Figure 3

Nelson-Aalen curves and estimated numbers at risk at each time point. (Left) Screenshot of the Nelson-Aalen curves in the original NEJM report of the ERSPC and (right) numbers at risk at each time point after randomization, derived from the PostScript file. The large numbers censored exactly at the end of follow-up years 8, 9, 10, and 11 are because the men in the Finnish portion of the trial were randomized on January 1, 1996, 1997, 1998, and 1999 and were still alive on December 31, 2006. The shallower slope of the curve in years 1 to 8 is due to deaths, while the steeper slope of the curve in years 9 to 13 reflects the staggered entries, beginning in different years in the seven different countries.