Skip to contents

Extract Survival Analysis Results from Kaplan-Meier Table Using Person-Years Method

Source: R/extractfromKM.R
extractfromKM.Rd

This function estimates hazard rates and median survival time from published Kaplan-Meier table data using the person-years method, assuming exponential distribution within each interval.

Usage

extractfromKM(time_points, n_risk, n_censored, warn_negative_events = TRUE)

Arguments

time_points

Numeric vector of time points (length n+1 for n intervals)

n_risk

Numeric vector of number at risk at each time point (length n+1)

n_censored

Numeric vector of cumulative number censored at each time point (length n+1)

warn_negative_events

Logical, whether to warn about negative events (default TRUE)

Value

List containing:

hazard_table

Data frame with interval summaries and hazard rates

median_survival

Estimated median survival time (NA if not reached)

Details

The function calculates:

  • Number of events in each interval (accounting for risk set changes)

  • Person-time using trapezoidal rule

  • Hazard rates as events/person-time

  • Median survival time using exponential survival function

Examples

# Example 1: Basic usage
time_points <- c(0, 6, 12, 18, 24)
n_risk <- c(200, 165, 130, 95, 65)
n_censored <- c(0, 15, 35, 53, 65)

result <- extractfromKM(time_points, n_risk, n_censored)
print(result$hazard_table)
#>   interval n_at_risk_start n_censored_interval n_events hazard_rate
#> 1    [0,6)             200                  15       20  0.01826484
#> 2   [6,12)             165                  20       15  0.01694915
#> 3  [12,18)             130                  18       17  0.02518519
#> 4  [18,24)              95                  12       18  0.03750000
print(paste("Median survival:", result$median_survival, "months"))
#> [1] "Median survival: NA months"