Table 3. Meta-regression (inverse variance weights) for the potential source of heterogeneity.

Study characteristic	P value	RDOR	95% CI
PAX1 methylation
Patient size (<100 vs. ≥100)	0.4626	1.41	0.53–3.79
Control size (<100 vs. ≥100)	0.8981	0.93	0.28–3.12
Cut-off value (PMR % <10 vs. ≥10)	0.0873	1.51	0.94–2.44
Method (QMSP vs. MS-HRM)	0.0768	0.44	0.17–1.10
Study quality (QUADAS ≥10 vs. <10)	0.4623	1.75	0.36–8.59
SOX1 methylation
Patient size (<100 vs. ≥100)	0.4833	0.76	0.32–1.80
Control size (<100 vs. ≥100)	0.9282	1.07	0.17–6.55
Cut-off value (PMR % <10 vs. ≥10)	0.1031	0.61	0.33–1.14
Study quality (QUADAS ≥10 vs. <10)	0.6732	1.2	0.45–3.20
HPV DNA testing
Patient size (<100 vs. ≥100)	0.2042	0.40	0.08–2.02
Control size (<100 vs. ≥100)	0.3471	1.83	0.45–7.33
Method (QMSP vs. MS-HRM)	0.0552	3.51	0.96–12.77
Study quality (QUADAS ≥10 vs. <10)	0.3654	2.06	0.34–12.47

CI, confidence interval; RDOR, relative diagnostic odds ratio; QUADAS, quality assessment for studies of diagnostic accuracy; QMSP, quantitative methylation-specific polymerase chain reaction; MS-HRM, methylation-sensitive high-resolution melting; HPV, human papillomavirus.