Risk factors for intrafamilial spread of hepatitis B in northeastern Bosnia and Herzegovina.

BACKGROUND: Accurate estimations of hepatitis B virus transmission risk for any region in Bosnia and Herzegovina are not clearly established. We aimed to determine levels of risk associated with intrafamilial transmission of hepatitis B infection within families in our region. PATIENTS AND METHODS: Family members of 81 chronic carriers of hepatitis B surface antigen (>6 months positive and considered as index case) were tested for hepatitis B markers. For family members, we recorded their age, sex, and family relationship to the index case, and vaccination status. RESULTS: The proportion of HBsAg positive family members was 25/207 (12.1%), while the proportion of family members with evidence of exposure to HBV was 80/207 (38.6%). Only 17/207 (8.2%) family members had evidence of past HBV vaccination. Age was found to be a significant predictor of HBV exposure of family members (odds ratio 1.05, 95% CI 1.03-1.07, P<.001). In a multivariate analysis, HBsAg positivity was associated with a female index case (odds ratio 11.31, 95% CI 3.73-34.32, P<.001), HBeAg positivity in the index case (odds ratio 5.56, 95% CI 1.80-17.23, P<.005) and being a mother of the index case (odds ratio 9.82, 95% CI 2.43-39.68, P<.005). A female index case (odds ratio 4.87, 95% CI 2.21-10.72, P<.001), HBeAg positivity in the index case (odds ratio 3.22, 95% CI 1.15-9.00, P<.05) and being a mother of the index case (odds ratio 3.72, 95% CI 1.19-11.64, P<.05) were also risk factors for HBV exposure among family members. The combination of HBeAg positivity and female index case was a significant predictor for HBsAg positivity of family members (odds ratio 70.39, 95% CI 8.20-604.61, P<.001). CONCLUSIONS: Children of HBeAg positive mothers are at highest risk for becoming chronic carriers themselves and generally, the combination of female sex and HBeAg positivity dramatically increases the chances of HBV transmission within the family.

T he World Health Organization (WHO) div v vides the world into areas of low, intermediv v ate and high endemicity for hepatitis B virus (HBV) infection. Bosnia and Herzegovina, according to WHO estimates and published reports, is in the area of intermediate endemicity for HBV infection, with an HBsAg prevalence of 3.6% among firstvtime blood donors. 1 Transmission of HBV within family or household contacts is not a rare phenomenon and family members of chronic HBV carriers are a wellvesv v tablished risk group. 2v8 Although the precise means of HBV spread within families is not well defined, there are several risk factors associated with it. 4,5,7 Several studies have evaluated routes of intrafamilial transmisv v sion of HBV in our region, and the difference in HBV exposure between family members of chronic HBV carv v

Risk factors for intrafamilial spread of hepatitis B in northeastern Bosnia and Herzegovina
riers and firstvtime blood donors. 3,7v9 However, the true levels of risk associated with facilitated HBV spread within the family have not been analyzed in detail. We conducted a hospitalvbased study with the intention to analyze levels of risk for risk factors associated with inv v trafamilial transmission of hepatitis B infection within families in our region, thus determining the pattern of risk factors that eventually results in more HBVvposiv v tive family members of chronic hepatitis B patients.

PATIENTS AND METHODS
The study was conducted prospectively during a 2vyear period (2004v2006) at the Tuzla University Hospital, Bosnia and Herzegovina, and was approved by the University Clinical Center Tuzla Ethical Committee. Index cases were 81 consecutive patients that satisfied the criteria of being positive for HBsAg for a minimum of 6 months and were the first detected chronic HBsAg carv v riers within their families. They were tested for HBsAg and hepatitis B e antigen (HBeAg) with the ELISA test (Abbott Diagnostics, Wiesbaden, Germany), and their age and sex were noted. We defined chronic HBsAg carriers as those positive for HBsAg for at least a 6vmonth period. Exposure to HBV was defined as positivity for any HBV marker in the absence of HBV vaccination evidence. We offered HBV serology testing to family members of all chronic HBsAg carriers living in the same household. Two hundred and seven famv v ily members of index cases were tested for hepatitis B markers (HBsAg, HBeAg, antivHBsAg, antivHBcAg, antivHBeAg) and their age, sex, family relationship to the index case and vaccination status were noted.
Statistical tests were performed using SPSS 15.0 (SPSS Inc, Chicago, IL, USA). Index cases were strativ v fied by age, sex, and presence of HBeAg, while family members were stratified by sex, age, presence of HBV markers, relationship to index case, size of their correv v sponding family and vaccination status. Standard tests for descriptive statistics were used for determination of baseline characteristics of groups. Betweenvgroup differences in frequencies were investigated by the chiv square test. The Spearman rankvorder correlation was used for determining correlation between age, HBV exv v posure and HBsAg seropositivity. Univariate logistic rev v gression was used to determine factors associated with HBsAg positivity or HBV exposure. Factors found to be significant in univariate analysis were tested using stepwise multivariate regression analysis. All tests were performed with a 95% statistical significance level.
The proportion of HBsAgvpositive family members was 25/207 (12.1%), while the proportion of famv v ily members with evidence of exposure to HBV was 80/207 (38.6%). The highest proportion of HBsAgv positive family members was found among parents of index cases, while the highest ratio of HBV exposure was found within siblings and parents, respectively (Table 1). For family members whose index case was a female there were significantly more HBsAg positives than among those whose index case was a male (Table  2). Correspondingly, there were significantly more famv v ily members with evidence of HBV exposure among those with a female index case than among those with a male index case. Family members of HBeAg positive index cases (10/28; 35.7%) had a higher proportion of HBsAg positives (c 2 =14.56; P<.001) in compariv v son to family members of HBeAg negative index cases (15/179; 8.4%). The proportion of HBV exposed was also higher (c 2 =16.32; P<.001) among family members of HBeAg positive (21/28; 75%) than among HBeAg negative index cases (59/179; 33%). We found a positive but not strong correlation between age and HBV expov v sure in family members (Spearman' s r=0.35; P<.001), yet we found no significant correlation between age and HBsAg seropositivity (Spearman' s r=v0.04; P=.54).
Using univariate logistic regression analysis, signifiv v cant predictors of HBsAg positivity were a female inv v    03v1.07, P<.001), while family member sex and being a spouse of an index case were not significant predictors (P>.05). We used factors that were significant in the univariate regression to test their significance and level of risk in a multivariv v ate analysis for association with HBsAg positivity and HBV exposure (Table 3). HBeAg positivity and female index case combined was a significant predictor for HBsAg positivity of family members with an odds ratio of 70.39 (95% CI 8.20v604.61; P<.001). Nevertheless, the combination of HBeAg positivity and female index case was not associated with HBV exposure of family members (P=.98).

DISCUSSION
Reports from countries in our region report a prevav v lence of HBsAg in the general population that ranges from 2% to 10% in endemic areas with an overall HBV prevalence of 14% to 60%. 3,7,10,11 Several studies that analyzed rates of HBV exposed family members of HBsAg carriers reported overall HBV prevalence rates ranging from 23% to 48%. 3,7,9 We found comparable rates of HBV exposure in family members of chronic HBV carriers in our sample.
That the highest rate of HBV exposure and HBsAg positivity was among parents of index cases (Table 1) emphasizes the vertical route of transmission within families in our region. We have also found a relatively high number of HBV exposed cases (47.5%) among spouses of index cases, but the ratio of HBsAg posiv v tive cases among them was very low (1.7%). This rev v flects the fact that sexual transmission, although less predominant in our setting, usually does not end with a chronic HBsAg carrier.
It is well known that about 90% of newborns inv v fected perinatally become chronic HBsAg carriers and most of them eventually end up with chronic hepativ v tis B. 12 Vaccination programs for all infants have been established in Bosnia and Herzegovina since 2001, yet most of the adult population was not vaccinated. As a result, we have few family members (1/12 of the samv v ple) with evidence of vaccination.
Patients positive for HBeAg are considered to be more infectious and they usually have higher titers of HBV DNA. 13 Our figures support this since we found significantly more HBV exposed and HBsAg carriers among family members of HBeAg positive index cases.
It has been reported in previous studies that rates of HBV exposure are positively correlated with age. 14, 15 We have also found that each additional year of age within family members increases the chances of HBV exposure by 5% on average. This is a logical consequence of number of exposures and number of infection reserv v voirs within a family that increases with age.
We have demonstrated that female gender, HBeAg positivity and being a mother of an index case are strong and significant independent risk factors associated with HBV spread within the family. This is a result of a domiv v nant mothervtovchild transmission pattern, which is the most significant means of HBV spread in our sample. The presence of HBeAg facilitates this process further, rendering family members of an HBeAg positive female index carrier around 70 times more susceptible to bev v come chronic HBV carriers themselves.
It is important to note some limitations of this study. All included index cases are from one hospital and therefore, these cases cannot necessarily represent the northveastern region of Bosnia. The exact occurrence of HBV infection of each index case is very hard to dev v termine and determination of the time sequence of this infection in relation to other family members is even harder. Therefore, labelling the first detected HBsAg carrier within the family is purely arbitrary and is as best a guess as possible. This methodological approach had been used in number of previous studies. 3,5,7 All of our results indicate that the main risk factors for HBV transmission inside the family are female sex, HBeAg seropositivity in the index case and an HBsAgv positive mother in the family. This pattern of risk factors facilitates the vertical, mothervtovchild route of transv v mission, which leads to a higher percentage of chronic HBsAg carriers, especially in younger age groups. Age appears to be another important risk factor for HBV exposure, as a direct result of the number of exposures to HBV and infection reservoirs within the family.
Children of HBeAg positive mothers are at the highest risk for becoming chronic carriers themselves and generally, the combination of female gender and HBeAg positivity dramatically increases the chances of HBV transmission within the family. Considering the intolerably low rate of HBV vaccination within family members, it is of essential importance to insist on the vaccination of this risk group in our region. This leads to the possibility that it would be a prudent strategy to investigate costveffectiveness of prepartal screening of pregnant women for HBsAg positivity, as a first step toward a nationwide program of pregnancy screening for HBsAg that would essentially block the main route of HBV spread in our region. Also, an important adv v dition to this strategy would be the screening and vacv v cination at the high school age or university age, or at the time of marriage to protect the susceptible cases born before 2001.