CHAPTER 21 Summarizing and Graphing Survival Data 303

Recognizing that survival times

aren’t normally distributed

Even though survival times are numerical quantities, they’re almost never

normally distributed. Because of this, it’s generally not a good idea to use the

following:»

» ^{Means and standard deviations to describe survival times»}

» ^{T tests and ANOVAs to compare survival times between groups»}

» ^{Least-squares regression to investigate how survival time is influenced by}

other factors

If non-normality were the only problem with survival data, you’d be able to sum-

marize survival times as medians and centiles instead of means and standard

deviations. Also, you could compare survival between groups with nonparametric

Mann-Whitney and Kruskal-Wallis tests instead of t tests and ANOVAs. But time-

to-event data are susceptible to a specific type of missingness called censoring.

Typical parametric and nonparametric regression methods are not equipped to

deal with censoring, so we present survival analysis techniques in this chapter.

Considering censoring

Survival data are defined as the time interval between a selected starting point and

an endpoint that represents an event. But unfortunately, the time the event takes

place can be missing in survival data. This can happen in two general ways:»

» ^{You may not be able to observe all the participants in the data until they}

have the event. Because of time constraints, at some point, you have to end

the study and analyze your data. If your endpoint is death, hopefully at the

end of your study, some of the participants are still alive! At that point, you

would not know how much longer these participants will ultimately live. You

only know that they were still alive up to the last date they were measured in

the study as part of data collection, or the last date study staff communicated

with them in some way (such as through a follow-up phone call). This date is

called the date of last contact or the last-seen date, and would be the date that

these participants would be censored in your data.»

» ^{You may lose track of some participants during the study. Participants}

who enroll in a study may be lost to follow-up (LFU), meaning that it is no

longer possible for study staff to locate them and continue to collect data for

the study. These participants are also censored at their date of last contact,

but in the case of LFU, this date is typically well before the observation

period ends.