I'm not sure there is a simple answer to this problem. The data are the data, and some events of interest can be low frequency. I'm assuming the data are right-censored, meaning that the subjects were lost to follow-up or the observation period ended before the event of interest happened. If much of the data was censored, perhaps the interveal of observation need to be longer to look for the event? Or, if there is an effect of interest, perhaps a power analysis can determine the number of necessary events needed to detect the effect of interest (e.g., the reduction of the hazard ratio by a certain amount).
I am not aware of options in PROC LIFETEST that address this. I've only used LIFETEST to describe the data (e.g., make Kaplan Meier curves or life tables).
