Table 4.

Statistical tests for publication bias by regression analysis of ln(DOR) against inverse of square root of effective sample size.

	Bias coefficient (SE)	P value	Intercept (SE)	P value
HE4	− 4.45 (7.46)	0.557	4.10 (0.56)	0.001
CA125	− 9.90 (5.79)	0.100	3.30 (0.44)	0.001
ROMA	− 15.30 (9.16)	0.110	4.32 (0.64)	0.001
ROMA-pre-menopause	0.24 (6.89)	0.973	3.00 (0.64)	0.001
ROMA-post-menopause	− 3.42 (4.99)	0.499	4.24 (0.51)	0.001

Human epididymis secretory protein 4 (HE4), Cancer antigen (CA125) and Risk of Ovarian Malignancy Algorithm (ROMA). Standard error (SE).