jsurvival

The survival-analysis sub-module — eight analyses covering Kaplan-Meier, Cox and the data preparation around them; the advanced methods this page used to claim are in the umbrella, not here.

Purpose

Handle the common time-to-event questions in clinicopathological research, with publication-ready KM curves including risk tables and p-value annotations.

Data used

Any cohort with a time variable and an event variable. No project-specific data.

Methods

Package jsurvival, version 1.0.1, 8 analyses, 35 vignettes.

Analysis	What it is
survival	Survival Analysis — KM, log-rank
multisurvival	Multivariable Survival Analysis — Cox
survivalcont	Survival Analysis for Continuous Variable
singlearm	Single Arm Survival
outcomeorganizer	Outcome Organizer for Survival Analysis
timeinterval	Comprehensive Time Interval Calculator
datetimeconverter	DateTime Converter
oddsratio	Odds Ratio Table and Plot

Worth noticing that half of the module is data preparation — outcome coding, time-interval computation, date conversion. Given that the documented pitfall below names event-variable coding as the group’s most common mistake, that emphasis looks deliberate rather than incidental.

Correction: the advanced methods are in the umbrella

This page previously listed Nelson-Aalen, Turnbull NPMLE, AFT, frailty, landmark analysis, competing risks, multi-state models, recurrent events, cure models, IPCW for treatment switching, tumour growth kinetics, random survival forests, gradient-boosted survival and regularised Cox as capabilities of this module. Read from source on 2026-07-26, none of them is in this repository. They exist, but in ClinicoPath jamovi Module — flexmultistate, markovmultistate, curemodels, recurrentsurvival, finegray, landmarkanalysis, four frailty variants, robustaft, treatmentswitching, tumorgrowth, gradientboosting, lassocox, treeensemble, extratrees, highdimcox, ncvregcox.

The distinction is practical, not pedantic: installing the focused jsurvival module gets you the eight analyses above and none of the rest. The same pattern applies to all five sub-modules and is tabulated on the umbrella page.

Current state / open questions

Active; last commit 2026-07-16 at version 1.0.1, in step with the other five.

Documented pitfalls, all statistical: the event variable must be a factor with the event level explicitly specified — silent 0/1 coding errors are called out as the single most common mistake in the group; never report Cox hazard ratios without the Schoenfeld test, which the module prints by default; when a non-trivial fraction of patients die of other causes, cause-specific or Fine-Gray is correct rather than naive KM on disease-specific death — which now carries the caveat that Fine-Gray lives in the umbrella, so that advice requires the larger install.

Related reading

Concordance Index — the open question is answered. That page and this one both recorded that nothing documented which C-index variant was computed. The umbrella’s concordanceindex analysis offers Harrell’s method (default) and Uno’s IPCW method, with a time-dependent concordance option; this module’s multisurvival also reports a C-index with IPCW weighting. Antolini’s Ctd is not named anywhere. So the variant is documented after all — in the .a.yaml, which is not where anyone was looking.

Gradient-Based Feature Attribution — partly corrected. This page previously said the ML options report no feature importance. The umbrella has dedicated interpretability analyses — explainableai (“SHAP values, and other interpretability methods”), biomarkerdiscovery and clinicalprediction (both “SHAP/LIME”), plus variable importance in extratrees, highdimcox and ncvregcox. The original claim does hold for multisurvival specifically, where the SHAP option is present but commented out in the .a.yaml.

Survival Analysis Workflow — carries a covariate-admissibility rule this module’s documented pitfalls do not include: treatment-received variables must stay out of a prognostic model. Two external repositories evaluated here both broke it.

Related: ClinicoPath jamovi Module — the umbrella, and where every advanced survival method listed above actually lives.

Derived from: repository working tree read 2026-07-26 — DESCRIPTION, the 8 jamovi/*.a.yaml files, jamovi/multisurvival.a.yaml, vignettes/, plus git ls-files/git log; and ClinicoPathJamoviModule/jamovi/concordanceindex.a.yaml for the C-index finding. Earlier content derived from jamovi/jsurvival.qmd.