Skip to main content

Table 4 Result of meta-analysis, publication bias and fill-trim method for prevalence estimate and corresponding 95% confidence interval of gastric cancer in H.pylori positive patients

From: The global prevalence of gastric cancer in Helicobacter pylori-infected individuals: a systematic review and meta-analysis

Subgroup

Meta-analysis

Publication bias (Egger’s test)

Fill-trim method

NS

Heterogeneity index

Pooled prevalence%

(95% CI)

Coefficient

(95% CI)

p-value

Adj-pooled prevalence%

(95% CI)

Study design

 Cross sectional

69

I2 = 98.59%; p < 0.001

19.46 (18.34 to 20.57)

7.09 (5.82 to 8.36)

 < 0.001

7.89 (6.78 to 9.01)

 Prospective cohort

13

I2 = 98.99%; p < 0.001

2.49 (2.09 to 2.90)

8.59 (4.33 to 12.84)

0.001

1.13 (0.65 to 1.61)

 Prospective case series

56

I2 = 97.70%; p < 0.001

23.13 (20.41 to 25.85)

6.07 (5.15 to 6.98)

 < 0.001

16.23 (13.76 to 18.69)

 Retrospective case series

8

I2 = 98.66%; p < 0.001

11.14 (8.09 to 14.19)

6.30 (-1.45 to 14.05)

0.094

–-

 Retrospective cohort

3

I2 = 0.10%; p < 0.001

1.17 (0.55 to 1.78)

5.79 (-7.04 to 18.62)

0.110

–-

Continents

 Asia

114

I2 = 98.62%; p < 0.001

12.96 (12.38 to 13.55)

6.33 (2.03 to 10.63)

0.010

4.37 (0.03 to 8.75)

 America

20

I2 = 98.84%; p < 0.001

18.06 (16.48 to19.63)

6.89 (5.87 to 7.92)

 < 0.001

6.43 (7.02 to 21.43)

 Africa

6

I2 = 88.39%; p < 0.001

9.52 (5.92 to 13.12)

3.41 (-3.42 to 10.26)

0.239

–-

 Europa

9

I2 = 98.40%; p < 0.001

16.26 (12.02 to 20.50)

8.09 (5.54 to 10.64)

 < 0.001

7.10 (5.58 to 8.63)

  1. CI Confidence interval, NS Number of studies