Table 2.

Sensitivity analysis for the progression from SIL to cervical cancer in the HPV-positive subset

N = number of studies (number of individuals)
OR (95% CI)
P value
Subset analysisRecessive modelDominant modelAdditive model
HR-HPV onlyN = 14 (1,638)N = 12 (1,392)N = 12 (1,392)
1.40 (1.14–1.72)1.09 (0.74–1.60)1.25 (0.99–1.57)
P = 0.001P = 0.67P = 0.06
Methodologically soundN = 10 (1,335)N = 10 (1,335)N = 10 (1,335)
1.42 (1.15–1.76)1.38 (0.93–2.04)1.30 (1.04–1.62)
P = 0.001P = 0.11P = 0.02
Optimal tissue sourceN = 7 (919)N = 6 (786)N = 6 (786)
1. 26 (0.91–1.74)0.90 (0.58–1.42)1.08 (0.78–1.49)
P = 0.16P = 0.66P = 0.41
Optimal genotyping methodN = 15 (1,753)N = 14 (1,640)N = 14 (1,640)
1.44 (1.19–1.76)1.24 (0.92–1.69)1.30 (1.05–1.61)
P < 0.001P = 0.16P = 0.01
WhiteN = 9 (979)N = 9 (979)N = 9 (979)
1.59 (1.26–2.01)1.27 (0.70–2.29)1.40 (1.07–1.87)
P < 0.001P = 0.43P = 0.01
East AsianN = 7 (921)N = 5 (675)N = 5 (675)
1.24 (0.92–1.67)1.62 (1.00–2.65)1.27 (0.93–1.73)
P = 0.16P = 0.05P = 0.13
Total: Cervical cancer vs. any SILN = 19 (2,339)N = 17 (2,093)N = 17 (2,093)
1.37 (1.15–1.62)1.30 (0.98–1.73)1.25 (1.04–1.51)
P < 0.001P = 0.07P = 0.02