Figure 5.

Trial Sequential Analysis of chronic hepatitis C virus for interferon‐alpha versus no intervention was performed using an alpha error of 1.6%, power of 90% (beta error of 10%), relative risk reduction (RRR) of 20%, control group proportion observed in trials (Pc = 84%; upper figure) and Pc = 20% (lower figure), and observed diversity in the trials (0%). The upper figure with Pc = 84% shows that the accrued sample size was only a small fraction of the diversity‐adjusted required information size (DARIS). The Z‐curve (blue line) crosses the conventional boundaries (dotted green line), but it does not cross any of the trial sequential monitoring boundaries (dotted red lines). The lower figure with Pc = 20% shows that the accrued sample size was so small that trial sequential monitoring boundaries were not drawn. There is a high risk of random errors.