Skip to main content
. 2020 Jan 13;18:9. doi: 10.1186/s12957-020-1786-2

Table 3.

Subgroup meta-analyses of the relationship between AGR and OS via multivariate analyses

Potential confounding factor No. of studies No. cases Hazard ratio with 95%CI P value I2 (%) for Heterogeneity P value for Heterogeneity
Overall survival (OS) 17 6704 1.75 (1.54–2.00) 0.000 34.0 0.084
Methods for choosing AFR cut-off value
 X-tile 6 3060 1.47 (1.19–1.81) 0.000 34.6 0.177
 ROC 9 3085 2.07 (1.76–2.43) 0.000 0.0 0.886
 Score 2 559 1.62 (1.06–2.47) 0.024 50.2 0.156
Cut-off value of AFR
 > 9.7 7 3215 1.76 (1.36–2.28) 0.000 61.4 0.017
 ≤ 9.7 8 2930 1.82 (1.53–2.15) 0.000 0.0 0.623
 Score = 0 2 559 1.62 (1.06–2.47) 0.024 50.2 0.156
Study designed type
 Retrospective 14 5437 1.78 (1.52–2.08) 0.000 42.6 0.046
 Prospective 3 1267 1.71 (1.31–2.23) 0.000 0.0 0.457
Number of cases
 < 360 9 1948 2.04 (1.66–2.50) 0.000 21.5 0.252
 ≥ 360 8 4756 1.53 (1.35–1.74) 0.000 2.7 0.409
Cancer classification
 Lung cancers 3 1211 1.83 (1.44–2.33) 0.000 0.0 0.956
 Digestive cancers 10 3700 1.58 (1.32–1.90) 0.000 43.9 0.075
 Gynecological cancers 2 499 2.16 (1.47–3.17) 0.000 0.0 0.944
 Other cancers 3 1294 2.37 (1.68–3.34) 0.000 0.0 0.431
Publication time
 After 2019 5 1441 1.70 (1.33–2.17) 0.000 23.9 0.262
 Before 2019 12 5263 1.79 (1.52–2.11) 0.000 42.0 0.062
Treatment option
 Surgical resection 12 5298 1.73 (1.50–2.01) 0.000 33.4 0.123
 Others 5 1406 1.80 (1.31–2.49) 0.000 45.3 0.120