Prevalence , risk factors and hepatitis B immunization : helping fill the gap on hepatitis B epidemiology among homeless people , Goiânia , Central Brazil

Prevalencia, factores de riesgo e inmunización contra la hepatitis B: ayudando a completar las lagunas en la epidemiología de la hepatitis B entre personas sin techo en Goiânia, Goiás, Brasil Paulie Marcelly Ribeiro dos Santos Carvalho 1 Marcos André de Matos 1 Regina Maria Bringel Martins 2 Raquel Silva Pinheiro 1 Karlla Antonieta Amorim Caetano 1 Márcia Maria de Souza 1 Megmar Aparecida dos Santos Carneiro 2 Sheila Araujo Teles 1


Introduction
Street population experiences prejudice and discrimination, and has a strong intersection with drug users and sex workers, representing a vulnerable population that mostly has poor access to health services 1 .Globally, the number of people living on the streets is estimated at 100 million.These people have death rates four times higher than the general population and a high prevalence of psychiatric disorders, abuse of licit and illicit drugs, and infectious diseases 2 .In Brazil, about 50,000 people live on the streets in state capitals, the Federal District and cities with over 300,000 inhabitants 3 .
It is estimated that 2 billion people have been infected with hepatitis B virus (HBV) worldwide, and that every year 600,000 people die due to the consequences of viral infection such as chronic hepatitis, cirrhosis, and hepatocellular carcinoma 4 .In Brazil, where the overall prevalence of HBV is low, the burden of the infection is concentrated mainly in populations with social vulnerability and risk behaviors 5,6 .
Homeless people have been disproportionately affected by infectious diseases including HBV infection.In Brazil, the only study on hepatitis B in this population was carried out in the beginning of 2000 among 330 shelter users in the city of São Paulo, Southeastern Region of the country.The author showed 30.6% were previously exposed to HBV and 3.3% were chronic HBsAg carriers 7 , suggesting a high prevalence of hepatitis B among this population.
Hepatitis B vaccine has been available for almost 20 years in our country 8 , and therefore it should be of great interest to decision-makers to know whether this policy has reached this target-population.Therefore the purpose of this study was to investigate the prevalence, risk factors and HBV immunization status among homeless people being served in a public shelter in the city of Goiânia, Central Brazil.

Methodology
This is a cross-sectional study carried out among homeless people served at a public shelter in the city of Goiânia (1,302,001 inhabitants).Data collection was conducted from August 2014 to June 2015.This shelter was opened in 2007, and has the capacity to serve approximately 240 individuals.This facility represented the only public shelter in the city during the period of investigation.
The Brazilian National Survey on Homelessness carried out in 2008 estimated the number of homeless people in Goiânia at 563 9 .The study sample was calculated on the basis of an alpha and beta error of 5% and 20%, respectively, with an expected HBV prevalence of 30.6% 7 and a precision of 1.5%.According to these data, the minimum sample size necessary was 311 individuals.
Those included in the study were individuals served by the public shelter, aged over 18 years.Subjects were excluded if they were under the effects of medication/psychoactive drugs at the time of the interview; and/or displayed behavior that would preclude the collection of a blood sample.
Data collection was conducted three days a week during the morning, afternoon, or evening.All individuals were interviewed in a private place, using a structured questionnaire containing questions about sociodemographic characteristics and predictors of HBV infection.10mL of blood collected by venipuncture was subjected to detection of HBV markers: HBsAg, total anti-HBc and anti-HBs (Bioelisa, Biokit S.A., Barcelona, Spain).Individuals whose sera were positive only for anti-HBs were considered vaccinated against HBV.
Data were analyzed using Stata SE, version 13 (StataCorp LP, College Station, USA).Prevalence was calculated with 95% confidence intervals (95%CI).Descriptive analysis was conducted by frequency distribution, arithmetic mean and standard deviation.For analysis of potential predictors of exposure to HBV (HBsAg and/or total anti-HBc positive), individuals who had serological profile of previous vaccination against HBV were excluded.This study used the chi-square test (χ 2 ) or Fisher's exact test to assess differences between proportions.For calculation of the adjusted prevalence ratio (PR), variables that presented a p-value < 0.10 were included in a Poisson regression model with robust error variance.P-values < 0.05 were considered statistically significant.
This project was approved by the Ethics Committee of the Federal University of Goiás (protocol n. 045/13).Since individuals living on the streets have great geographical mobility and poor access for Cad.Saúde Pública 2017; 33(7):e00109216 returning surveys, rapid testing was carried out for the detection of HBsAg markers (VIKIA -HBsAg, BioMérieux Brasil S/A, Rio de Janeiro, Brazil), following the recommendations of the Brazilian Ministry of Health.Individuals HBsAg positive were referred to the public health service, per prior agreements with the Municipal Health Secretariat of Goiânia, for evaluation and treatment (if necessary).
It was found that the prevalence of HBV exposure (anti-HBc) increased from 10.8% in subjects aged 18-30 years to 46.8% in individuals older than 50 years.The prevalence of immunized (anti-HBs isolated) participants dropped from 53.6% in subjects aged 18-30 years to 1.4% in individuals older than 50 years (p < 0.001).There is no difference between the proportions of vaccinated vs. unvaccinated homeless regarding sex, ethnicity, and years of education (p > 0.05).Statistically significant differences were observed in the proportion of individuals exposed to HBV, considering the following variables: age, sleeping on the street, and sexual intercourse with people living with HIV/AIDS (p < 0.05).Also, the variables ethnicity (self-identified), sexual orientation, and history of sexual violence tended to be associated (p < 0.10) (Table 2).
These variables, as well as sex, were included in a Poisson regression model with robust error variance, and age > 50 years (adjusted PR: 3.06), black (adjusted PR: 1.81), and homosexuality or bisexuality (adjusted PR: 2.60 and 1.80, respectively) were independently associated with exposure to HBV (Table 3).

Discussion
Corroborating findings observed by other investigators in countries of low to intermediate endemicity 10,11,12,13 , a high prevalence of HBV markers (21.8%; 95%CI: 17.8-26.4)was observed among homeless people in Goiânia.This was higher than that estimated for the adult population of capitals in the same region of the country (13.2%; 95%CI: 10.9%-15.5%) 14, and also for those found in other   money or drugs = 2; sex with partner living with HIV/AIDS = 61; sex with drug users = 11; anal sex = 2; previously incarcerated = 2; previous STI = 13; previous blood transfusion = 7.

Table 3
Multivariate analysis of variables associated with hepatitis B markers.

Variable Adjusted PR (95%CI) p-value
Age (years) vulnerable populations from the same region who have a large intersection with homeless such as recyclable waste collectors (n = 431; 12.8%; 95%CI: 9.8-16.2) 15 and users of illicit drugs (n = 852; 14%; 95%CI: 11.7-16.5) 16.However, the prevalence found among homeless people in Goiânia was lower than that found ten years before by Brito et al. 7 (30.6%;95%CI: 25.9-35.8) in São Paulo.In Goiânia, the participants were younger and had better education levels.These characteristics very likely contributed to these findings.The gold standard for assessment of vaccine coverage is the immunization record (a card issued by the public health system).However, in adult populations there is little chance of access to this information, especially if the population has little access to public health services such as those people living on the streets 1 .Furthermore, the verbal report of previous hepatitis B vaccination is poorly reliable 17,18 .Therefore, despite methodological limitations, detection of anti-HBs in serum alone is used as indicator prior immunization.
In our study, only 19.5% of participants presented a serological profile of prior immunization, confirming the low frequency of hepatitis B vaccination among vulnerable populations in our region and elsewhere 16,19,20,21 .However, it should be noted the wide variation observed in the distribution by age group, with prevalence of 53.6% in individuals aged 18 to 30 years dropping to 1.4% among those aged above 50 years.These better results among younger subjects very likely reflects the outcome of the Brazilian National Immunization Program policy, which from the 2000s effectively implemented hepatitis B vaccination in children and adolescents nationwide.This policy was also gradually extended to older people and vulnerable populations, and is currently available to older people 22 .These findings and that of higher HBV prevalence among older homeless reinforce the need of health managers to invest in more persuasive strategies to vaccinate these specific populations, particularly older subjects.
We found that self-reported homosexual or bisexual orientation was statistically associated with HBV exposure (p = 0.009 and p = 0.030), corroborating previous studies that have shown a higher risk of hepatitis B among men who have sex with men (MSM) 23,24 .Many MSM have multiple sexual partners and use condoms inconsistently during sexual intercourse, behaviors that put them at increased risk for sexually transmitted infections, including hepatitis B 23,25 .
In our country, social racial categorization depends on appearance of the individual, being classified as white, black, mixed race, indigenous and yellow 26 .According to the 2010 demographic census 2010, the Brazilian population is 47.7% white, 43.1% mixed race, and 7.6% black 27 .This composition is dramatically different from that found among the homeless studied here, with 78.6% self-declared mixed race or black.In addition, in our country and elsewhere 28,29,30 , the black population is subject to greater economic and social vulnerability, which can contribute to disproportionate susceptibility to infectious diseases.Our findings reinforce these assumptions, since ethnicity was predictive of HBV infection (adjusted PR: 1.70; p = 0.024).
Some limitations of the study should be presented.The study was conducted in a public institution that houses people on the streets and may not be representative of the entire homeless population in the city.However, the characteristics of the individuals studied were similar to those in the Brazilian National Survey on Homelessness 3 , giving the findings external validity.The prevalence data for vaccination coverage may be underestimated, since the date of the last dose of vaccine is unknown, and according to the kinetics of this marker, the titers of anti-HBs decline over the years.However, our results are consistent with other studies in vulnerable populations.Finally, many questions are morally sensitive, and despite the team's efforts to assure the respondent of privacy and anonymity, some answers may not have been truthful.
Although the study has been limited to individuals housed in a shelter, it is believed that the obtained data will contribute to the body of knowledge on the homeless population, and encourage more specific health policies for this population, which in most cases is outside the public health services and presents social and individual vulnerabilities that put them at risk of infections such as hepatitis B.

Table 1
Prevalence of hepatitis B virus (HBV) markers among 353 homeless individuals in Goiânia, Central Brazil.

Table 2
Bivariate analysis of risk factors associated with exposure to hepatitis B virus (HBV) among 353 homeless individuals in Goiânia, Central Brazil.