OptSurvCutR: Validated Cut-point Selection for Survival Analysis

The stratification of subjects based on continuous predictors is a common yet challenging task in time-to-event analysis (or survival), particularly when relationships are non-linear and require multiple thresholds. Arbitrary cut-point selection can lead to overfitting and spurious conclusions, highlighting the need for a statistically rigorous and reproducible framework.

We present OptSurvCutR (Optimal Survival Cut-Points in R programming), an R package designed to provide a complete, end-to-end workflow for robust and interpretable cut-point analysis. The package’s methodology is structured into a logical three-step process. First, the find_cutpoint_number() function provides a data-driven approach to determine the optimal number of cut-points by comparing models using information criteria (e.g., AIC, AICc or BIC). Second, the find_cutpoint() function identifies the precise locations of these thresholds by optimising survival metrics, such as the log-rank statistic or hazard ratio, using either an exhaustive systematic search or an efficient genetic algorithm. Finally, the validate_cutpoint() function assesses the stability of the identified thresholds by generating 95% confidence intervals through bootstrap resampling.

We demonstrate the package’s full workflow using two case studies: a plant science example modelling the non-monotonic effect of temperature on rapeseed germination and a clinical bioinformatics analysis stratifying colorectal cancer patients by a microbial biomarker. These examples show how OptSurvCutR can uncover complex survival patterns that would be missed by traditional dichotomisation. OptSurvCutR provides a transparent and extensible framework for validated cut-point selection, making it broadly applicable for researchers working with time-to-event data. The package, including source code and documentation, is freely available from GitHub at https://github.com/paytonyau/OptSurvCutR .