Competing Risks — Fine-Gray + Aalen-Johansen CIF (multi-cause survival)
VerifiedAdvanced & specialized
Run this test straight away on a free built-in teaching dataset — no data of your own needed — or bring your own. Either opens the guided workspace: variable setup, assumption diagnostics, results with effect sizes and confidence intervals, figures, and APA-ready reporting.
Loading teaching datasets…
Or use your own dataset
Loading your datasets…
Time-to-event analysis when subjects can experience one of ≥ 2 mutually exclusive event types (e.g., cancer death vs other-cause death; graft rejection vs patient death).
Standard KM overestimates the cumulative incidence of each cause when competing events exist; the Aalen-Johansen estimator gives the correct CIF. Reports cause-specific CIF per group, Gray's test for between-group differences in CIF, plus cause-specific Cox AND Fine-Gray subdistribution-hazards regression. Standard tool in oncology, transplantation, cardiology, and aging research.
Worked example
Does a drug change the chance of relapse when death is a competing event?
Patients could relapse OR die without relapsing (a competing event). Cumulative incidence functions and Gray's test compare relapse incidence across arms without over-counting.
The end-of-follow-up cumulative incidence of relapse was lower on the drug (18% vs 30%), while death-without-relapse did not differ.
In a competing-risks analysis, the cumulative incidence of relapse was lower on the drug (18% vs 30%).
Try it yourself: Load this ready-made sample and follow the run above.
When to use it
- Two-group competing-risks comparisonTrial of adjuvant chemotherapy vs control in early-stage cancer.
- k-group competing-risks comparisonComparing 5-year cardiovascular death across 4 statin regimens (low, moderate, high intensity, ezetimibe), with non-CV death as competing event.
- Covariate-adjusted competing-risks regressionCancer-specific mortality adjusted for age, tumour stage, and treatment arm via Fine-Gray regression with non-cancer death as competing event.
When NOT to — use instead
- Single event type onlyUse standard Cox / KM — no competing event to adjust for. → Cox Proportional Hazards \u2014 multivariable survival with adjusted HRs
- Recurrent (non-terminal) eventsSubjects can experience the event multiple times — use Andersen-Gill or recurrent-event models. → Cox Proportional Hazards \u2014 multivariable survival with adjusted HRs
- Multi-state transitions (more than terminal causes)When subjects can transition through intermediate states, multi-state models are the appropriate framework. → Cox Proportional Hazards \u2014 multivariable survival with adjusted HRs
- Few events per cause (< 15)Fine-Gray and cause-specific Cox both need adequate events per cause for stable estimation. → Kaplan-Meier + log-rank \u2014 survival analysis with censoring
Hypotheses
Parameter tested: CIF per cause per group; log-subdistribution-hazard coefficients β + log-cause-specific-hazard coefficients β.
Assumptions (and what to do if they fail)
Check: Visual −log(1 − CIF) plot — parallel lines indicate PH.
If violated: Non-PH inflates / attenuates the subdistribution HR; the effect on the CIF may be time-varying.
Check: Per-group censoring rates within 30 pct pts.
If violated: Informative censoring biases the CIF and the subdistribution HR.
Check: Max |score residual| per subject.
If violated: An outlier can drive a single subdistribution coefficient.
Ready to run a Competing Risks — Fine-Gray + Aalen-Johansen CIF (multi-cause survival) on your own data?
Guided setup, automatic assumption checks, effect sizes, figures and an APA write-up.
Run this test →