Epidemiology of hepatitis E virus infection in pregnant women in sekondi-Takoradi Metropolis, Ghana

A thesis, Degree of Doctor of Philosophy

Infections caused by hepatitis E virus (HEV) constitute a global public health burden. Worldwide, over 20 million new cases of asymptomatic infections, 3.4 million illnesses and 70 000 deaths due to HEV occur each year. Infections with HEV are the leading cause of oro-fæcally acquired viral hepatitis and outbreaks have occurred in more than 61 countries. In developing countries particularly in Asia and Africa HEV outbreaks cause 10-30% case fatality rates in pregnant women and the infection is more common and fatal during the third trimester of pregnancy. Several reports in Ghana have pointed to the country being HEV endemic and sporadic maternal deaths and abortions due to HEV infections. However, there is not a surveillance system for HEV infections and the Ghana Ministry of Health (MOH) expressed concerns over knowledge gaps about the infection and particularly among the most HEV-vulnerable pregnant population.This is important because in the absence of vaccine to control HEV infection individual actions by the affected pregnant women to lower the effect of risk factors becomes necessary. There is also the policy implication of this study. In the absence of vaccines to control HEV infections the strategy of education becomes an important tool to lower the effects of the risk factors. The objectives of this thesis therefore were to determine seroprevalence of subclinical infections of HEV in the third trimester of pregnancy, risk factors and the delivery outcomes due to infections with HEV in the Sekondi-Takoradi Metropolis, Ghana. Asymptomatic and apparently healthy pregnant women in third trimester, of 18 years and above were purposively selected in a cross-sectional study. The third trimester was selected because this is the period reported to be associated with most HEV-related vulnerabilities of abortions, faetal and maternal mortalities. To reduce confounding effects on estimates pregnant women with hepatitis B, C and active liver disease profiles examined earlier as part of antenatal care in routine checks and found not to be free of these infections were excluded from the study. Because reports have associated human immunodeficiency virus (HIV) with increased infections of HEV infections those with HIV have also been excluded. Blood samples were collected and analyzed for HEV infection. Socio-demographic and household data were collected. Household proximity to wetlands and domestic pig farms were estimated and farmed swine from two districts from where the sampled pregnant women originated were also tested for HEV infection. Data on diagnoses and mode of delivery of the pregnant women were collected from books on their discharge and also from the Regional Hospital Database. Bivariate and multivariate logistic regression (LR) analysis was done in SPSS version 20 and by Microsoft Excel, respectively. R code version 4.1 was used to check model assumptions and geographical dependence in the dataset. Anti-HEV IgM was 22.5% (81/360) 95% CI:18.2-26.8: and the anti-HEV IgG 11.0% (11/100) 95% CI: 5.6-18.8. In bivariate LR analysis statistically significant associations were found between recent HEV infections and age-groups, level of education, access to household flash water toilet systems. Three out of 12 (25%) domestic pig farms were infected with HEV. Proximity to 20% (5/20) farms were significantly associated with recent infections with HEV. Infection with HEV was significantly associated with complications at delivery (P = 0.029: OR 1.24 95% CI 1.021-1.496). Surprisingly, among 22 diagnoses recorded in ward discharge log books, normal pregnancy (NP) was the only significant outcome (P = 0.000: OR 0.349 95% CI 0.232-0.525), an indication of recent infections with HEV protective of normal deliverys. Seroprevalence of active infection with HEV in the third trimester of pregnancy is 22.5% in Sekondi Takoradi Metropolis of Ghana. The absence of water closet toilets and proximity to domestic pig farms are risk factors for HEV infection, whilst proximity to wetlands is not indicative of an infection risk to HEV. Active infection with HEV is not indicative of having an adverse outcomes on deliveries at delivery. Metropolitan, municipal and local government authorities should increase support for water closet toilet facilities in households. Improved swine farm effluent management systems should also be encouraged. Health authorities should encourage public awareness creation about HEV infection in pregnancy. Even though the outcomes during delivery of recent HEV infections appear to be protective of normal pregnancy, clinical realities may be different. Therefore, research institutions should consider swine HEV as a One Health challenge to reduce the potential of swine-HEV risks for the development of liver cancers in pregnant women. The current study provided insights for dealing with geospatial analysis and provided additional analytic approach in dealing with geospatial data. The seroprevalence data on recent HEV infections adds to existing data on HEV during pregnancy and the first study of HEV in swine in Ghana. In the absence of vaccine against HEV infections education against HEV during pregnancy by the Ministry of Health and related institutions will improve disease prevention and control.