Table 3.

Effectiveness of locoregional treatment on overall survival in Cox proportional hazard model adjusted by propensity score. Sensitivity analysis.

Model	HR (95% CI)	P
RFA versus PEI (n = 47 + 253)
Adjustment by propensity score (n = 47 + 239)
Linear	1.11 (0.79–1.57)	0.53
Cubic spline	1.09 (0.77–1.54)	0.63
Stratified (quintiles)	1.11 (0.78–1.58)	0.56
Inverse probability weighting	1.13 (0.82–1.57)	0.46
Adjustment by prognostic covariates (n = 47 + 253)	1.21 (0.87–1.71)	0.25
Surgery versus RFA/PEI (n = 34 + 303)
Adjustment by propensity score (n = 34 + 255)
Linear	0.95 (0.64–1.41)	0.79
Cubic spline	0.95 (0.64–1.41)	0.81
Stratified (quintiles)	0.95 (0.61–1.48)	0.82
Inverse probability weighting	0.80 (0.52–1.24)	0.32
Adjustment by prognostic covariates (n = 34 + 303)	0.96 (0.64–1.42)	0.82
TAE/TACE versus RFA/PEI (n = 80 + 303)
Adjustment by propensity score (n = 80 + 287)
Linear	0.88 (0.66–1.17)	0.38
Cubic spline	0.89 (0.67–1.18)	0.41
Stratified (quintiles)	0.86 (0.65–1.15)	0.32
Inverse probability weighting	0.98 (0.73–1.32)	0.90
Adjustment by prognostic covariates (n = 80 + 303)	0.83 (0.63–1.10)	0.20

PEI: percutaneous ethanol injection, RFA: radiofrequency ablation, TACE/TAE: transarterial embolization with (TACE) or without (TAE) chemotherapy.