Comparison | No of studies | OR (95% CI) | p value* | Publication bias | Heterogeneity test† | |||
---|---|---|---|---|---|---|---|---|

Egger | Begg | p | I^{2} | H | ||||

OR, odds ratio; CI, confidence interval. | ||||||||

*Random effects models were used, weighted by the inverse variance. All statistical tests are two sided. | ||||||||

†p<0.1 is considered statistically significant for Q statistics; I^{2} is interpreted as the proportion of total variation contributed by between-study variation; H<1.2 suggests no heterogeneity among studies. | ||||||||

TNF2 v TNF1 | 14 | 1.37 (1.02 to 1.84) | 0.04 | p>0.2 | p>0.2 | <0.01 | 0.75 | 1.98 |

TNF2/2 v TNF1/1 | 8 | 2.01 (1.26 to 3.20) | 0.009 | p>0.2 | p>0.2 | 0.029 | 0.53 | 1.46 |

TNF2/2 v TNF2/1 | 8 | 1.51 (1.02 to 2.22) | 0.041 | p>0.2 | p>0.2 | >0.1 | 0 | 1 |

TNF2/1 v TNF1/1 | 11 | 1.47 (1.01 to 2.13) | 0.045 | p>0.2 | p = 0.04 | <0.01 | 0.74 | 1.97 |