Plot Spline Effect from Cox Proportional Hazards Model

Generates a plot showing the spline-transformed effect of a continuous variable in a Cox model using predicted marginal effects with confidence intervals.

Usage

cox_plot_spline(
  data_model,
  spline_var = "age_bs",
  title = "Spline effect",
  xlab = NULL,
  ylab = "Predicted risk",
  color = "#0072B2"
)

Arguments

data_model: A fitted Cox proportional hazards model object created with `coxph()` and spline terms (e.g., using `splines::bs()`).
spline_var: Character string specifying the name of the spline variable in the model to visualize (e.g., `"age_bs"` or `"bmi"`).
title: Character string specifying the title of the plot. Default is `"Spline effect"`.
xlab: Optional label for the x-axis. If `NULL`, the variable name will be used.
ylab: Label for the y-axis. Default is `"Predicted risk"`.
color: Character string with a valid color for the line and confidence ribbon. Default is Bootstrap blue `"#0072B2"`.

Value

A `ggplot` object showing the effect of the spline-transformed variable with a confidence ribbon.

Details

The function uses `ggeffects::ggpredict()` to compute marginal effects across the full range of the spline variable. It assumes the model was created using natural splines via `splines::bs()` and includes the variable named in `spline_var`.

Examples

if (FALSE) { # \dontrun{
library(survival)
library(splines)

# Example model with spline
model <- coxph(Surv(time, status) ~ splines::bs(age, df = 4), data = lung)

# Plot the spline effect
p <- cox_plot_spline(model, spline_var = "age", title = "Effect of Age")
print(p)
} # }