3 | Results

Our search strategy generated 3949 citations from 2 databases. 3773 articles were removed after exclusion of duplicates and screening of titles and abstracts. Of the remaining 176 studies, 163 studies were excluded after reviewing the full text. In total, 13 articles including 260,029 participants met the inclusion criteria and were included in the meta-analysis.11,24-35 The flow diagram of trial identification and selection is shown in fig 1. Descriptions and baseline characteristics of included studies are detailed in Table 1. No problems were encountered with participant data deficiency during the data integrity check.
Two cohorts were conducted in Americas, five in Europe and two in Asia. Two Cross-sectional studies were conducted in Europe, one was in Asia and one in America. One cohort study was conducted in Africa. None of the included studies were at low risk of bias (rated A) as all trials had an element of pragmatism in using different methods. Nine included studies, including 87% participants, were deemed to be at moderate risk of bias (rated B). We judged four studies including only 13% participants to have high risk of bias (rated C) in the field of study participation or statistical reporting.

3.1 | Association of BCG Vaccination in Childhood with Incidence of Allergic Disease

In the pooled analysis, we found that participants received BCG in childhood associated with a lower risk of allergic disease than that of non-BCG group (OR=0.86, 95%CI 0.75 to 0.97; fig 2).
12 studies involving 139035 participants reported the relationship between BCG vaccination and the risk of asthma. Compared with non-BCG group, received BCG in early childhood was associated with a significantly reduced risk of asthma (OR=0.74, 95%CI 0.61 to 0.91; fig 2).
8 studies including 58,825 participants reported the association between rhinitis and BCG vaccine and 9 studies including 61,109 participants studied eczema and BCG vaccine. Compared with the control group, use of BCG vaccine showed no significant effect in preventing eczema and rhinitis (OR=0.87 and 1.03, 95%CI 0.68 to 1.11 and 0.87 to1.22, respectively; fig 2).

3.2 | Association between BCG Vaccination and Demographics Factors on the Risk of Allergic Disease

We evaluated the association between BCG vaccination and participants demographics on the preventative effect of allergic disease. Participants from Europe and America were associated with a significantly lower risk of developing asthma when administered BCG vaccine in early childhood (pooled OR=0.59 and 0.90, 95%CI 0.40 to 0.88 and 0.82 to 0.98, respectively; fig 3). However, in Asia and Africa, participants who received BCG vaccine in early childhood were not associated with a significant reduced risk of allergic disease (pooled OR=0.97 and 1.16, 95%CI 0.51 to 1.87 and 0.68 to 1.97, respectively; fig 3).
Use of BCG vaccine was not associated with the risk of eczema in the subgroup analysis of different continents (fig 4). Similar results were obtained in participants with rhinitis (fig 5). The results from our study did not observe an association of BCG vaccination with reduction in risk of eczema or rhinitis in Europe, America, Asia or Africa.

3.3 | Publication Bias

The total publication bias is outlined in the funnel plot (fig 6). From visual inspections of the funnel plots and by Egger’s test36, it is suggested that publication bias did not impact our estimates. (bias coefficient for the main analysis 1.18, 95%CI 0.38 to 2.75, P=0.13)