table 3 Relative odds of hepatitis B infection for sexual behavioral factors among non-vaccinated persons 13 69 years of age in three regions in Brazil, 2005* No. OR (95% CI) P No. OR (95% CI) P No. OR (95% CI) P Initiated sexual life No 382 1.00 284 1.00 124 1.00 Yes 1,233 2.72 (1.28 5.76) 0.009 1,238 2.89 (1.05 7.97) 0.04 662 2.83 (0.58 13.6) NS Bisexual partner No 1,145 1.00 1,160 1.00 626 1.00 Yes 42 4.49 (2.06 9.81) < 0.0001 14 1.52 (0.39 5.90) NS 11 1.10 (0.17 7.16) NS Does not know 39 4.38 (1.83 10.47) 0.001 66 0.34 (0.12 0.96) 0.034 27 0.68 (0.18 2.49) NS Another sexual partner besides current partner No 489 1.00 554 1.00 225 1.00 Yes 675 1.05 (0.66 1.67) NS 636 1.79 (1.20 2.68) 0.004 408 3.27 (1.45 7.36) 0.004 Previous STD No 1,053 1.00 1,059 1.00 569 1.00 Yes 166 1.74 (1.16 2.60) 0.007 165 1.35 (0.87 2.09) NS 91 2.03 (0.89 4.64) 0.09 Current partner had hepatitis No 1,031 1.00 1,097 1.00 561 1.00 Yes 36 0.84 (0.29 2.39) NS 41 2.24 (1.15 4.36) 0.018 21 1.53 (0.31 7.43) NS Previous sexual relationship with known HBV carrier No 1,149 1.00 1,164 1.00 620 1.00 Yes 34 0.41 (0.10 1.59) NS 25 2.64 (1.08 6.46) 0.033 23 1.31 (0.28 6.08) NS * OR = odds ratio; CI = confidence interval; NS = not significant; STD = sexually transmitted disease; HBV = hepatitis B virus. Weighted ORs adjusted for random effect and age. Results of the final multivariate model showed that age and socioeconomic conditions remained independent risk factors in all regions (Table 5). Male sex and having initiated sexual activity were strongly associated with HBV infection in the Northeast and Central-West regions. In Federal District, having a healthcare-related job and previous hospitalization were independently associated with infection. Having had a tattoo was kept in the final model (Northeast region) despite having a P value of 0.14 because it was likely to be caused by the lack of statistical power. In addition, collinearity between tattoo and illegal drug use was found in the intermediate steps of the multivariate analysis. Therefore, the variable tattooing may also reflect the use of illegal drugs and the association between tattooing and HBV infection should be interpreted with caution. DISCUSSION This hepatitis B survey conducted among a representative sample of the population 13 69 years of age in Brazil showed a prevalence of HBsAg of less than 1% in the Northeast and Central-West regions and in the Federal District of Brasília, representing an overall estimated 30,000 individuals chronically infected with HBV. Our findings classify these areas as having low endemicity for HBV, rather than intermediate endemicity defined by previous studies. 3 In our study, an Table 4 Relative odds of hepatitis B infection for drug use-related factors among non-vaccinated persons 13 69 years of age in three regions in Brazil, 2005* No. OR (95% CI) P No. OR (95% CI) P No. OR (95% CI) P Ever used smoked drugs No 1,449 1.00 1,417 1.00 701 1.00 Yes 164 1.88 (1.09 3.22) 0.022 95 0.40 (0.15 1.06) 0.066 88 1.34 (0.53 3.36) NS Ever used inhaled drugs No 1,588 1.00 1,496 1.00 782 1.00 Yes 27 2.07 (0.55 7.71) NS 22 0.67 (0.08 5.36) NS 7 6.32 (1.21 32.9) 0.028 Ever used sniffed drugs No 1,533 1.00 1,467 1.00 752 1.00 Yes 85 2.21 (1.04 4.71) 0.039 48 0.76 (0.24 2.49) NS 37 2.01 (0.51 7.88) NS Ever used injected drugs No 1,604 1.00 1,505 1.00 786 1.00 Yes 4 8 2 Previous use of glass syringe No 1,260 1.00 1,174 1.00 592 1.00 Yes 335 1.12 (0.77 1.64) NS 322 1.07 (0.68 1.67) NS 190 0.88 (0.39 2.00) NS Alcohol consumption None 985 1.00 964 1.00 358 1.00 Light 457 1.16 (0.81 1.67) NS 471 1.30 (0.86 1.98) NS 381 1.14 (0.60 2.16) NS Heavy 169 1.84 (1.08 3.13) 0.025 82 0.73 (0.29 1.96) NS 48 1.75 (0.91 3.34) 0.09 Dependent 5 8.12 (0.36 184.65) NS 7 3.67 (0.80 16.91) 0.095 3 7.78 (0.57 106.0) NS * OR = odds ratio; CI = confidence interval; NS = not significant. Weighted ORs adjusted for random effect and age.

HEPATITIS B INFECTION AND RISK ASSESSMENT 245 OR (95% CI) P OR (95% CI) P OR (95% CI) P Age 1.03 (1.02 1.05) 0.0001 1.04 (1.03 1.06) 0.0001 1.05 (1.03 1.08) 0.0001 Sex Female 1.0 1.0 1.0 Male 1.71 (1.20 2.44) 0.003 2.09 (1.50 2.92) 0.0001 1.89 (0.83 4.62) 0.127 Schooling Illiterate 1.0 Basic level 0.59 (0.32 1.08) 0.086 Secondary level and university 0.28 (0.14 0.55) 0.0001 Paid work past week Yes 1.0 No 0.36 (0.15 0.86) 0.021 Dental treatment Never 1.0 Past 12 months 0.47 (0.22 1.01) 0.054 Ever 0.46 (0.22 0.99) 0.047 Hospitalization Never 1.0 1.0 Past 12 months 2.14 (1.14 4.01) 0.018 1.30 (0.29 5.85) 0.729 Ever 1.01 (0.68 1.51) 0.954 2.01 (1.03 3.91) 0.040 Tatoo Yes 1.62 (0.82 3.20) 0.166 Healthcare-related job Yes 2.40 (1.04 5.54) 0.040 Sharing of cutting tools Yes 0.58 (0.30 1.12) 0.104 Initiated sexual life 1.0 Yes 2.90 (1.28 6.56) 0.010 2.92 (1.06 8.03) 0.038 Ever used smoked drugs Yes 0.31 (0.12 0.82) 0.018 Ever used inhaled drugs (cheira cola) Yes 4.24 (0.78 22.9) 0.094 * OR = odds ratio; CI = confidence interval; NS = not significant. Adjusted ORs corrected for random effect and weighted for age group. Table 5 Factors associated with hepatitis B infection in three regions of Brazil, 2005* upward trend for HBV infection (antibodies against HBc) with age showed a similar force of infection, when evaluated using the simple catalytic model, which indicates comparable transmission rates in all regions. The HBV infection data refer to the non-vaccinated group because it represents the population at risk. In the three settings, the overall prevalence of HBsAg for the unvaccinated population was less than 1%. In accordance with these findings, surveys of blood donors in Brazil also showed an HBsAg prevalence of approximately 1% in the Central-West and Northeast regions 14 and less than 1% in selected state capitals across the country. 13 In this population-based study, HBV genotypes A and F were present in the Northeast region and genotypes A, D, and F were present on the Central-West region. These results are consistent with predominant genotypes circulating in all regions in Brazil, as identified in blood donors 28 or among chronically infected patients. 29,30 It is worth noting that the F genotype identified may indicate population movement from the Amazon region where this genotype predominates. 7 In our study, a higher prevalence of infection occurred in males than in females at the population level, as in the United States National Health and Nutrition Survey. 21 Other studies have also presented higher frequencies of HBV infection among males recruited from restricted areas and in volunteers or clinic outpatients from different regions in Brazil. 8,31,32 In the three regions in Brazil, age was an independent predictor of HBV infection, which indicates length of time for viral exposures and initiation of sexual activity, which is considered the main risk factor for HBV acquisition in most countries. 2,33 Initiation of sexual activity was strongly associated with HBV infection in the northeast and Central-West regions, in accordance with most studies, 33 which indicates the importance of reinforcing safe sexual behavior. Having a partner with homosexual experience and tattooing were also significantly associated with HBV infection in the Northeast region. However, issues relating to sexual practices should be interpreted with caution, because these data were collected in one interview in the household. Healthcare-related employment was positively associated with the HBV in one area (Federal District), suggesting missing opportunities for HBV vaccination and safe medical practices. Ever having used inhaled drugs was associated with infection and this factor could be considered either a risk factor or a surrogate marker for other practices. Because drug use is an illegal practice, this information may be underestimated in a population-based study. Interestingly, a lower level of education was a risk factor for HBV infection in the Northeast region, which is the most deprived area in the study.

246 PEREIRA AND OTHERS Education was considered a reliable marker of socioeconomic conditions when hepatitis A infection was analyzed in the same regions. 20 Although most of the predictor factors found in our study are well-described risk factors for HBV, 33 some unexpected associations were found, such as paid employment and dental treatment as a protective factor. These risk factors could be interpreted as markers for socioeconomic conditions that the multivariable analysis was not able to control. Our study aimed to evaluate HBV infection and chronically infected individuals in capital cities. These results are valid for the most populous cities in the study regions, where the largest human reservoir of HBV is expected to be found. However, a limitation exists in extrapolating these findings to smaller cities or rural communities. In Brazil, most studies reporting areas in which HBV is highly endemic were conducted in limited areas. 7,8,18 Another potential limitation was that because our study was a household survey, underreporting of sensitive questions (drug use, sexual behavior) was likely. The association of HBV with these variables should be interpreted in that light. This multicentric study in the Northeast and Central-West regions is the first national effort to obtain knowledge of hepatitis B infection and its magnitude and diversity of risk factors in large urban areas. Among the study participants, the non-vaccinated group was older than the vaccinated group and represented approximately 70% of the total population. This finding appears to be compatible with national HBV vaccine policy that has currently achieved a high degree of coverage. Vaccination for HBV was later extended to adolescents in the public sector. 11,31 Our data indicated that approximately 4% of HBV infections among the vaccinated group could be explained by previous natural infection among adolescents and high risk groups because serologic screening is not a public health recommendation before HBV vaccination. Estimating the prevalence of immunization is outside the scope of this study because the test for detecting antibody against HBs was not performed for the whole study population. However, it was possible to identify the non-vaccinated population with a reasonable degree of certainty because there were highly positive and negative predictive values for selfreported information regarding vaccination compared with the vaccination card. Our finding of approximately 30% of HBV-vaccinated individuals was similar to the results of HBV vaccine coverage in the U.S. National Health Interview Survey, based on self-reporting of HBV vaccination status among the adult population. 3 Our findings that age and initiation of sexual activity are risk factors for HBV reinforce the need for extensive HBV vaccine coverage among adolescents to prevent viral infection. All individuals in the risk groups identified by our survey should be considered candidates for HBV vaccine and educational measures. The impact of vaccination can be assessed by estimating incidence of infection from prevalence data using catalytic modeling as reported by two population surveys during different periods. 1,21 Such evaluation showed no reduction in HBV infection during the time in which high-risk adults were targeted for hepatitis B vaccination. 34 Therefore, vaccination upon request for individuals without specific risk should be considered, in accordance with the current recommendations of the Centers for Disease Control and Prevention (Atlanta, GA). 3 The range of risk factors found in our study over the three areas underlines the complexity involved in drawing up a national policy to target specific risk groups. This survey shows the importance of generating baseline population-based information on HBV infection and conducting periodic cross-sectional surveys to assess the impact of HBV vaccination and to facilitate educational strategies that take into account regional differences. Received October 26, 2008. Accepted for publication March 25, 2009. Acknowledgment: We thank University of Pernambuco Foundation for administrative support. Hepatitis Study Group: Zulma Medeiros (Centro de Pesquisa Aggeu Magalhães FIOCRUZ), Demócrito de Barros Miranda Filho (Universidade de Pernambuco), Maria Mabel Melo (Laboratório Central de Saúde Pública - Dr. Milton Bezerra Sobral), Conceição Sá (Hemocentro Centro de Hematologia e Hemoterapia do Piauí), Carlos Henrique Nery Costa (Universidade Federal do Piauí), Arnaldo de Jesus Dominici (Universidade Federal do Maranhão), Maria das Graças Aragão (Universidade Federal do Maranhão), Elizabeth de Souza Lima (Laboratório Central de Saúde Pública do Maranhão), José Milton de Castro Lima (Universidade Federal do Ceará), José Wellington Oliveira Lima (Fundação Nacional de Saúde-Ceará), Maria do Carmo Vidal Gadelha Lima (Laboratório Central de Saúde Pública Ceará), Gilmar Amorim de Sousa (Universidade Federal do Rio Grande do Norte), Márcia Araújo Barreto (Universidade Federal do Rio Grande do Norte), Manoel Jaime Xavier Filho (Universidade Federal da Paraíba), Jória Viana Guerreiro (Universidade Federal da Paraíba), Dayse Mércia Cavalcanti de Oliveira (Fundação Nacional de Saúde-AL), Maria Rosileide Bezerra Alves (Secretaria Estadual de Saúde de Alagoas), Ivoneide Moreira de Oliveira Barros (Laboratório Central de Saúde Pública), Tereza Virgínia Silva B. Nascimento (Universidade Federal de Sergipe), Lúcia Maria Sayde de Azevedo Tavares (Fundação Nacional de Saúde), Raymundo Paraná Ferreira Filho (Universidade Federal da Bahia), José Tavares- Neto (Universidade Federal da Bahia), Maria Alice Sant Anna Zarife (Laboratório Central de Saúde Pública da Bahia ), Rodrigo Sebba Aires (Secretaria Estadual de Saúde de Goiás), Renato Maurício de Oliveira (Instituto de Patologia Tropical e Saúde Pública/Universidade Federal de Goiás), Beatriz Maranhão Bariani (Laboratório Central de Saúde Pública Goiás), José Ivan Aguiar (Universidade Federal do Mato Grosso do Sul), Eugênio Barros (Secretaria Municipal de Saúde Campo Grande), Gilza Bastos dos Santos (Laboratório Central de Saúde Pública Mato Grosso do Sul ), Francisco José Dutra Souto (Universidade Federal do Mato Grosso), Cor Jesus Fernandes Fontes (Universidade Federal do Mato Grosso), Virgínia Correia de Azevedo (Laboratório Central de Saúde Pública), Roberto de Melo Dusi (Secretaria Estadual de Saúde Distrito Federal ), Lídia Maria Pinto Luna (Laboratório Central de Saúde do Distrito Federal), Gabriela Coral (Fundação Faculdade de Ciências Médicas de Porto Alegre), Airton Tetelbom Stein (Fundação Faculdade de Ciências Médicas de Porto Alegre), José Carlos da Fonseca (Universidade Federal do Amazonas), Leila Melo Brasil (Fundação Medicina Tropical do Amazonas), Kátia Biscuola de Campos (Programa Nacional de Prevenção e Controle das Hepatites Virais/Secretaria de Vigilância em Saúde/Ministério da Saúde) Financial support: This study was supported by the Brazilian Ministry of Health and the Pan American Health Organization. The authors were partially supported by CNPq (scholarship 307963/2004-7 to C.M.T.M, 303049/2007-3 to M.R.A.C, and 300917/2006-6 to R.A.A.X). Authors addresses: Leila M. M. B. Pereira, Ulisses R. Montarroyos, Luiz C. A. de Alencar, and Ricardo A. A. Ximenes, Faculdade de Ciências Médicas de Pernambuco, Hospital Universitário Oswaldo Cruz, Universidade de Pernambuco, Rua Arnóbio Marques, 310, Santo Amaro, CEP 50100130, Recife, Pernambuco, Brazil. Celina M. T. Martelli and Marília D. Turchi, Instituto de Patologia Tropical e Saúde Publica, Departamento de Saúde Coletiva, Universidade Federal de Goiás, Rua Delenda Rezende de Mello, S/N, Sala 405, Setor Universitário, CEP 74605050, Goiania, GO, Brazil. Edgar Merchán-Hamann, Faculdade Ciências da Saúde, Departamento de Saúde Coletiva, Faculdade de Saúde, Universidade de Brasília, Campus Universitário Darcy Ribeiro, Asa Norte, CEP 70910-900, Brasília, Distrito Federal, Brazil. Maria C. Braga and Maria L. C. de Lima, Fundação Oswaldo Cruz, Centro de Pesquisas Aggeu Magalhães, Departamento de Parasitologia, Av Moraes Rego, s/n,

HEPATITIS B INFECTION AND RISK ASSESSMENT 247 Cidade Universitária, CEP 50000-230, Recife, PE, Brazil. Maria R. A. Cardoso, Faculdade de Saúde Publica, Departamento de Epidemiologia, Universidade de São Paulo, Avenida Dr. Arnaldo, 715, Cerqueira César, CEP 01246-904, São Paulo, SP, Brazil. Marcelo A. Costa, Hospital de Base do DF, Área Especial, Asa Sul, CEP70.335-900, Brasília, DF, Brazil. Regina C. Moreira, Instituto Adolfo Lutz, Av. Dr. Arnaldo, no. 355, Cerqueira Cezar, CEP 01246-902, São Paulo, SP, Brazil. Gerusa M. Figueiredo, Ministério da Saúde, Programa Nacional de Prevenção e Controle das Hepatites Virais. SCS Quadra 04, Bloco A Edifício Principal 4 andar, CEP 70304-000, Brasília, DF, Brazil. REFERENCES 1. McQuillan GM, Kruszon-Moran D, Kottiri BJ, Curtin LR, Lucas JW, Kington RS, 2004. Racial and ethnic differences in the seroprevalence of 6 infectious diseases in the United States: data from NHANES III, 1988 1994. Am J Public Health 94: 1952 1958. 2. Shepard CW, Simard EP, Finelli L, Fiore AE, Bell BP, 2006. Hepatitis B virus infection: epidemiology and vaccination. Epidemiol Rev 28: 112 125. 3. Mast EE, Weinbaum CM, Fiore AE, Alter MJ, Bell BP, Finelli L, Rodewald LE, Douglas JM Jr, Janssen RS, Ward JW, 2006. A comprehensive immunization strategy to eliminate transmission of hepatitis B virus infection in the United States: recommendations of the Advisory Committee on Immunization Practices (ACIP) Part II: immunization of adults. MMWR Recomm Rep 55: 1 33. 4. World Health Organization, 2000. Hepatitis B. Available at: http:// www.who.int/mediacentre/factsheets/fs204/en/print.html. Accessed July 1, 2008. 5. Lewis E, Shinefield HR, Woodruff BA, Black SB, Destefano F, Chen RT, Ensor R, 2001. Safety of neonatal hepatitis B vaccine administration. Pediatr Infect Dis J 20: 1049 1054. 6. Souto FJ, Fontes CJ, Gaspar AM, 2001. Prevalence of hepatitis B and C virus markers among malaria-exposed gold miners in Brazilian Amazon. Mem Inst Oswaldo Cruz 96: 751 755. 7. Viana S, Parana R, Moreira RC, Compri AP, Macedo V, 2005. High prevalence of hepatitis B virus and hepatitis D virus in the western Brazilian Amazon. Am J Trop Med Hyg 73: 808 814. 8. Souto FJ, Fontes CJ, Oliveira JM, Gaspar AM, Lyra LG, 1997. Epidemiological survey of infection with hepatitis B virus in the savannah and wetlands (Pantanal) of central Brazil. Ann Trop Med Parasitol 91: 411 416. 9. Clemens SA, da Fonseca JC, Azevedo T, Cavalcanti A, Silveira TR, Castilho MC, Clemens R, 2000. Hepatitis A and hepatitis B seroprevalence in 4 centers in Brazil. Rev Soc Bras Med Trop 33: 1 10. 10. Echevarria JM, Leon P, 2003. Epidemiology of viruses causing chronic hepatitis among populations from the Amazon Basin and related ecosystems. Cad Saude Publica 19: 1583 1591. 11. Secretaria de Saúde de São Paulo, 2006. Hepatitis B vaccine. Rev Saude Publica 40: 1137 1140. 12. Pereira LM, Melo MC, Lacerda C, Spinelli V, Domingues AL, Massarolo P, Mies S, Saleh MG, McFarlane IG, Williams R, 1994. Hepatitis B virus infection in schistosomiasis mansoni. J Med Virol 42: 203 206. 13. Nascimento MC, Mayaud P, Sabino EC, Torres KL, Franceschi S, 2008. Prevalence of hepatitis B and C serological markers among first-time blood donors in Brazil: a multi-center serosurvey. J Med Virol 80: 53 57. 14. Martelli CM, Turchi M, Souto FJ, Saez-Alquezar A, Andrade AL, Zicker F, 1999. Anti-HBc testing for blood donations in areas with intermediate hepatitis B endemicity. Rev Panam Salud Publica 6: 69 73. 15. Marchesini AM, Pra-Baldi ZP, Mesquita F, Bueno R, Buchalla CM, 2007. Hepatitis B and C among injecting drug users living with HIV in Sao Paulo, Brazil. Rev Saude Publica 41 (Suppl 2): 57 63. 16. Bertolini DA, Pinho JR, Saraceni CP, Moreira RC, Granato CF, Carrilho FJ, 2006. Prevalence of serological markers of hepatitis B virus in pregnant women from Parana State, Brazil. Braz J Med Biol Res 39: 1083 1090. 17. Souto FJ, Santo GA, Philippi JC, Pietro BR, Azevedo RB, Gaspar AM, 2001. Prevalence of and factors associated with hepatitis B virus markers in a rural population of central Brazil. Rev Panam Salud Publica 10: 388 394. 18. Foccacia R, da Conceição O, Sette H, Sabino E, Bassit L, Nitrini D, Lomar A, Lorenço R, Vieira de Souza F, Kiffer C, Santos E, Gonzales M, Sáez-Alquézar A, Riscal J, Fischer D, 1998. Estimated prevalence of viral hepatitis in the general population of the municipality of São Paulo, measured by a serologic survey of a stratified, randomized and residence-based population. Braz J Infect Dis 2: 269 284. 19. Kiesslich D, Rocha JE Jr, Crispim MA, 2002. Prevalence of hepatitis E virus antibodies among different groups in the Amazonian basin. Trans R Soc Trop Med Hyg 96: 215. 20. Ximenes RA, Martelli CM, Merchán-Hamann E, Montarroyos UR, Braga MC, de Lima ML, Cardoso MR, Turchi MD, Costa MA, de Alencar LC, Moreira RC, Figueiredo GM, Pereira LM for the hepatitis study group, 2008. Multilevel analysis of hepatitis A infection in children and adolescents: a household survey in the Northeast and Central-west regions of Brazil. Int J Epidemiol 37: 852 861. 21. McQuillan GM, Coleman PJ, Kruszon-Moran D, Moyer LA, Lambert SB, Margolis HS, 1999. Prevalence of hepatitis B virus infection in the United States: the National Health and Nutrition Examination Surveys, 1976 through 1994. Am J Public Health 89: 14 18. 22. Kaneko S, Miller RH, Feinstone SM, Unoura M, Kobayashi K, Hattori N, Purcell RH, 1989. Detection of serum hepatitis B virus DNA in patients with chronic hepatitis using the polymerase chain reaction assay. Proc Natl Acad Sci USA 86: 312 316. 23. Sitnik R, Pinho JR, Bertolini DA, Bernardini AP, da Silva LC, Carrilho FJ, 2004. Hepatitis B virus genotypes and precore and core mutants in Brazilian patients. J Clin Microbiol 42: 2455 2460. 24. Sanger F, Nicklen S, Coulson AR, 1977. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74: 5463 5467. 25. Anderson RM, May RM, 1992. Dynamic aspects of eradication and control. Anderson RM, May RM, eds. Infectious Disease of Human: Dynamics and Control. First edition. Oxford, United Kingdom: Oxford University Press, 172 207. 26. Fonseca JC, 1999. Epidemiologia da infecção pelo vírus da hepatite C no Brasil. Gastroenterologia Endoscopia Digestiva 18: S3 S8. 27. Lyra LG, Damasceno AP, Cotrim P, Mota E, Silva L, 1986. Prevalence of antibody to hepatitis B virus in an urban population of northeast Brazil. Rev Inst Med Trop Sao Paulo 28: 406 409. 28. Mello FC, Souto FJ, Nabuco LC, Villela-Nogueira CA, Coelho HS, Franz HC, Saraiva JC, Virgolino HA, Motta-Castro AR, Melo MM, Martins RM, Gomes SA, 2007. Hepatitis B virus genotypes circulating in Brazil: molecular characterization of genotype F isolates. BMC Microbiol 7: 103. 29. Ribeiro NR, Campos GS, Angelo AL, Braga EL, Santana N, Gomes MM, Pinho JR, De Carvalho WA, Lyra LG, Lyra AC, 2006. Distribution of hepatitis B virus genotypes among patients with chronic infection. Liver Int 26: 636 642. 30. Devesa M, Pujol FH, 2007. Hepatitis B virus genetic diversity in Latin America. Virus Res 127: 177 184. 31. Silveira TR, da Fonseca JC, Rivera L, Fay OH, Tapia R, Santos JI, Urdeneta E, Clemens SA, 1999. Hepatitis B seroprevalence in Latin America. Rev Panam Salud Publica 6: 378 383. 32. El Khouri M, Duarte LS, Ribeiro RB, da Silva LF, Camargo LM, dos Santos VA, Burattini MN, Corbett CE, 2005. Seroprevalence of hepatitis B virus and hepatitis C virus in Monte Negro in the Brazilian western Amazon region. Clinics 60: 29 36. 33. Goldstein ST, Alter MJ, Williams IT, Moyer LA, Judson FN, Mottram K, Fleenor M, Ryder PL, Margolis HS, 2002. Incidence and risk factors for acute hepatitis B in the United States, 1982 1998: implications for vaccination programs. J Infect Dis 185: 713 719. 34. Coleman PJ, McQuillan GM, Moyer LA, Lambert SB, Margolis HS, 1998. Incidence of hepatitis B virus infection in the United States, 1976 1994: estimates from the National Health and Nutrition Examination Surveys. J Infect Dis 178: 954 959.