Cox Model Survival Analysis

About the Dataset

The dataset, haberman.csv, includes data from a study conducted between 1958 and 1970 at the University of Chicago's Billings Hospital. It focuses on the survival of breast cancer surgery patients.

• Number of Instances: 306
• Number of Attributes: 4
  • Age of patient at time of operation (numerical)
  • Patient's year of operation (year - 1900, numerical)
  • Number of positive axillary nodes detected (numerical)
  • Survival status (class attribute):
    • 1 = Survived 5 years or longer
    • 2 = Died within 5 years
• Missing Values: None

Files

• haberman.csv: The dataset used for survival analysis.
• cox-survival.md: R Markdown file for performing survival analysis.

Usability

• License: Unknown
• Expected Update Frequency: Not specified

Analysis Overview

1. Introduction: Explains the dataset and the objectives of the survival analysis.
2. Load Required Libraries: Lists and loads the libraries needed for the analysis.
3. Data Import and Preview: Imports the dataset and provides an initial preview.
4. Data Profiling: Provides diagnostic statistics for numeric variables.
5. Defining Time and Event Variables: Sets up variables for survival analysis.
6. Kaplan-Meier Estimator: Fits and plots the Kaplan-Meier survival curve.
7. Stratified Kaplan-Meier Curves: Creates and plots survival curves for different patient cohorts based on positive axillary nodes detected.
8. Cox Proportional Hazards Model: Fits a Cox model to the data and summarizes the results.
9. Visualizing Cox Model Coefficients: Generates a forest plot of Cox model coefficients.
10. Predicting Survival Curves for Specific Patients: Predicts and plots survival curves for selected patients.

Example Code

Below is an example of loading the dataset and performing a Kaplan-Meier survival analysis:

library(tidyverse)
library(survival)
library(survminer)
library(dlookr)
library(gridExtra)

# Load the dataset
data <- read.csv('haberman.csv', header = FALSE)
colnames(data) <- c('Age', 'Operation_year', 'Nb_pos_detected', 'Surv')

# Kaplan-Meier Estimator
km_fit <- survfit(Surv(Age, Surv) ~ 1, data = data)
ggsurvplot(km_fit, data = data, conf.int = FALSE, ggtheme = theme_minimal(), title = "Kaplan-Meier Estimate")

Conclusion

The analysis explores patient survival probabilities using Kaplan-Meier estimation and Cox Proportional Hazards modeling. It provides insights into how various factors influence survival outcomes for breast cancer patients.

---

Feel free to adjust any section or add additional details as needed!