Table 2.

Subgroup analyses of volume effect on postoperative mortality in gastric cancer patients undergoing gastrectomy.

Subgroup	Number of patients	Number of studies	Odds ratio (95% CI)	P	Test of heterogeneity
					I 2 value (%)	P
Total	586 993	30	0.65 (0.56–0.76)	<0.001	81.9	<0.001
Study period
1982–1999	34 561	3	0.73 (0.64–0.83)	<0.001	0	0.774
2000–2018	519 749	24	0.61 (0.50–0.73)	<0.001	79.5	<0.001
Country
Eastern	418 575	9	0.50 (0.38–0.66)	<0.001	72.3	<0.001
Western	168 418	21	0.72 (0.61–0.85)	<0.001	79.3	<0.001
Sample size
<5000	32 871	16	0.80 (0.66–0.96)	0.016	45.4	0.025
≥5000	554 122	14	0.57 (0.46–0.70)	<0.001	88.3	<0.001
Hospital number
<100	15 062	10	0.77 (0.65–0.92)	0.003	0	0.968
≥100	507 542	12	0.58 (0.46–0.73)	<0.001	85.7	<0.001
Study design
Retrospective	577 117	26	0.65 (0.56–0.77)	<0.001	84.0	<0.001
Prospective	9876	4	0.66 (0.42–1.03)	0.066	27.7	0.246
Study quality
High	559 670	23	0.66 (0.58–0.76)	<0.001	72.0	<0.001
Moderate	27 323	7	0.54 (0.27–1.11)	0.092	92.1	<0.001
Volume grouping
Dichotomies	42 076	9	0.65 (0.58–0.74)	<0.001	0	0.852
Tertiles	319 182	11	0.75 (0.57–0.98)	0.036	89.7	<0.001
Quartiles	180 783	7	0.54 (0.35–0.84)	0.006	76.3	<0.001
Quintiles	44 952	3	0.57 (0.36–0.91)	0.019	75.8	<0.001
Adjusted OR
Yes	561 908	19	0.63 (0.54–0.73)	<0.001	77.7	<0.001
Unknown	25 085	11	0.69 (0.44–1.08)	0.102	87.4	0.001

OR, odds ratio.