Fig. 2.

Estimated sample sizes under the two assumptions in the Prevention Of Pelvic (POPI) trial. Plotted curves represent the estimated sample size needed per group while varying power of the study; for the original POPI trial (green lines) and for two types of progression; the one where PID develops at a constant rate throughout infection (dashed-dotted lines), and the one where PID develops at the end of infection (dashed lines). The third type of progression where PID develops immediately after infection is not shown. Panels a and b separate the two different assumptions about the projected follow-up incidence of PID in the original POPI trial: scenario 1 with 2 % per year (Panel a), and scenario 2 with 3 % per year (Panel b). The red circle represents the sample size with 80 % power in the original POPI trial. PID, pelvic inflammatory disease; POPI trial, Prevention Of Pelvic Infection trial