6.

Trial Sequential Analysis of all‐cause mortality at maximal follow‐up for transarterial chemoembolisation (TACE) versus percutaneous alcohol injection (PAI) versus PAI. We used an alpha error of 2.5%, power of 90% (beta error of 10%), a relative risk reduction (RRR) of 20% (upper figure) and that observed in trials (lower figure), control group proportion observed in the trials (30% mortality in about 2 to 3 years), and heterogeneity of 0% (upper figure) and that observed in the trials (I² = 75%) (lower figure). The accrued sample size (202 trial participants) is only a fraction of the information size (IS) (3351) or heterogeneity‐adjusted information size (HIS) (13,240 trial participants). As shown in all the comparisons, the cumulative Z‐curves (blue line) do not cross any of the trial sequential monitoring boundaries (red lines). They crossed the conventional alpha boundary of 2.5% (green line).