What is Survival Analysis?
Survival analysis is a statistical technique used to analyze the time-to-event data, where the event of interest can be failure, death, recurrence, or any other outcome. It involves modeling the probability that an individual will experience the event over time.
Key Concepts:
- Time-to-event: The time it takes for the event to occur.
- Survival function: The probability that an individual has not experienced the event at a given time (S(t)).
- Hazard function: The instantaneous rate of occurrence of the event at a given time (h(t)).
Types of Survival Analysis:
- Kaplan-Meier estimator: A non-parametric method for estimating the survival function.
- Cox proportional hazards model: A semi-parametric model that estimates the hazard ratio between groups.
Example:
Suppose we want to investigate the relationship between smoking and lung cancer mortality using a cohort study.
Data:
- We have 1,000 individuals with no prior history of lung cancer.
- 500 individuals are smokers (Group A), and 500 individuals are non-smokers (Group B).
- The follow-up period is 5 years.
- Mortality rates are recorded at the end of each year.
Kaplan-Meier Estimator:
| Time | Survival Function (S(t)) |
| --- | --- |
| 1 | 0.95 |
| 2 | 0.85 |
| 3 | 0.75 |
| 4 | 0.65 |
| 5 | 0.55 |
The Kaplan-Meier estimator plots the survival function for both groups, showing that smokers have a higher mortality rate than non-smokers.
Cox Proportional Hazards Model:
Using the Cox model, we can estimate the hazard ratio (HR) between smoking and lung cancer mortality.
| Covariate | Hazard Ratio (HR) |
| --- | --- |
| Smoking | 2.3 |
The Cox model estimates that smokers are approximately 2.3 times more likely to die from lung cancer than non-smokers, controlling for other factors.
Interpretation:
Survival analysis provides a powerful framework for analyzing time-to-event data in statistics. The Kaplan-Meier estimator and Cox proportional hazards model can be used to identify significant relationships between covariates and the event of interest. In this example, smoking is associated with a higher mortality rate from lung cancer.
I hope this summary and example help!