Survival Analysis for Continuous Explanatory Variable

Details

This function provides advanced survival analysis specifically designed for continuous explanatory variables. It includes:

Core Features:

Optimal cut-off determination using maximally selected rank statistics
Multiple cut-offs analysis with 4 different methods (quantile, recursive, tree-based, minimum p-value)
Person-time analysis with interval stratification
Date-based time calculation with multiple format support
Multiple event level support (overall, cause-specific, competing risks)
Landmark analysis for time-dependent effects

Advanced Analytics:

Restricted Mean Survival Time (RMST) analysis
Cox model residual diagnostics (Martingale, Deviance, Score, Schoenfeld)
Log-log plots for proportional hazards assessment
Enhanced error handling and data validation

Visualization Options:

Kaplan-Meier survival curves with optimal cut-offs
Multiple cut-offs histogram with cut-point annotations
Cumulative events and hazard plots
KMunicate-style plots for publication
Residual diagnostic plots (4-panel layout)
Log-log plots for assumption checking

References

Hothorn, T., & Zeileis, A. (2008). Generalized maximally selected statistics. Biometrics, 64(4), 1263-1269.

Royston, P., & Parmar, M. K. (2013). Restricted mean survival time: an alternative to the hazard ratio for the design and analysis of randomized trials with a time-to-event outcome. BMC Medical Research Methodology, 13(1), 152.

Morris, T. P., et al. (2019). Proposals on Kaplan–Meier plots in medical research and a survey of stakeholder views: KMunicate. BMJ Open, 9(9), e030874.

Super classes

jmvcore::Analysis -> ClinicoPath::survivalcontBase -> survivalcontClass

Methods

Public methods

survivalcontClass$clone()

Inherited methods

Method `clone()`

The objects of this class are cloneable with this method.

Usage

survivalcontClass$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.

Examples

if (FALSE) { # \dontrun{
# Basic survival analysis with optimal cut-off
data("lung", package = "survival")
lung$status_binary <- ifelse(lung$status == 2, 1, 0)

result1 <- survivalcont(
  data = lung,
  elapsedtime = "time",
  outcome = "status_binary", 
  contexpl = "age",
  findcut = TRUE,
  sc = TRUE
)

# Multiple cut-offs analysis with different methods
result2 <- survivalcont(
  data = lung,
  elapsedtime = "time",
  outcome = "status_binary",
  contexpl = "ph.karno",
  multiple_cutoffs = TRUE,
  num_cutoffs = "three",
  cutoff_method = "recursive",
  min_group_size = 15,
  sc = TRUE
)

# RMST analysis with residual diagnostics
result3 <- survivalcont(
  data = lung,
  elapsedtime = "time",
  outcome = "status_binary",
  contexpl = "wt.loss",
  findcut = TRUE,
  rmst_analysis = TRUE,
  rmst_tau = 500,
  residual_diagnostics = TRUE,
  loglog = TRUE
)

# Person-time analysis with date calculation
# Create sample data with dates
set.seed(123)
n <- 200
sample_data <- data.frame(
  biomarker = rnorm(n, 100, 25),
  event = rbinom(n, 1, 0.6),
  dx_date = as.Date("2020-01-01") + sample(0:365, n, replace = TRUE),
  fu_date = as.Date("2020-01-01") + sample(366:1095, n, replace = TRUE)
)

result4 <- survivalcont(
  data = sample_data,
  tint = TRUE,
  dxdate = "dx_date",
  fudate = "fu_date",
  timetypedata = "ymd",
  timetypeoutput = "months",
  outcome = "event",
  contexpl = "biomarker",
  person_time = TRUE,
  time_intervals = "6, 12, 24",
  rate_multiplier = 1000,
  calculatedtime = TRUE
)

# Comprehensive analysis with all features
result5 <- survivalcont(
  data = lung,
  elapsedtime = "time",
  outcome = "status_binary",
  contexpl = "meal.cal",
  findcut = TRUE,
  multiple_cutoffs = TRUE,
  num_cutoffs = "two",
  cutoff_method = "quantile",
  rmst_analysis = TRUE,
  rmst_tau = 400,
  residual_diagnostics = TRUE,
  person_time = TRUE,
  time_intervals = "100, 300, 500",
  sc = TRUE,
  ce = TRUE,
  ch = TRUE,
  kmunicate = TRUE,
  loglog = TRUE,
  ci95 = TRUE,
  risktable = TRUE,
  calculatedcutoff = TRUE,
  calculatedmulticut = TRUE
)
} # }