Included studies
A total of 1,553 articles were screened through search strategy (Fig. 1). After removal of duplicates, we screened 952 titles and abstracts for inclusion and reviewed the full texts of 142 studies. Finally, 34 articles met the inclusion and exclusion criteria, including 21 articles on immunogenicity [18–38] and 25 articles on safety [11, 18, 19, 21, 23–25, 27–31, 34, 39–50].
Table 1 summarizes the characteristics of the selected studies. Of the 21 studies on immunogenicity, 11 were randomized controlled trials [18, 21, 22, 24, 25, 27–29, 32, 35, 37] and 10 were cohort studies [19, 20, 23, 26, 30, 31, 33, 34, 36, 38]. Eighteen and seven studies evaluated the immunogenicity of 1 – [18, 19, 21–25, 28–38] and 2-dose [20, 22, 23, 26, 27, 31, 38] Varv, respectively. Five articles evaluated the immunogenicity of imported Varv [18, 22, 25, 32, 35]. Of the 25 studies on safety, 11 were randomized controlled trials [11, 18, 21, 24, 25, 27–29, 39, 44, 48] and 14 were cohort studies [19, 23, 30, 31, 34, 40–43, 45–47, 49, 50]. Twenty-three studies evaluated the safety of 1 dose of Varv [18, 19, 21, 23–25, 28–31, 34, 39–50]; five studies evaluated the safety of 2 doses of Varv [11, 23, 27, 31, 45]. The safety of imported Varv was evaluated in 3 articles [18, 25, 48].
Immunogenicity of domestic Varv and imported Varv
According to the random effects models, the pooled proportion of seroconversion in all populations after Varv vaccination was 89% (95%CI: 87%-92%). Studies showed high heterogeneity (I2 = 97.03%, P < 0.001). The pooled seroconversion proportion of domestic Varv was 89% (95%CI: 86%-91%) and that of imported Varv was 93% (95%CI: 88%-98%) (Fig. 2). There was no statistically significant difference in proportion of seroconversion between domestic Varv and imported Varv (χ2 = 1.97, P = 0.160).
In susceptible populations, the pooled seroconversion proportion of domestic Varv was 96% (95%CI: 95%-97%). The pooled seroconversion proportion of imported Varv was 93% (95%CI: 87%-100%) (Fig. 3). In non-susceptible populations, the pooled seroconversion proportion of domestic Varv was 80% (93%CI: 75%-86%). The pooled seroconversion proportion of imported Varv was 81% (93%CI: 67%-96%) (Fig. 3). There was no statistically significant difference in proportion of seroconversion between domestic Varv and imported Varv in susceptible (χ2 = 0.59, P = 0.440) and non-susceptible populations (χ2 = 0.02, P = 0.900).
Safety of domestic Varv and imported Varv
According to the random effects models, the pooled proportion of systemic reactions was 11% (95%CI: 10%-13%), showing high heterogeneity (I2 = 98.49%, P < 0.001). The pooled proportion of systemic reactions for domestic Varv was 11% (95%CI: 10%-13%). The pooled proportion of systemic reactions for imported Varv was 8% (95%CI: 6%-10%) (Fig. 4). The proportion of systemic reactions for domestic Varv was higher than that of imported Varv (χ2 = 8.04, P < 0.001). The pooled proportions of fever and headache reactions in domestic Varv were 9% (95%CI: 8%-10%) and 1% (95%CI: 1%-1%), respectively (Supplementary file 3: Fig. S1; Fig. S2). The pooled proportions of fever and headache reactions of imported Varv were 6% (95%CI: 5%-8%) and 0% (95%CI: 0%-1%), respectively (Supplementary file 3: Fig. S1; Fig. S2).
The pooled proportion of local reactions was 3% (95%CI: 2%-5%), showing high heterogeneity (I2 = 94.32%, P < 0.001). The pooled proportion of local reactions for domestic Varv was 3% (95%CI: 2%-3%). The pooled proportion of local reactions for imported Varv was 7% (95%CI: 3%-10%) (Fig. 4). The proportion of systemic reactions for domestic Varv was lower than that of imported Varv (χ2 = 5.30, P = 0.020). The pooled proportions of redness and swelling reactions of domestic Varv was 1% (95%CI: 1%-1%) and 1% (95%CI: 0%-1%), and the pooled proportions of redness and swelling reactions of imported Varv was 1% (95%CI: 0%-2%) and 1% (95%CI: 0%-1%) (Supplementary file 3: Fig. S3; Fig. S4).
Quality assessment
Of the 34 studies included, 15 randomized controlled trials had quality scores ranging from 2 to 7, with 5 having low scores (< 4) [24, 28, 32, 37, 48] and 10 having high scores (≥ 4) [11, 18, 21, 22, 25, 27, 29, 35, 39, 44]. The quality scores of 19 cohort studies ranged from 2 to 6, with 4 scoring low (< 4) [26, 34, 47, 50] and 15 scoring high (≥ 4 points) [19, 20, 23, 30, 31, 33, 36, 38, 40–43, 45, 46, 49]. See Supplementary file 2 for details.
Subgroup analyses
Results revealed strong heterogeneity among the selected studies. To determine potential factors of heterogeneity, we conducted subgroup analyses as detailed in Table 2 and Table 3. Because there are few studies on imported Varv, we performed subgroup analyses together with studies on domestic Varv. Subgroup analyses of immunogenicity showed a lower proportion of seroconversion in studies in 2015 and beyond (86%, 95%CI: 82%-89%) compared with studies prior to 2015 (95%, 95%CI: 93%-97%). According to the stratification of vaccine manufacturer, the pooled proportion of seroconversion in the vaccine studies of Changchun BCHT Biotechnology Co., Ltd. was higher (95%, 95%CI: 93%-98%), and the pooled proportion of seroconversion in the vaccine studies of Changchun Qijian Biological Products Co., Ltd. was lower (75%, 95%CI: 70%-80%). When stratified by study quality, high-quality studies had a lower proportion of seroconversion (87%, 95%CI: 84%-90%) compared with low-quality studies (95%, 95%CI: 92%-97%) (Table 2).
Table 2
Subgroup analyses of immunogenicity based on literature publication year, vaccine doses, manufacturer and quality score
Subgroup variable | Number of Studies | Serum sample size | Proportion of seroconversion (95% CI) | I2 (%) | Heterogeneity (χ2) | P value | Interaction test (χ2) | P value |
Age | ≤ 13 | 26 | 7,027 | 90 (88–92) | 95.90 | 609.54 | < 0.001 | 0.72 | 0.400 |
| > 13 | 6 | 2,931 | 86 (78–94) | 98.84 | 430.28 | < 0.001 | | |
Publication year | Before 2015 | 13 | 2,939 | 95 (93–97) | 90.08 | 121.00 | < 0.001 | 19.28 | < 0.001 |
| 2015 and beyond | 19 | 7,019 | 86 (82–89) | 97.83 | 830.35 | < 0.001 | | |
Vaccine doses | 1 dose | 22 | 6,485 | 91 (88–93) | 95.81 | 501.73 | < 0.001 | 2.04 | 0.150 |
| 2 doses | 10 | 3,473 | 86 (81–92) | 98.33 | 540.16 | < 0.001 | | |
Vaccine manufacturer | Shanghai Institute of Biological Products | 9 | 1,913 | 88 (82–93) | 93.21 | 117.90 | < 0.001 | 58.33 | < 0.001 |
| Changchun Qijian Biological Products | 5 | 2.021 | 75 (70–80) | 84.31 | 25.50 | < 0.001 | | |
| Changchun BCHT Biotechnology | 8 | 2,896 | 95 (93–98) | 93.62 | 109.65 | < 0.001 | | |
| Other domestic manufacturersa | 6 | 2,170 | 93 (90–97) | 94.09 | 84.59 | < 0.001 | | |
| Glaxosmithkline Biologicals | 4 | 958 | 91 (88–93) | 95.38 | 64.88 | < 0.001 | | |
Quality score | < 4 | 9 | 2,431 | 95 (92–97) | 92.42 | 105.51 | < 0.001 | 12.90 | < 0.001 |
| ≥ 4 | 23 | 7,527 | 87 (84–90) | 97.59 | 913.20 | < 0.001 | | |
CI: confidence interval. aOther domestic manufacturer: Changchun Changsheng Biotechnology(1), Beijing Wantai Biological Pharmacy Enterprise (1), Changchun Institute of Biological Products(1), and Shanghai Rongsheng Bio-Pharmaceutical (3). |
Table 3
Subgroup analyses of safety based on literature publication year, vaccine doses, manufacturer and quality score
Subgroup variable | Number of Studies | Serum sample size | Proportion of reactions (95% CI) | I2 (%) | Heterogeneity (χ2) | P value | Interaction test (χ2) | P value |
Systemic reactions |
Age | ≤ 13 | 36 | 28,381 | 10 (9–12) | 98.39 | 2171.87 | < 0.001 | 61.77 | < 0.001 |
| > 13 | 2 | 1,197 | 22 (19–24) | - | - | - | | |
Publication year | Before 2015 | 19 | 19,107 | 9 (7–11) | 98.31 | 1065.02 | < 0.001 | 7.25 | 0.010 |
| 2015 and beyond | 19 | 10,471 | 14 (11–17) | 98.67 | 1350.64 | < 0.001 | | |
Vaccine doses | 1 dose | 30 | 24,991 | 11 (10–13) | 98.49 | 1925.31 | < 0.001 | 0.07 | 0.790 |
| 2 doses | 8 | 4,587 | 10 (5–16) | 98.50 | 466.64 | < 0.001 | | |
Vaccine manufacturer | Shanghai Institute of Biological Products | 7 | 5,298 | 21 (10–32) | 98.92 | 553.36 | < 0.001 | 10.72 | 0.030 |
| Changchun Qijian Biological Products | 7 | 4,726 | 10 (4–17) | 98.27 | 346.28 | < 0.001 | | |
| Changchun BCHT Biotechnology | 12 | 13,521 | 8 (5–10) | 97.18 | 389.43 | < 0.001 | | |
| Other domestic manufacturersa | 10 | 5,312 | 13 (9–17) | 98.68 | 682.71 | < 0.001 | | |
| Glaxosmithkline Biologicals | 2 | 721 | 8 (6–10) | - | - | - | | |
Quality score | < 4 | 11 | 4,310 | 17 (11–23) | 98.74 | 793.21 | < 0.001 | 5.38 | 0.020 |
| ≥ 4 | 27 | 25,268 | 9 (8–11) | 98.37 | 1597.25 | < 0.001 | | |
Local reactions |
Age | ≤ 13 | 32 | 26,859 | 3 (2–3) | 94.30 | 544.34 | < 0.001 | 1.33 | 0.250 |
| > 13 | 2 | 1,197 | 3 (2–4) | - | - | - | | |
Publication year | Before 2015 | 20 | 19,165 | 3 (2–3) | 93.24 | 280.99 | < 0.001 | 1.13 | 0.290 |
| 2015 and beyond | 14 | 8,891 | 4 (2–5) | 95.33 | 278.41 | < 0.001 | | |
Vaccine doses | 1 dose | 29 | 24,381 | 2 (2–3) | 93.06 | 403.73 | < 0.001 | 13.20 | < 0.001 |
| 2 doses | 5 | 3,675 | 5 (4–6) | 65.95 | 11.72 | 0.020 | | |
Vaccine manufacturer | Shanghai Institute of Biological Products | 7 | 5,298 | 3 (2–4) | 76.52 | 25.55 | < 0.001 | 7.07 | 0.130 |
| Changchun Qijian Biological Products | 5 | 4,033 | 3 (0–5) | 92.91 | 56.45 | < 0.001 | | |
| Changchun BCHT Biotechnology | 9 | 12,634 | 4 (2–6) | 96.18 | 209.30 | < 0.001 | | |
| Other domestic manufacturersa | 10 | 5,312 | 2 (1–3) | 93.04 | 129.28 | < 0.001 | | |
| Glaxosmithkline Biologicals | 3 | 779 | 7 (3–10) | - | - | - | | |
Quality score | < 4 | 10 | 4,023 | 4 (3–5) | 77.85 | 40.63 | < 0.001 | 7.85 | 0.010 |
| ≥ 4 | 24 | 24,033 | 2 (2–3) | 94.26 | 400.96 | < 0.001 | | |
CI: confidence interval. aOther domestic manufacturer: Changchun Changsheng Biotechnology(2), Beijing Wantai Biological Pharmacy Enterprise (2), Changchun Institute of Biological Products(3), Shanghai Rongsheng Bio-Pharmaceutical (2), and Sinovac (Dalian) Vaccine Technology (1). |
For subgroup analyses of systemic reactions, there was a higher proportion of systemic with in studies in 2015 and beyond (14%, 95%CI: 11%-17%) compared with studies prior to 2015 (9%, 95%CI: 7%-11%). The pooled proportion of systemic reactions was higher in the vaccine studies of Shanghai Institute of Biological Products (21%, 95%CI: 10%-32%) and lower in the vaccine studies of Changchun BCHT Biotechnology (8%, 95%CI: 6%-10%). Higher quality studies had a lower proportion of systemic reactions (9%, 95%CI: 8%-11%) compared with lower quality studies (17%, 95%CI: 11%-23%) (Table 3). Sensitivity analysis showed that omission of any single studies did not significantly change the overall estimate, which indicated that our analysis was stable.