Study design, site and participants
This was a cross-sectional survey conducted among children born after 2006 and their respective mothers in Nanoro health and demographic surveillance system area (HDSS), Centre-west region of Burkina Faso. The study site covers an updated population size of 63,000 inhabitants. The study area was described elsewhere [17]. Nanoro is a rural area located in the Centre-West of the country. In this area, health care is provided by seven peripheral health posts and one referral hospital. Based on an anticipated hepatitis B surface antigen (HBsAg) prevalence in the rural area of 17.3% in Burkina Faso [3], confidence level of 95%, precision of 5, and 10% maximum missing data, the minimum sample size for mothers is 240 participants. As for children required sample size is less than 240, because the prevalence among children is thought as lower than that among adults. Therefore, the required sample size was totally 240 pairs of children and their mothers. The study on HBsAg prevalence in rural area of Burkina Faso is limited. The study area was randomly selected among all rural area of Burkina Faso. This study population covers 0.46% of all rural population in Burkina Faso. All 24 villages of the Nanoro HDSS were selected. We conducted sampling procedure as follows: First mothers were selected by random sampling with probability proportional to village size. Second, children were selected using simple random sampling from children of selected mothers when mother have more than one eligible child. Then, a list of 240 pairs of mothers above 15 years old and their child was generated from the HDSS database. Selected potential participants were visited at home by study investigators. Informed consent was obtained before study related activities were performed. If the woman has more than an eligible child aged one to 11 years old, we selected the participant (child) among eligible children using an automatic random number generator (RNG).
Questionnaire survey
Trained study investigators administered a standardized questionnaire to all study participants (Additional file 1: Appendix 1a and 1b). Socio-demographic data such as age, sex, ethnic group, educational level, occupation, marital status, vaccination history; and potential risk factors including surgery, blood transfusion, injections, tattoos, skin-piercings, scarification, birthplace, were collected. The vaccination history was taken mainly from vaccination card. If the child has no vaccination card, the information was obtained by the recall-memory of their mothers. For the mother, the information could only be obtained by their recall memory.
Serological and molecular assays
A rapid point of care Alere Determine™ HBsAg test strips, sensitivity 95–100% and specificity 96–100% (Abbott Japan Co., Ltd.) using 1 drop (20 μl) was performed using blood obtained from participants finger pricks. Completely vaccinated referred to children who received 3 doses of pentavalent vaccine. Three drops of blood (60 μl) were collected on HemaSpot™ sampling devices, dried and stored at minus 80 °C in Nanoro and shipped to Hiroshima university for laboratory analysis. The HemaSpot™ comprises absorbent filter paper separated into eight fragments in a fan-shape with a central hole. Three fragments were detached and eluted at room temperature in 600 μL of elution buffer [Tris-buffered saline (TBS): 50 mM Tris, 150 mM NaCl, 0.1% Proclin 300 and 0.05% Tween 20 at pH 7.2]. The mixtures were stirred in shaking plates for 1 hour and subsequently centrifuged at 12000 rpm for 5 min at 4 °C and the supernatants were separated. Using Lumipulse G1200 (Fujirebio Inc. Inc., Japan), a chemiluminescent enzyme immunoassay (CLEIA) was used to detect HBsAg (Lumipulse® -II HBsAg, Fujirebio Inc., Japan with reported sensitivity of 100% and specificity of 99.7% [18]) with a cut-off index (COI) of 1.0, HBcAb (Lumipulse® HBcAb-N, Fujirebio Inc., Japan, with reported sensitivity of 88.4% and specificity of 95.2% [18]) COI of 0.6, HBsAb (Lumipulse® HBsAb-N, Fujirebio Inc., Japan, with reported sensitivity of 95% and specificity of 100% [18]) COI of 3.4, HBeAg (Lumipulse® -I HBeAg, Fujirebio Inc., Japan, with reported sensitivity of 97.6% and specificity of 98.2% [18]) COI of 1.0, HBeAb (Lumipulse® HBeAb-N Inc., Fujirebio, Japan, with reported sensitivity of 100% and specificity of 95.9% [18]) COI of 50.0. We considered the positivity of HBsAg or HBcAb sero-positivity as indicative of “HBV exposure”. Samples positive to HBsAb and negative for both HBsAg and HBcAb were classified as “serologically vaccinated”. Those with negative test results to HBsAg, HBsAb and HBcAb were designated as “susceptible” and at risk of HBV infection. To investigate HCV infection, HCV antibody (HCVAb) positivity (LumipulseII®Ortho- HCV, Ortho Clinical Diagnostics, Japan) with a cut-off value of 0.6 was measured using CLEIA.
Molecular assay and genome sequencing
In HBsAg positive samples, confirmatory tests were performed to detected and quantify HBV DNA by nucleic acid amplification test (NAT). At first, nucleic acid was extracted from HemaSpot™ using SMITEST EX-R&D (Medical and Biological Laboratories co., LTD, MA, USA). To detect HBV DNA, nested PCR assays with Applied Biosystems MiniAmp Plus Thermal Cycler (Thermo Fisher Scientific, Tokyo, Japan) were performed using primers obtained from Minegishi K et al work [19]. For viral load assessment, nucleic acid extracts were amplified and measured using a real-time PCR TaqMan Fast Universal PCR Master Mix (2X); Applied Biosystems, StepOne™ (Thermo Fisher Scientific, Tokyo, Japan) with primers and probe from HBV s-region gene [20, 21]. Hepatis C virus RNA was also detected using nested PCR amplification with in-house designed primer set from the conserved core region gene.
Genome sequencing for genotype analysis was conducted. To this regard, PCR products were generated using PrimeStar GXL (Takara Bio Inc., Shiga, Japan). Subsequently, a direct BigDye Terminator v3.1 Cycle Sequencing Kit; (Applied Biosystems, Foster city, CA, USA) was used.
Data management and statistical analysis
Study data were collected and managed using REDCap electronic data capture tools hosted at the Clinical Research Unit of Nanoro (CRUN). Data were analysed using Stata, Version 15 (StataCorp. 2017, TX). Proportions were estimated with 95% exact confidence intervals (95%CI). The χ2 or Fisher’s exact test were used as appropriate to compare proportions between groups. For risk factors of HBsAg positivity, HBV exposure and HBcAb positivity in children and mothers, crude odd-ratios (OR), adjusted OR and their 95% CIs were calculated by a univariate logistic regression and multivariate logistic-regression using a stepwise selection procedure (method of increasing and decreasing based on p-value threshold: both of probability to enter and probability to leave are 0.3) Covariate for analysis in children are age group, one pentavalent at least, completely vaccine, gender, ethnic, blood transfusion, birth assistance and traditional scarification. Covariate for analysis in mothers are age group, occupation, partner occupation, schooling status, birth place, pregnancy, genital mutilation, marital status, birth assistance, know about MTCT and Heard about hepatitis. P-value < 5% was considered statistically significant.
Ethical considerations
The study protocol was approved by the ethics committee for epidemiological research of Hiroshima University (Certificate E-1257) and the ethics committee of the ministry of health in Burkina Faso (Deliberation CERS-2018-02-023). All participants signed an informed consent before they can enter the study. Participants tested positive to hepatitis infection were referred to a specialized centre.