Pseudo-observations methods for direct modeling of survival probabilities and restricted mean survival time (RMST). Enables regression analysis of survival probability at specific time points and RMST with standard statistical methods.
Usage
pseudosurvival(
time_var,
status_var,
covariates = NULL,
stratification_vars = NULL,
analysis_type = "survival_probability",
time_points = "12,24,36,60",
tau_rmst = 60,
quantile_probs = "0.25,0.5,0.75",
jackknife_method = "standard",
regression_method = "ols",
cluster_var = NULL,
confidence_level = 0.95,
bootstrap_samples = 500,
robust_se = FALSE,
model_diagnostics = TRUE,
prediction_intervals = FALSE,
survival_curves = TRUE,
rmst_comparison = TRUE,
sensitivity_analysis = FALSE,
competing_risks = FALSE,
weighted_analysis = FALSE
)Arguments
- time_var
Survival time variable
- status_var
Event status indicator variable
- covariates
Covariates to include in the pseudo-regression model
- stratification_vars
Variables for stratifying the analysis
- analysis_type
Analysis approach using pseudo-observations
- time_points
Time points for pseudo-observation calculation
- tau_rmst
Maximum follow-up time for RMST analysis
- quantile_probs
Survival quantiles for pseudo-observation analysis
- jackknife_method
Method for computing pseudo-observations
- regression_method
Statistical method for pseudo-observation regression
- cluster_var
Variable defining clusters for GEE or cluster jackknife
- confidence_level
Confidence level for parameter estimates
- bootstrap_samples
Bootstrap resamples for standard error estimation
- robust_se
Whether to compute robust standard errors
- model_diagnostics
Whether to include model diagnostics
- prediction_intervals
Whether to compute prediction intervals
- survival_curves
Whether to plot survival probability curves
- rmst_comparison
Whether to perform RMST group comparisons
- sensitivity_analysis
Whether to perform sensitivity analysis
- competing_risks
Whether to include competing risks modeling
- weighted_analysis
Whether to apply inverse probability weighting
Value
A results object containing:
results$instructions | a html | ||||
results$pseudo_summary | a html | ||||
results$regression_results | a html | ||||
results$covariate_effects | a html | ||||
results$rmst_analysis | a html | ||||
results$survival_probability_results | a html | ||||
results$group_comparisons | a html | ||||
results$model_diagnostics_summary | a html | ||||
results$sensitivity_results | a html | ||||
results$pseudo_observations_plot | an image | ||||
results$regression_plots | an image | ||||
results$survival_curves_plot | an image | ||||
results$rmst_comparison_plot | an image | ||||
results$diagnostics_plots | an image | ||||
results$sensitivity_plot | an image |
Examples
# Example: RMST regression with covariates
pseudosurvival(
data = cancer_data,
time_var = survival_time,
status_var = death_status,
covariates = c("age", "stage", "treatment"),
analysis_type = "rmst_regression",
tau_rmst = 60
)