Does risk for ovarian malignancy algorithm excel human epididymis protein 4 and ca125 in predicting epithelial ovarian cancer: A meta-analysis

Table 5 Summary estimates of comparison among ROMA, HE4 and CA125 for EOC prediction

Settings (study numbers) [analysis model]	Mean Sen (95% CI)	Mean Spe (95% CI)	Mean DOR (95% CI)	Mean LR+ (95% CI)	Mean LR- (95% CI)	AUC (95% CI)
EOC- ROMA	0.86 (0.81-0.91)*	0.84 (0.79-0.88) *^#	32.72 (12.42-86.21)	5.35 (4.09-7.01)	0.17 (0.07-0.40)	0.92 (0.86-0.97)
(n = 3) [uni]	[I² =75.5%]	[I² =0.0%]	[I² =64.4%]	[I² =0.0%]	[I² =80.5%]
EOC-HE4	0.80 (0.73-0.85)	0.94 (0.90-0.96)	67.99 (31.97-144.60)	12.21 (4.25-35.11)	0.22 (0.17-0.29)	0.95 (0.92-0.98)
(n = 3) [uni]	[I² =0.0%]	[I² =78.2%]	[I² =19.2%]	[I² =79.0%]	[I² =0.0%]
EOC-CA125	0.84 (0.78-0.89)	0.78 (0.73-0.83) *	19.15 (7.26-50.53)	3.81 (2.66-5.46)	0.23 (0.09-0.58)	0.88 (0.81-0.96)
(n = 3) [uni]	[I² =90.0%]	[I² =79.6%]	[I² =68.8%]	[I² =41.4%]	[I² =87.6%]

Sen: sensitivity; spe: specificity; DOR: diagnostic odds ratio; LR+: positive likelihood ratio; LR-: negative likelihood ratio; CI: confidence interval; uni: univariate model (random effects model); * compared to HE4; ^# compared to CA125. Cells labeled with characters have significant difference (p < 0.05) in corresponding estimates with the compared groups. I² was calculated for estimates (sen, spe, DOR and LR±).

ISSN: 1471-2407