Bias associated with delayed verification in test accuracy studies: accuracy of tests for endometrial hyperplasia may be much higher than we think!

Table 3 Results of meta-regression analysis.

Outcome Explanatory variables	Univariable analysis		Multivariable analysis
	Ratio dOR (95% CI)^†	P value	Ratio dOR (95% CI)^†	P value
Endometrial hyperplasia
▪ Delay in verification of diagnosis (> 24 hours vs. < 24 hours)	0.31 (0.08–0.84)	0.089	0.26 (0.07–0.99)	0.048
▪ Study quality (category II vs. category I)	1.36 (0.34–5.53)	0.664	3.46 (0.79–15.0)	0.098
▪ Test (Endometrial biopsy vs. ultrasound endometrial thickness)	3.22 (0.83–12.55)	0.091	5.0 (1.2–20.7)	0.027

^†The dependent variable is the natural logarithm of the diagnostic odds ratio (dOR). Results are presented as the ratio of diagnostic odds ratios (RdOR); an RdOR < 1 means that the diagnostic accuracy is reduced and a RdOR > 1 means that it is increased in relation to the reference category, < 24 h, category I, and ultrasound endometrial thickness, respectively.

ISSN: 1741-7015