Seroprevalence of Hepatitis B and C and Associated Risk Factors among HIV-1 Infected Patients in a High Risk Border Region of South West Cameroon

Background: Coinfection of HIV with human retroviruses such as hepatitis B (HBV) and hepatitis C (HCV) has been associated with adverse liver disease outcomes including reduced survival, cancer and antiretroviral induced hepatotoxicity. In this study, we evaluated the prevalence of hepatitis B surface antigen (HBsAg) and hepatitis C antibody (HCVAb) among HIV-infected individuals diagnosed within six months in South West Cameroon and identified risk factors of coinfection. Methods: We performed a systematic screening for HBsAg and HCVAb among 299 newly enrolled HIV-positive patients South West of Cameroon. Enzyme Linked Immuno-sorbent Assay (ELISA) confirmed the presence of HBV/HCV infection status. Risk factors of coinfection were assessed by logistic regression using SPSS V20 and Epi-Info v7. A P value < 0.05 was considered significant. Results: The overall prevalence of Hepatitis B/C was 16.05% with 12.0% (95% CI: 1.96-2.03) for HBV and 4% (95% CI: 1.97-2.02) for HCV. Participants in urban environment were less likely to be infected with HBV [aOR: 0.171 (0.044-0.660), P = 0.010] compared to rural environment. A history of blood transfusion increased odds of hepatitis B positivity six times [aOR: 5.8 (2.04-16.94), P = 0.001] and 13 times for hepatitis C [13.14 (3.04-56.81, P = 0.001]). Conclusions: The seroprevalence of hepatitis B and C among diagnosed HIV patients observed was 12% and 4% respectively. A history of blood transfusion was independently ORigiNAl ARTiCle


Background
Viral hepatitis has emerged as an important public health problem globally. It is characterized by high prevalence, high burden of morbidity and mortality, and suboptimal diagnosis and management approaches [1,2]. Hepatitis B (HBV) and C (HCV) are two among the numerous forms of infections whose clinical degeneration, morbidity-mortality and low immune responsiveness in people living with human immunodeficiency virus (HIV) are highly evident. Co-infections between HIV, HBV and/or HCV have been associated with reduced survival, increased risk of progression to liver diseases and increased risk of hepatotoxicity associated to antiretroviral therapy [3,4].
While varying in their transmission efficiencies, HIV, HBV and HCV share common routes of transmission, and as such, the prevalence of HBV and HCV are gener-ally higher in HIV-infected individuals [5,6]. Studies have shown that HBV, HCV and HIV are endemic in Africa and especially the sub-Saharan Africa [7]. However, co-infection rates among HIV-infected individuals remain controversial (international journal of infectious diseases, 2010). Scientists have reported that the worldwide prevalence of HBV ranges from 0.1-20% [8,9].
Hepatitis B virus is detected in blood and body fluids (semen, saliva, and nasopharyngeal fluids), and the four major modes of transmission are: Sexual contact, mother to child transmission in pregnancy and at birth, blood to blood contact and through sharing of infected items [10]. The world's predominant mode of transmission of HBV is perinatal. In this case, children born of HBV positive mothers have 90% chance of having the infection too, among whom 25% may die in adult life of chronic liver diseases or cancer [11]. Sub-Saharan Africa (SSA) has the highest burden of infectious diseases and remains the epicenter of the global HIV epidemic [12,13]. Despite this, data for HBV and HCV are lacking due to limited diagnostic capacity and lack of prioritization in public health programs.
In one meta-analysis, HIV/HBV frequency in SSA varied from 0 to 28.4% (median 7.8%) [2]. One study carried out in South Africa among HIV patients receiving anti-retroviral therapy (ART) showed a prevalence of 5% for HBV [14]. Also, a prospective study carried out among HIV patients admitted to a large government hospital in Johannesburg found a co-infection rate of 6% [15]. Further, a retrospective control laboratory-based study showed a HBV rate of 16.2% among the HIV patients from a more rural area of the country [16].
HCV is also detected in blood and body fluids, and transmitted from an infected person to a non-infected person in like manner as for HBV. HIV/HCV co-infection is being recognized as a separate entity from HIV or HCV mono-infections [17]. The prevalence of HCV among the HIV patients taking antiretroviral therapy according to a cohort study carried out in Switzerland is 37.2% [18]. Two systematic reviews suggest an overall HCV prevalence of 3% among people living with HIV in SSA with significant regional variations (0-55.9%) [3,19].
In Cameroon, a hospital prevalence of 8.9% (HBV = 6.1% and HCV = 2.8%; [20]) among HIV patients in two regional facilities was observed. With increasing access to ARV therapy and increasing HBV/HCV infections, the burden of the latter among the HIV patients in resource limited countries are expected to increase in all ages contrary to the case in Europe and North America [21,22]. Understanding the prevalence and disease characteristics of HBV/HCV amongst the HIV patients is thus very essential [23]. Furthermore liver co-morbidities may compromise response to antiretroviral drugs and worsen outcome of HIV patients even when these drugs are efficacious. Although the guideline for the clinical management of HIV patients recommends screening for viral hepatitis among newly diagnosed HIV patients; it is not systematically implemented in Cameroon, as it is not included in the package of baseline laboratory tests [20]. Consequently, the data on hospital prevalence of new HIV infections is lacking, especially in conflict regions and humanitarian settings such as in the Kumba Health District. The aim of this study was to determine the seroprevalence of HBsAg and HCVAb in newly diagnosed HIV-positive patients in Kumba Health District, South West region of Cameroon to contribute to better management of HIV patients in the region.

Study design and setting
This was a prospective cross-sectional study carried out between November 2018 and January 2019. Newly diagnosed HIV patients or HIV patients diagnosed within six months were consecutively recruited for the determination of the HBV and HCV sero-status. This study targeted selected HIV-positive patients enrolled at the three health facilities that care for HIV patients in Kumba. These were the district hospital Kumba, Kumba integrated health centre and the Apostolic hospital in Banga-Bakundu. Screening was offered free-of-charge to all selected consenting participants. The Kumba district hospital is the referral facility for the health district.

Study population sampling and recruitment
HIV patients were consecutively sampled in each of the three recruitment centres in the Kumba district. Data were collected from the 299 study participants after having completed an informed consent. Those who were newly diagnosed as well as those who reportedly knew their status within six months were also included in the study population. Data on their socio-demographic characteristics as well as clinical data including their WHO HIV defined clinical staging and co-morbidity were collected. Additional data was obtained from participants through a semi-structured questionnaire interview and medical chart review. Selected participants were invited for screening during the routine HIV appointment, during which samples were collected and analyzed in the laboratories at the study sites.

Sample collection and serological assays
For each consenting participant, 3 mls of whole venous blood was collected into heparinised tubes and centrifuged to obtain plasma. Two tests were performed on each of the samples collected to determine circulating HBV surface antigen and HCV antibodies. The samples were first screened using a rapid diagnostic test (RDT) for both HBsAg and HCV antibody using the Diaspot® (Nantong, 226010, P.R. China) HBsAg and Diaspot® (Nantong, 226010, P.R. China) for HCVAb respectively. The procedure was based on manufacturer's instructions. After 15 minutes the RDT test results were read and recorded as either positive (red lines on test the threshold of significance.
After conducting a bivariate analysis using chisquare, the following variables were entered in to a multivariate binary logistic regression model to control for the effects of other explanatory variables: Area of residence, history of blood transfusion, educational level, take ART regularly, level of knowledge categories on HCV, level of knowledge categories on HBV.

Ethical considerations
The ethical clearance for this study was obtained on the 15 th November 2018, from the Faculty of Health Science Institutional Review Board, University of Buea following the protocol number 814-05.

Sociodemographic characteristics of the study participants
A total of 299 HIV positive patients were screened for HBV and HCV in the Kumba Heath District. A majority of the study participants were from the District hospital Kumba 257 (86.0%), were female (208/299, 69.6%), aged 35-44 (99/299, 33.1%), and with a primary level of education (126/299, 42.1%). In addition, a majority of the study participants were Single/Cohabiting/Separated (160/299, 53.5%), Christians (282/299, 94.3%), (100/299, 33.4%), from rural areas 202 (202/299, 67.6%) and 89.3% (267/299) reported no history of blood transfusion (Table 1). window and control window) or negative (red line on control window and no line on test window) or invalid (no line on control window irrespective of presentation of test window). Invalid tests were repeated on a new RDT. All HBV and/or HCV RDT positives samples underwent a further ELISA immunoassay (Golden Bio Technology ELISA) to obtain the definitive test outcome.

Sample size
Our sample size calculation used the Normal approximation to the Binomial distribution of our sample population. The sample size is based on the following calculation Where, X = (Z α/2 ) 2* p * (1-p)/e 2 , p = 8.9 based on studies by Luma, et al. [20], e = margin of error estimated at 3.5. By inputting these values, we obtained a sample estimate of 254 participants. We accounted for a non-response rate of 15% and a test sensitivity and specificity of 99% and 100% as reported by the test manufacturers for both HBV and HCV. We therefore estimated that we needed a minimum of 292 participants for the study.

Statistical methods and data analysis
Our primary study outcome was the proportion of newly HIV positive individuals with hepatitis B surface antigen and hepatitis C antibody, as confirmed by ELI-SA immunoassay. Newly diagnosed HIV patients were defined as HIV infected patients diagnosed during the study or within six months prior to the study starting. The different variables included demographic information like age, sex, marital status, level of education, religion, occupation, setting, civil status and factual information like: Weight, WHO clinical and biological staging, HIV-1 therapy, smoking status, alcohol status, blood transfusion history, hepatitis status. These data were entered into Microsoft Excel 2010. Descriptive statistics were used to summarize the distribution of these data in the study e calculated as the proportion of confirmed HBV and HCV infections in the total number of participants tested. This was presented with the corresponding 95% confidence intervals. We estimated the odds of having hepatitis infection given risk factors fitted in a binary logistic regression model. We first conducted a bivariate analysis and calculated R values of all selected predictors. We found that for all analysis, the R values were less than 0.7. Next we, generated co-linearity diagnostic to check for variance inflation by multi-co-linearity of predictors with tolerance value for all predictors set at P > 0.1. All variance inflation factors were less than 5, indicating together that predictors were not highly inter-correlated. We reported the adjusted odds ratios and their corresponding P values where appropriate. All analyses were done using SPSS version 20 and Epi info version 7.0, and a P value of = 0.05 was used as

Prevalence of HBV/HCV among newly infected HIV patients
Our primary outcome was the prevalence of HBV and HCV infections among newly diagnosed HIV patients in the Kumba health district. Out of the 299 HIV positive patients screened for HBV and HCV within the study period, a total of 48 newly infected HIV patients were infected with either the HBV or HCV virus, giving a seroprevalence of 16.0%. Of this 36/299 (12.0%) were positive for HBV and 12 (4%, 95% CI: 1.97-2.02) were infected with the HCV by the ELISA immunoassay. In addition the prevalence of hepatitis B virus was higher in participants with no formal education (18/88 (20.5%), X 2 = 11.139, P = 0.011) ( Table 2). The prevalence of hepatitis B virus was significantly higher in participants who live in rural areas, 33/202 (16.3%) compared to those who live in urban areas, 3/97 (3.1%) (X 2 = 10.853 P = 0.001) ( Table 2); among males 13/91 (13.4%) compared with females 23/208 (11.05%).
The prevalence of hepatitis C virus was significantly higher in participants from Kumba urban hospital ( Table  2).
As pertaining to blood transfusion, the prevalence  someone were 5.88 times more likely to be infected with hepatitis B (95% CI: 2.04-16.94, P-values: 0.001), and 13.14 times more likely to be infected with hepatitis C (95% CI: 3.04-56.81, P-values: 0.001), than participants who have never received blood from someone (Table  3). Participant's area of residence also had a significant effect on hepatitis B virus prevalence. Participants from rural areas were 5.84 (1/0.171) times more likely to be infected with hepatitis B than those from urban areas (95% CI: 0.044-0.660, P-values: 0.010) ( Table 3). Also, the prevalence of HBV and HCV were higher in males than female i.e for HBV, females 23/208 (11.05%) (X 2 = 0.251, P = 0.616) and males 13/91 (13.4%) (X 2 = 0.251, P = 0.616) and for HCV, females 8/208 (3.8%) (X 2 = 0.050, P = 0.824) and males 4/91 (4.4%) (X 2 = 0.050, P = 0.824), which were not statistically significant. Furthermore, participant's treatment centres had a significant effect on hepatitis C virus prevalence. The odds are low for participants from Kumba District hospital (0.24 times that of those from Kumba Subdivisional hospital). Thus, participants from Kumba Subdivisional hospital 4.14 (1/0.24) times more likely to be infected with hepatitis C virus than those from Kumba District hospital (95% CI: 0.06-0.84, P-values: 0.027) ( Table 3). of both HBV (13/32 (40.6%), P < 0.0001) and HCV (6/32 (18.8%), X 2 = 20.202, P < 0.0001) virus was significantly higher in participants who reported that they have received blood from someone. Furthermore, participants who reported that they do take their anti-retroviral therapy (ART) regularly significantly had a low prevalence of hepatitis B, 21/231 (9.1%) compared to those who do not take their ART regularly 15/68 (22.1%) (X 2 = 8.342, P = 0.004) ( Table 2). As regarding the level of participant's knowledge on the modes of transmission and the cardinal signs/symptoms of hepatitis viruses, participants who recorded poor level of knowledge on hepatitis B virus significantly had a high prevalence of hepatitis B virus, (26/80 (32.5%) compared to those who recorded good level of knowledge on hepatitis B virus, 10/219 (4.6%) (X 2 = 43.172, P < 0.0001). In addition, participants who recorded poor level of knowledge on hepatitis C virus significantly had a high prevalence of hepatitis C virus, 9/79 (11.4%), compared to those who recorded good level of knowledge on hepatitis C virus, 3/220 (1.4%) (X 2 = 15.176, P < 0.0001) ( Table 2).

Risk factors of Hepatitis B and Hepatitis C viruses
Participants that have ever received blood from Also, a majority of these participants were cohabiting/single/separated (53.5%). On the perspective of the occupation, a variety of occupations were presented among the participants. The predominance of farmers was remarkable because most of the participants were from the rural areas highly involved in farming.

Knowledge on hepatitis B and C virus
Only 9.4% of the participants knew the signs and symptoms of hepatitis B and C leaving 90.6% who did not know any signs and symptoms of these infections. This proportion was not good because it means that the participants had a very poor knowledge of these health threats, which means that the infections may continue to spread among the people in the community and may lead to poor management. Only 9.4% of the par-

Discussion
Guidelines for clinical management of HIV patients recommends screening for viral hepatitis, unfortunately this is not standard practice in Cameroon, as it is not included in the package of baseline laboratory tests [20]. Given the limited resources available for population screening efforts, the present study is aimed to establish the prevalence of hepatitis B and C virus, and the associated risk factors among the HIV patients placed on highly active anti-retro viral therapy (HAART), which could contribute towards the understanding of the burden of the viral hepatitis-co-infections in Kumba Health District and to guide prioritisation for testing of key atrisk and geographic populations.

Demographic characteristics
A total of 299 HIV patients taking ART in three different treatment centres in Kumba Health District were interviewed. The differences in the recruitment rate at the various selected treatment centres respected the patient's population sizes. The highest number of participants was recruited from the Kumba District hospital because of the high HIV patient's population density of the health facility. A total of 208 (69.6%) females and  As regarding to the level of participant's knowledge on the modes of transmission and the cardinal signs/ symptoms of hepatitis viruses, participants who recorded poor knowledge on HBV significantly had a high prevalence of HBV 26/80 (32.5%) compared to those who recorded good level of knowledge about the infection, 10/219 (4.6%) (X 2 = 43.172, P < 0.0001), which is raised a high need to carryout effective education/sensitization in the different communities.

Prevalence of hepatitis C virus after laboratory analysis
The prevalence of HCV was 4% which was much lower than that obtained among the HIV patients taking antiretroviral therapy according to a study carried out in Switzerland with 37.2% [18], but was slightly lower than that obtained at the Buea and Limbe regional hospitals in HIV patients of age 21 years and above 2.8% [20]. On the other hand, this prevalence was slightly similar to that obtained in two systematic reviews among people living with HIV in SSA 3% [3,19].
The prevalence of HCV was significantly higher in participants from Kumba Sub-divisional hospital (4/11 (8%), X 2 = 6.178, p = 046). In the same light with HBV, as pertaining to blood transfusion, hepatitis C virus (6/32 (18.8%), X 2 = 20.202, P < 0.0001) was also significantly higher in participants who reported to have been transfused blood. In addition, participants who recorded poor level of knowledge on hepatitis C virus significantly had a high prevalence of hepatitis C virus, 9/79 (11.4%), compared to those who recorded good level of knowledge on hepatitis C virus, 3/220 (1.4%) (X 2 = 15.176, P < 0.0001).
The result showed that participants that have ever received blood from someone were 5.88 times more likely at risk to be infected with hepatitis B (95% CI: 2.04-16.94, P-values: 0.001), and 13.14 times more likely at risk to be infected with hepatitis C (95% CI: 3.04-56.81,P-values:0.001), than participants who have never received blood from someone. There is thus a high need for hospitals to improve on the quality of screening for blood products with strict respect of screening protocols. The participant's area of residence also had a significant effect on HBV prevalence. The odd ratios were low for participants from urban areas (0.171 times that of those from rural areas). Therefore, participants from rural areas are 5.84 (1/0.171) times more at risk to be infected HBV than those from urban areas (95% CI: 0.044-0.660, P-values: 0.010). This different could be due to the low level of education in the rural areas, limited to tele-communication networks which has attributed to very low knowledge of the participants on the infection. Further-ticipants knew that hepatitis B and C are transmitted through contact with infected blood meanwhile 90.6% did not know the cause of these infections. For the majority of the participants not knowing the cause of these infection could be attributed to the high level of very low level of education and illiteracy discovered among this people since 42.1% has attended only primary level of education while 29.4% had never attended school. In addition, the poor knowledge among the participants could also be linked to inadequate counseling during follow up for HIV therapy as well as implementation of social groups among HIV patients which limit regular contact between the care giver and the clients.

Prevalence of hepatitis B virus
The prevalence of HBV was 12% among HIV patients, which was higher than that obtained in a similar studied carried out in South African in HIV patients receiving ARV therapy 5% [14], Buea and Limbe regional hospitals in HIV patients of age 21 years and above 6.1% [20] and National hospital of Tropical Disease in Vietnam 8.4% [24].
On the other hand, this prevalence was slightly lower than that documented from a retrospective control laboratory-based study carried out in South Africa among the HIV patients from a more rural area of the country at 16.2% [16]. In similar manner, this prevalence was slightly similar to that obtained in Australia 13.1% [25]. 3/97 (3.1%) (X 2 = 10.853 P = 0.001) and this is consistent with a study conducted in Rwanda [7]. As pertaining to blood transfusion, the prevalence of hepatitis B virus (13/32 (40.6%), X 2 = 27.648, P < 0.0001) was significantly higher in participants who reported to have been transfused blood. Furthermore, participants who reported that they do take their ART regularly significantly had a low prevalence of hepatitis B, 21/231 (9.1%) compared to those who do not take their ART regularly 15/68 (22.1%) (X 2 = 8.342, P = 0.004).
The prevalence of hepatitis B virus was higher in participants with no formal education (18/88 (20.5%), X 2 = 11.139, P = 0.011). This prevalence was also significantly higher in participants who live in rural areas, 33/202 (16.3%) compared to those who live in urban areas, where the population get more educative talks directly from all the radio stations and other health partners carrying out multiple health activities in the community 3/97 (3.1%) (X 2 = 10.853 P = 0.001) and this is consistent with a study conducted in Rwanda [7]. As pertaining to blood transfusion, the prevalence of hepatitis B virus (13/32 (40.6%), X 2 = 27.648, P < 0.0001) was significantly higher in participants who reported to have been transfused blood. Furthermore, participants who reported that they do take their ART regularly significantly had a low prevalence of hepatitis B, 21/231 (9.1%) compared to those who do not take their ART regularly 15/68 (22.1%) (X 2 = 8.342, P = 0.004). Lamivudine use in the