Sixteen case-control studies involving 21,986 women (5,568 cases and 16,418 controls) were included.
Current IUD use.
Pregnant controls: OR 10.63 (95% CI: 7.66, 14.74), chi-squared homogeneity 23.37 (P < 0.01); adjusted for confounder, OR 6.29 (95% CI: 4.23, 9.34), chi-squared homogeneity 1.59 (P > 0.05).
Weighting by quality scores and inverse variance, and including confounding, OR 6.37 (95% CI: 4.09, 9.92).
Non-pregnant controls: OR 1.06 (95% CI: 0.91, 1.24), chi-squared homogeneity 133.29 (P < 0.01); adjusted for confounder, OR 1.06 (95% CI: 0.89, 1.28), chi-squared homogeneity 59.96 (P < 0.01).
Weighting by quality scores and inverse variance, and including confounding, OR 1.00 (95% CI: 0.83, 1.32).
Past IUD use.
Pregnant controls: OR 1.33 (95% CI: 1.08, 1.63), chi-squared homogeneity 13.55 (P > 0.05); adjusted for confounder, OR 1.38 (95% CI: 1.12, 1.70), chi-squared homogeneity 8.78 (P > 0.05).
Weighting by quality scores and inverse variance, and including confounding, OR 1.37 (95% CI: 1.08, 1.72).
Non-pregnant controls: OR 1.45 (95% CI: 1.23, 1.72), chi-squared homogeneity 5.99 (P > 0.05); adjusted for confounder, OR 1.45 (95% CI: 1.23, 1.72), chi-squared homogeneity 5.99 (P > 0.05).
Weighting by quality scores and inverse variance, and including confounding, OR 1.46 (95% CI: 1.12, 1.74).
Total pooled past IUD use: OR 1.40 (95% CI: 1.23, 1.59), chi-squared homogeneity 19.99 (P > 0.05); adjusted for confounder, OR 1.42 (95% CI: 1.25, 1.62), chi-squared homogeneity 14.92 (P > 0.05).
Weighting by quality scores and inverse variance, and including confounding, OR 1.42 (95% CI: 1.24, 1.64).
The average quality scored was 0.78.
Funnel plots used to assess the possibility of publication bias showed asymmetry round the estimated pooled OR for current IUD use and pregnant controls, and approximate symmetry round the pooled OR for current IUD use and pregnant controls, and for past IUD use regardless of controls.