Our study showed that, under the currently available immunoprophylaxis against hepatitis B, children of HBsAg-positive mothers delivered by ECS and VD had a similar prevalence of HBsAg, demonstrating that ECS does not reduce the risk of mother-to-child transmission of HBV. Therefore, ECS should not be used as a measure to prevent mother-to-child transmission of HBV.
Although HBV DNA may be detected in maternal blood and other body fluids of HBV carrier mothers and ECS may shorten the time of delivery [14
], we found in the present study that the HBV infection rate in the children delivered by ECS was comparable with that in the other children delivered by VD (Table ). Since the percentage of HBeAg-positive mothers, the use of hepatitis B vaccine and HBIG in children and other general characteristics were comparable between the ECS and VD groups (Table ), the infection rate in either group was not influenced by these factors. Additionally, the HBeAg-positive rate (24.8%) in HBV-infected pregnant women in the present study was in accordance with the rates reported in previous studies [15
], suggesting that the study subjects were representative. Since infants of HBeAg-positive mothers are more prone to be infected [17
], we further analyzed the HBV infection in the 137 children of HBeAg-positive mothers. The results demonstrated that ECS does not reduce the risk of mother-to-child transmission of HBV, even in the children born to HBeAg-positive mothers (Tables and ).
Before the availability of HBIG and hepatitis B vaccine, Chen et al. [8
] described a cohort of 23 infants born by ECS and 73 infants born by VD, whose mothers were all asymptomatic chronic HBV carriers; the mother-to-child transmission was similar in infants (≥ 6 months) delivered by ECS and VD (39.1% vs. 43.8%). In addition, the rates of acquisition of HBsAg were comparable between the infants born after the first stage of labor >9 hours and ≤9 hours (41.9% vs. 45.2%). The data show that prolonged labor and uterine contractions at delivery play little role in mother-to-child transmission of HBV. Wang et al. [9
] also indicated no significant effects of delivery mode on the prevention of mother-to-child transmission of HBV. Recently, using highly sensitive real-time PCR, Papaevangelou et al. [18
] reported that HBV DNA in peripheral blood of newborns was more often detected than HBsAg, however, no difference in the incidence of neonatal viremia was observed between the babies born by ECS and VD (21.9% vs. 26.5%, p
0.685). Our data in the present study are in accordance with these reported results.
Lee et al. [4
] advised ECS for HBeAg-positive pregnant women because they found higher rate of HBV infection in infants delivered vaginally within 6 months of age and serum HBV DNA at birth. Similarly, a recent study in India presented higher transmission of HBV in babies delivered by vaginal route than by cesarean section [5
]. However, these two studies did not follow the infants to 9–18 months old, the age point by which the perinatal infection can be defined [19
]. Furthermore, although a systematic review suggested that ECS appears to be effective in preventing mother-to-child transmission of HBV [6
], high risk of bias included in the analyses should not be overlooked.
It is concerning to see that the birth dose vaccine was delayed in 10.4% of the newborns and only half of the infants in the present investigation received HBIG within 24 hours after birth. The untimely use of first dose vaccine and the low rate of HBIG administration indicated that there are considerable gaps in the immunoprophylaxis against hepatitis B between the national recommendations and routine practices in China [20
]. Therefore, more measures should be taken in the future to achieve full adherence to the recommended prophylaxis in preventing mother-to-child transmission of HBV.
The main limitation in our study is that the mothers’ delivery modes were not randomly assigned. However, as mentioned above, maternal and neonatal general characteristics were comparable between ECS and VD groups. Furthermore, maternal HBeAg-positive rates and HBV DNA levels in HBeAg-positive mothers were also similar between the two groups. Thus, it was less likely the non-randomized design in the present study may result in significant bias. Additionally, ethical considerations will not allow such a randomized study. The other minor flaw is that the use of second and third doses of hepatitis B vaccine in some 10% of the children was defined by the interview with their mothers. However, we consider that the data were reliable because all newborns in China have received three doses of hepatitis B vaccine without charge since 2002 [21
]. Additionally, the overall HBsAg-positive rate (2.4%, 13/546) and the rate (71.6%) of anti-HBs ≥10 mIU/ml also indicate that these children had been vaccinated against hepatitis B.