Criteria for considering studies for this review
Types of studies
We included RCTs only.
Types of participants
HBV-infected Pregnant women with HBV DNA-positive (HBV DNA > 103copies/ml) in sera of blood and their babies.
Types of intervention
ECS versus vaginal delivery.
Types of outcome measures
Primary outcomes (HBV transmission-related)
HBV-infection in neonates: HBV DNA-positive in umbilical blood or peripheral blood after birth.
Secondary outcomes (morbidities related to the actual method of delivery)
(1)Maternal morbidity: types of maternal morbidity evaluated includes: febrile morbidity, endometritis, hemorrhage or severe anemia, pneumonia, and urinary tract infections.
(2)Infant morbidity: types of infant morbidity evaluated includes: respiratory morbidity (respiratory distress syndrome and transient tachypnea of the newborn) and skin lacerations.
Search methods for identification of studies
We searched the electronic databases as follows: Cochrane Pregnancy and Childbirth Group's Trials Register (January, 2008), the Cochrane Central Register of Controlled Trials (the Cochrane Library 2008, issue 1), PubMed (1950 to 2008), EMBASE (1974 to 2008), Chinese Biomedical Literature Database (CBM) (1975 to 2008), China National Knowledge Infrastructure (CNKI) (1979 to 2008), VIP database (1989 to 2008). We also searched additional trials by scanning the reference lists of relevant trials identified. The search strategy was iterative as follows:
2 HBV INFECT*
3 HBV INFECTIONS
4 HBV INFECTED
5 #1 OR #2 OR #3 OR #4
6 DELIVERY, OBSTETRIC
7 DELIVERY AND PREGNANCY
8 CAESAREAN SECTION
9 "MODE OF DELIVERY" AND PREGNANCY
10 #6 OR #7 OR #8 OR #9
11 INFANT MORTALITY
12 INFANT MORBIDITY
13 NEONATAL MORTALITY
14 NEONATAL MORBIDITY
15 MATERNAL MORTALITY
16 MATERNAL MORBIDITY
17 POSTPARTUM MORTALITY
18 POSTPARTUM MORBIDITY
19 #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR#18
20 #5 AND #10 AND #19
21 (ANIMAL OF ANIMALS) NOT HUMAN
22 #20 NOT #21
Other search strategies
Organizations (including the World Health Organization), individual researchers working in the field were contacted in order to obtain possible additional references, unpublished trials, or ongoing trials, confidential reports and raw data of published trials.
Selection of studies
The titles, abstracts and keywords of every record retrieved were scanned to determine which were possibly relevant to the review. Any record that appeared likely to meet the inclusion criteria was obtained in full text. If there was any doubt regarding eligibility from the information given in the title and abstract, the full article was retrieved for clarification. Differences in opinion between reviewers were resolved by discussion.
Two review authors (ZX, MY) independently extracted data concerning details of the study population, interventions and outcomes using a standard data extraction form, specifically designed for this review. We resolved differences in data extraction by consensus, and with reference to the original article. If necessary, we sought information from the authors of the primary studies. For dichotomous outcomes, number of events and total number in each group were extracted. For continuous outcomes, mean, standard deviation and sample size of each group were extracted.
Assessment of risk of bias in included trials
The risk of bias was assessed based largely on the quality criteria specified by the Cochrane Handbook for Systematic Reviews of Interventions 5.0.0 [22
]. In particular, the following factors were studied:
• Selection bias: a) was the randomization procedure adequate? b) was the allocation concealment adequate?
• Performance bias: were the patients and people performing the intervention blind to the intervention?
• Attrition bias: a) were withdrawals, dropouts and losses of follow-up completely described? b) was analysis performed by intention-to-treat?
• Detection bias: were outcome assessors blind to the intervention?
Based on these criteria, studies were broadly divided into the following three categories. This classification was used as the basis of a sensitivity analysis. Additionally, we intended to explore the influence of individual quality criteria in a sensitivity analysis.
• A: all quality criteria met – low risk of bias.
• B: one or more of the quality criteria only partly met-moderate risk of bias.
• C: one or more criteria not met – high risk of bias.
Each trial was assessed by two reviewers independently (ZX, MY). Disagreements were resolved, where necessary, by recourse to a third reviewer (YJ). In cases of disagreement, the rest of the group were consulted and a judgment was made based on consensus.
Statistical analysis was carried out by using Review Manager (version 4.2). Dichotomous data were presented as relative risk (RR) and continuous outcomes as weighted mean difference (WMD), both with 95% confidence intervals (CI). The overall effect was tested by using Z score with significance being set at P < 0.05. Heterogeneity was tested by using the chi-squared statistic and I square (I2) with significance being set at P < 0.1. Possible sources of heterogeneity were to be assessed by sensitivity and subgroup analyses. A fixed-effect model was to be used when the studies in the subgroup were sufficiently similar (P > 0.10, I2 < 50). A random effects model was to be used in the summary analysis when there was heterogeneity between the subgroups. Publication bias was to be tested by using the funnel plot or other corrective analytical method, depending on the number of clinical trials included in the systematic review.