The Risk Set at Each Time Point
What you'll learn: How to identify which individuals are "at risk" of experiencing the event at any given moment in a survival study.
What Is the Risk Set?
In survival analysis, the risk set at a particular time point is the group of individuals who could experience the event at that exact moment. Think of it like a race: at the starting line, everyone is in the race. As runners drop out or finish, only those still running can potentially cross the finish line at the next checkpoint.
More formally, the risk set at time t includes:
- All individuals who have not yet experienced the event by time t
- All individuals who have not been censored before time t
Why Does It Matter?
The Kaplan-Meier estimator (which we've been building toward) calculates survival probabilities by looking at what fraction of the risk set experiences the event at each time point. You can't include someone who already experienced the event yesterday, or someone whose observation ended last week due to censoring.
A Simple Example
Imagine tracking five patients:
- At day 0: Risk set = 5 (everyone starts at risk)
- At day 3: One patient experiences the event → Risk set = 4
- At day 7: One patient is censored (lost to follow-up) → Risk set = 3
- At day 10: One patient experiences the event → Risk set = 2
Just before each event time, we count how many individuals were still "in the game" — that's our risk set.
Key Takeaway: The risk set at time t is the count of individuals who haven't yet experienced the event or been censored before t, representing everyone eligible to experience the event at that moment.