Prevalence and risk factors for HBV and HCV among incarcerated people who inject drugs in Iran: a cross sectional study

Background: This study aimed at determining the prevalence of and risk factors for hepatitis B virus (HBV) and hepatitis C virus (HCV) among incarcerated people who inject drugs (PWID) in Iran during 2015-16. Methods : The required data was collected from a database provided by Iranian national bio-behavioral surveillance surveys (BBSSs) of 11988 prisoners who were selected using multi-stage sampling method from among 55 prisons in 19 provinces during 2015-16. The data on demographics and behavioral status of subjects were collected through interviews and the status of HBV and HCV exposure were determined using ELISA blood test. A total of 1387 individuals with a lifetime history of drug injection were enrolled into the study. Data were analyzed using the survey package in Stata/SE 14.0. Univariate and multivariate logistic regression were used to investigate the relationships between risk factors and outcomes. Results : The mean age of the incarcerated PWID was 36.83 ± 8.13 years. Of all, 98.46% were male and 50.97% were married. The prevalence of HCV and HBV were 40.52% and 2.46%, respectively. HCV prevalence was associated with age ≥ 30 years, being single, illiteracy and low level of education, prison term>5 years, history of piercing, and extramarital sex in a lifetime (P<0.05). Conclusions : The prevalence of HCV is alarmingly high. In general, it is recommended to adopt measures for HCV screening and treatment and HBV vaccination for incarcerated PWID without a history of vaccination.


Background
Viral hepatitis is currently one of the most important causes of mortality and morbidity worldwide (1).
Contrary to human immunode ciency virus that has a declining trend, the burden of hepatitis is increasing in the world, and 96% of the deaths from hepatitis are related to two main types of viral hepatitis, namely hepatitis B virus (HBV) and hepatitis C virus (HCV) (2). These two diseases are considered as health challenge, while both of them are preventable, and more importantly, HCV is even treatable.
The prevalence of HBV differs from that of HCV that is attributed to differences in their patterns of geographic distribution. HBV is more common in Africa and the Western Paci c Region while HCV is more common in the eastern Mediterranean and European region. However, according to recent reports published by the World Health Organization (WHO), the prevalence of HBV and HCV in the world in 2015 were 3.5% and 1%, respectively (2).
Studies have shown that a large proportion of cases with HBV and HCV are exclusively observed in speci c groups such as prisoners and people who inject drugs (PWID) (3), that might be attributed to different reasons. First, a signi cant percentage of prisoners are incarcerated due to drug-related crimes (4), therefore, there is a high population of PWID in prisons. Secondly, due to the lack of interventions and proper care and treatment services, shared injection is a common practice in prisons (5) and it has been proved that non-sterile and shared injections are the main routes of the transmission of some diseases (6,7). The prevalence of HBV and HCV in prisons in the world in 2014 were 4.8% and 15.1%, respectively (8); moreover, their prevalence among PWID in 2017 were 9.1% and 52.3%, respectively (9). Iran is no exception and has a similar status. The prevalence of HBV and HCV among the general population were 0.9% in 2012 (10) and 0.6% in 2015 (11), respectively. The mentioned gures, respectively, were 2.5% and 9.5% in prisons in 2015 (12), while they were reported to be 30.9% and 51.5%, respectively, among PWID in 2012 (10). Therefore, the prevalence of HBV and HCV among incarcerated PWID is expected to be signi cantly higher than the gures mentioned above.
Daneshmand and Davoodian's study on incarcerated PWID in Iran showed that the duration of imprisonment is one of the factors associated with the prevalence of HBV and HCV (13,14). A population-based study on Iranian prisoners conducted by Moradi et al. showed that the prevalence of HBV and HCV was associated with a history of drug use; in addition, the prevalence of HCV was associated with a history of tattooing and an age of >30 years old (12). Silverman et al. conducted a study in Mexico and showed that the prevalence of HCV in prisons had a positive relationship with injecting drug use, older ages, higher frequency of incarcerations, and a history of tattooing, while it had a negative relationship with higher levels of education and being married (15).
The high prevalence of blood-borne diseases (BBD) in prisons can be both a threat and an opportunity for the health system. It is a threat because the prison is not an environment isolated from the community and the increase in the prevalence of the diseases in such environments is associated with an increase in the burden of diseases in the community, especially when the prisoners are released. It is also an opportunity, because prisons can help to reach infected people and high-risk groups at a place that, in fact, facilitates the access to this group of people. Moreover, it enables health policy-makers to design and implement appropriate behavioral surveys to accurately estimate the burden of diseases, identify the risk factors associated with each disease, tailor appropriate interventions, and effectively implement diagnostic, preventive, and therapeutic programs especially for cases with HCV and for a large number of eligible people (16). This, in turn, will not only facilitate, but also speed up the achievement of the World Health Organization's (WHO) goal set for eliminating HBV and HCV by 2030 (2). The aim of this study was to determine the prevalence of and risk factors for HBV and HCV among Iranian incarcerated PWID using national bio-behavioral surveillance surveys (BBSSs) during 2015-2016.

Data source
The present cross-sectional study is part of a series of reiterating biobehavioral surveillance surveys (BBSSs) of HBV and HCV among Iranian prisoners that were carried out in 2015 and 2016. The methodology of BBSSs was the same in both periods and is presented in comprehensively published BBSSs (12,17), however, a summary is provided below. The statistical population included all Iranian prisoners who were in prison at the time of the study. The prisoners were sampled via multi-stage random sampling method. First, all the 31 provinces of Iran were categorized into three groups, namely the north, center, and south groups. Then, three provinces from each category, and 2-4 prisons from each province were selected randomly (9 provinces and 26 prisons in 2015 and 10 provinces and 29 prisons in 2016 were included in the study). Next, the required number of samples from each prison was determined using probability-proportional-to-size sampling method based on the average annual population of incarcerated people in each province and prison. Finally using systematic random sampling, eligible prisoners (Iranian, age >18 years, and with a history of prison term of at least one month at the time of the study) were recruited from the roster of all inmates in the prison and after obtaining a written informed consent were enrolled into the study. Accordingly, over the two rounds, 12,800 Iranian prisoners including 6200 in 2015 and 6600 in 2016 that accounted for 0.06 and 0.10 of the total prisoners in the selected cities were recruited. Of the 12800 selected subjects, 11988 (5507 in 2015 and 6481 in 2016) participated in the surveys. Of these participants (11988), 1387 prisoners had a history of injecting drug use in the lifetime and formed our study population.

Data collection tool
Demographic and behavioral data In both rounds, a questionnaire-based interview was used for collecting data on demographics and behavioral status of the participants. The questionnaires, included seven sections (demographic data, criminal records, history of sexually transmitted diseases (STD), knowledge about HBV and HCV, and three other sections about high-risk behaviors including drug abuse, tattooing and piercing). All demographic and behavioral data was collected via self-report. The variables were de ned as below: Illiteracy: Inability to read and write.
Primary school, Junior high school, or Diploma: Having the last degree of education at every educational level.
University education: A certi cate of general and specialized knowledge and skills acquired from a university or other formal educational centers.
History of imprisonment: At least a history of imprisonment other than the present term of incarceration.
Number of incarcerations: Frequency of records of imprisonment other than the present term of incarceration; it was calculated for those with a previous history of imprisonment.
Prison term in the lifetime: Sum of the years of imprisonment; it was calculated for those with a history of imprisonment.
History of tattooing in the lifetime: History of injecting ink into the skin in the lifetime in every place.
History of tattooing in prison: History of injecting ink into the skin in prison; it was calculated for those with a history of tattooing in the lifetime.
History of piercing in the lifetime: History of making a whole in ear, lip, mouth, navel, and breast by a needle and cupping in the lifetime in every place.
History of piercing in prison: History of making a whole in ear, lip, mouth, navel, and breast by a needle and cupping at prison; it was calculated for those with a history of piercing in the lifetime.
History of extramarital sex in the lifetime: Engagement of a married person in a sexual activity with someone other than his/her legal spouse.
Number of sexual partners in the lifetime: Number of people, other than a legal spouse, that a married person had sexual activity with, whether heterosexual or homosexual. It was calculated for those with a history of extramarital sex in the lifetime History of STD: Symptomatic self-reported history of wounds or secretion from the genital area that are abnormal in terms of amount, color, and odor and STDs diagnosed by a physician.
Serologic data Blood samples were collected for serological testing to examine HBV and HCV exposure. Dried blood spot samples were collected from individuals who consented to participate in the study and stored under a standard condition. HBsAg and HCV antibodies were determined through assessing optical density (od) and cut-off points in ELISA (Enzyme-Linked Immunosorbent Assay) using Dia Pro kits (Diagnostic Bioprobes Srl, Italy). The validity of the results was veri ed using repeat testing for both positive and negative cases. All the tests on blood samples were carried out in the Pasteur Institute of Iran.
The questioning team in each prison consisted of two trained individuals: a) an interviewer who was an employee of health and treatment department of the prison, b) a blood sampler who was a staff of a university of medical sciences.

Statistical analysis
Data were analyzed using the survey package in Stata/SE 14.0. The weighting, clustering, and strati cation of the survey design were applied for the main data with a total of 12800 samples. Weights of the samples were determined using poststrati cation method, and postweight instrument. The frequency distribution of poststrata was determined through assessing the data on people's willingness to participate in the study. Finally the data required for this study was analyzed using descriptive statistics (frequency and mean) and analytical statistics (chi-square test, univariate and multivariate logistic regression). Since there were a large number of variables, only the variables with a P < 0.2 in the chi-square test were entered into the multivariate model.
The results of chi-square test showed that the prevalence of HCV was signi cantly higher in individuals with 30 years of age and older (as compared with people under 30 years old), illiterate prisoners and those with a level education lower than university education (as compared with those with an university education), prisoners with a higher frequency and length of imprisonment (as compared with those with a lower frequency and length of imprisonment), subjects with a history of tattooing (as compared with those without a history of tattooing in the lifetime), and those with a history of piercing in the lifetime (as compared with those with those without a history of piercing). However, there was no signi cant difference between the mentioned subgroups of prisoners and the prevalence of HBV (Table 1).

Discussion
This study aimed at identifying risk factors for HBV and HCV among one of the most important high-risk groups, i.e. incarcerated PWID in Iran. Comparing the prevalence of HCV in our study with the prevalence observed in other studies in 2015, it was found that the ratio of the prevalence of HCV among incarcerated PWID to general population was 67.53 (40.52% vs 0.6%) (11), to prisoners was 4.3 (40.52% vs 9.48%) (12), and to PWID was 0.79 (40.52% vs 51.46%) (10). In addition, comparing the prevalence of HBV in our study with the prevalence reported by other studies, it was found that the ratio of the prevalence of HBV among incarcerated PWID to the above mentioned subgroups was 2.73 (2.46% vs 0.90%) (10), 0.99 (2.46% vs 2.48%) (12), and 0.08 (2.46% vs 30.90%) (10) in the same year. Based on the results of these comparisons, the prevalence of HCV and HBV among non-incarcerated PWID are higher than that among the three groups of incarcerated PWID, prisoners, and general population. The higher prevalence of this disease among the incarcerated PWID than among prisoners and general population, can be justi ed, as explained earlier in the introduction. However, the lower prevalence of the disease among this group than among non-incarcerated PWID may be attributed to the fact that although injections are practiced in prisons, it is a prohibited action; therefore, the frequency of injection, and in particular the frequency of the practice of shared and non-sterile injection are lower in prisons than in the community. Furthermore, prisoners are prone to hide their real injection status. However, the ndings of a study by Falla et al. in Europe showed that the prevalence of HBV and HCV in prisoners is higher than that among injecting drug users (3). It highlights the fact that imprisonment results in the gathering of a large group of patients and provides a chance to implement health-related interventions, including diagnosis, treatment, and prevention programs.
The simultaneous investigation of the prevalence of HBV and HCV in this study and other studies in Iran show that the prevalence of HCV is signi cantly higher than the prevalence of HBV among prisoners and drug users (10,12), while at the community level, HBV is slightly more prevalent than HCV (10). Hence, when resources and facilities are limited, it is recommended to give priority to HCV than HBV. In addition, in order to achieve the goal of eliminating hepatitis in the community, it is necessary to direct a large part of investment and health interventions toward HCV.
Based on the results of this study, the prevalence of HCV among incarcerated PWID was 40.52%; although it is a noteworthy gure, it is signi cantly lower than the gures reported for this group of people in other studies. The reported rate at the global level is 64% (18), and in the Snow's study in Australia it was reported to be 47.4% (19). Moreover, according to Behzadifar et al., its rate in Iran is 53%. The observed difference may be attributed to differences between the types of two studies. Behzadifar et al.'s study was conducted as a meta-analysis that investigated papers published from 2004 to 2016, and the highest prevalence rate was observed during the years prior to 2010 (20).
As the most important ndings of this study, it was found that the prevalence of HCV among incarcerated PWID was associated with an age of 30 years and older, being single, illiteracy and low level of education, more than ve years of imprisonment, history of piercing in the lifetime, and history of extramarital sex. The association between HCV prevalence and age and lengthy imprisonment may be due to the fact that people aged 30 years and older and those with longer imprisonment terms are more likely than people younger than 30 years and those with shorter imprisonment terms to have long term contact with highrisk groups; so they are more prone to be exposed to the risk factors. The relationship between HCV and being single can be attributed to the fact that married people have a kind of commitment and adherence to ethics that is less observed in single persons (21). The observed relationship between illiteracy and low level of education and HCV may also be attributed to the lack of information and knowledge or the lack of social prestige among illiterate people and those with a low level of education; it is worth noting that university education has a deterrent effect on socially negative behaviors. Piercing was also identi ed as a risk factor, which might be attributed to shared needle injection and incorrect sterilization of piercing tools (22). In addition, high-risk sexual activity was also among the most well-known routes of STD transmission. Some of the risk factors identi ed in this study are more or less sporadically reported in a limited number of studies conducted on incarcerated PWID. Silverman-Retana et al.'s study in Mexico showed that the prevalence of HCV was positively associated with older ages (age over 40) and the frequency of imprisonment, while it had a negative relationship with high level of education and being married (15). In a study by Dolan et al. in Australia, it was found that the history of imprisonment and high level of education were associated with the prevalence of HCV among incarcerated PWID (23). Davoodian et al.'s study in Iran showed that the prevalence of HCV was associated with the length of incarceration (14).
Based on the results of chi-square test, there was no P-value less than 0.2 when assessing the relationships between HBV and independent variables. Therefore, multivariate logistic regression analysis was not performed for assessing the prevalence of HBV (Table 2) and there was no risk factor for the prevalence of HBV among the incarcerated PWID. This may be attributed to the low prevalence of HBV infection among the studied people. This nding is consistent with the results of Davoodian et al.'s study (14); however, the results of Daneshmand's study in Iran indicated an association between HBV and the length of imprisonment (13). This controversy may be attributed to differences between the settings of the two studies. Daneshmand's study was conducted among incarcerated PWID in just one province (Isfahan) out of a total of 31 provinces of Iran, and the prevalence of HBV in the mentioned province was signi cantly higher (27.2%), while the prevalence reported in this study (2.46%) was achieved through conducting surveys in 19 provinces.
This study was part of the BBSSs on HBV and HCV in prisons in Iran, and the required data was collected through self-reports. Therefore, information bias, including recall bias and social desirability bias were the most important limitations of the study; in order to reduce their effects, complementary questions were used in each section to identify unreal answers to the main questions.

Conclusion
Prior to this study, there was no report on the prevalence of and risk factors for HBV and HCV among Iranian incarcerated PWID at the national level. Since PWID and imprisonment are two important risk factors for HCV, it is important to pay special attention to this group of prisoners to signi cantly reduce the prevalence of HBV and HCV. The results of this study showed that the prevalence of HBV and HCV among incarcerated PWID is alarmingly high. In general, it is recommended to perform HCV screening and treatment and HBV vaccination for incarcerated PWID without a history of vaccination. In case of the lack or shortage of resources and facilities in health care system, in order to reduce HCV prevalence more effectively, it is suggested to implement screening and treatment programs through prioritizing incarcerated PWID with short terms who will be released back into the community more rapidly and those who practice high-risk behaviors such as tattooing, piercing, and extramarital sex.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. STROBEchecklistcrosssectional.pdf