We conducted a case-control study at the INH of Lomé from October 2016 to March 2017. The study population consisted of individuals residing in Lomé who participated in a free screening campaign for HBsAg in October 2016. Information about the campaign was broadcasted on public radio and television to maximize the catchment of the campaign throughout the city of Lomé. A database containing serology data from the participants was built during this screening campaign. All participants were educated on the goals of the study and gave a written consent to participate. Prior to the study we obtained the approval of the INH internal ethics council.
Terms and definitions
Case: Anyone who participated in the free HBsAg screening test at INH in October 2016 and tested positive on the ELISA HBsAg assay.
Control: Anyone who participated in the free HBsAg screening test at INH in October 2016 and tested negative on the ELISA assay HBsAg.
Active viral replication: Simultaneous presence of HBsAg and HBeAg.
Acute phase of hepatitis B: Simultaneous presence of HBsAg and anti-HBc IgM antibodies.
Sampling and data collection
For a risk of 5%, a power of 80%, a case and control ratio of 1/3, a proportion of controls and cases with exposures of 40 and 57% respectively and an odd ratio of 2.1 we determined the minimal sample size to be 77 cases and 249 controls. We randomly selected controls and cases from records in database described in the study details. There were no exclusion criteria for the study. We gathered participant information through a questionnaire. In addition, we examined the laboratory register or the filled questionnaires to complete missing data and to delete duplications.
We were interested in socio-demographic information such as age, sex, ethnic group, level of education, marital status, and place of residence. We also considered knowledge about hepatitis B transmission route and means of prevention. These variables are of interest because they may correlate with behavioural measures to avoid acquisition of the virus. In addition, we documented possible avenues of exposure such as prior blood transfusion, surgery, piercing, vaccination, dental care or surgery, hospitalization, and neighbours suffering of hepatitis B. Finally, we tracked serological markers: HBsAg, HBeAg, anti-HBc IgM antibody, and hepatitis C markers.
All blood samples were analysed for the presence of the HBsAg by means of the ELISA with Monolisa Kit from Biorad made in France. Samples which tested positive for HBsAg were further analysed for the presence of the HBeAg and anti-HBc IgM antibody by Electro-chemiluminescence immuno-assay (ECLIA) on Cobas machine from Roche and for the presence of Hepatitis C virus markers by ELISA with Monolisa Kit from Biorad.
We completed the missing data, corrected the duplicates of the database and performed analyses using Epi Info 7. We calculated means, frequencies, proportions and measured associations by calculating the odds ratios (OR). We compared proportions by Fisher exact or Chi-square tests with a significance threshold of p ≤ 0.05.
Risk factors that were associated with HBsAg with p < 0.20 in bivariate analysis were introduced into a multivariable logistic regression model.
The explanatory variables were: previous history of vaccination, blood transfusion, hospitalization, surgery, dental care or surgery, ethnic group, sex, marital status, knowledge of HBV status, neighbour suffering of hepatitis B. The explained variable was the carriage of the HBsAg. We checked interactions of variables.
Subjects enrolled in the free screening gave a written consent. Prior to the use of the screening database, we obtained clearance from the internal ethics council of INH of Lomé.