Skip to contentScholarGate
LibraryBookshelfDeskReview StudioAssistant
Sign in
On this page
IntuitionHow it worksWhen to use itStrengths & limitationsCommon pitfallsApplicationsFrequently asked🔒 Read the full methodSourcesRelated methods
Cite this pageSpotted an issue on this page? Report or suggest a fix →
Home›Epidemiology›Matched Competing Risks Analysis — Propensity-Matched Subdistribution Hazard Models
Process / pipelineClinical / epidemiology

Matched Competing Risks Analysis — Propensity-Matched Subdistribution Hazard Models

Matched Competing Risks Survival Analysis · Also known as: matched Fine-Gray analysis, propensity-matched competing risks, matched cause-specific hazard analysis, matched subdistribution hazard analysis

Matched competing risks analysis combines subject-level matching (e.g., propensity-score matching) with competing risks survival methods to estimate the cause-specific or subdistribution hazard of an event of interest while accounting for competing events that preclude the occurrence of that event. It is widely used in clinical and epidemiological observational studies where patients may die from causes other than the primary outcome of interest, and where treatment groups differ on baseline confounders.

ScholarGate
  1. Process / pipeline
  2. v1
  3. 2 Sources
  4. PUBLISHED
Cite this page →
Tools & resources
Download slides
Learn & explore

Read the full method

Members only

Sign in with a free account to read this section.

Sign in

Method map

The neighbourhood of related methods — select a node to explore.

Matched Competing Risks Analysis
Cox proportional hazardsInverse Probability Weig…Kaplan-Meier EstimatorPropensity Score Matching

When to use it

Use matched competing risks analysis when: (a) the study is observational and groups differ on measured confounders requiring matching or weighting; (b) patients can experience mutually exclusive events, and one or more of these competing events preclude the primary outcome; and (c) the research question concerns the absolute probability (CIF) of the event of interest, not merely relative hazards. Common clinical applications include cardiovascular outcomes research, oncology (cancer-specific vs. other-cause mortality), and infectious disease studies. Do NOT use standard Kaplan-Meier or Cox regression as a substitute when competing events are present — KM overestimates CIF. Do not use this approach when follow-up is very short and competing events are rare enough to be negligible, or when the study is a properly randomised trial where matching is unnecessary.

Strengths & limitations

Strengths
  • Correctly quantifies the absolute risk of the event of interest in the presence of competing events, avoiding the well-known overestimation from the complement of Kaplan-Meier.
  • Matching reduces confounding from measured covariates in observational data, approximating the covariate balance of a randomised design.
  • The Fine-Gray subdistribution hazard ratio directly models the covariate effect on the cumulative incidence function, which is the clinically and prognostically relevant quantity.
  • Supports both relative (SHR, CSH) and absolute risk (CIF difference) summaries, providing clinically actionable numbers.
  • Applicable across a wide range of clinical domains — cardiology, oncology, nephrology, infectious disease — wherever multiple failure types compete.
Limitations
  • Matching controls only for measured confounders; unmeasured confounding remains a threat to causal inference, as in all observational designs.
  • Propensity-score matching discards unmatched subjects, potentially reducing sample size and generalisability to the full cohort.
  • The Fine-Gray SHR mixes direct and indirect effects (via competing events), making causal interpretation more complex than cause-specific hazard models.
  • Correctly propagating matched-pair uncertainty through Fine-Gray models requires specialised software and careful implementation of robust variance estimators.
  • Very small event counts for either the primary or competing event reduce the precision of CIF estimates and the reliability of model-based inference.

Frequently asked

Why not just censor competing events and use standard survival analysis?

Censoring competing events treats them as non-informative — it assumes that subjects who experience a competing event would, if the competing event had not occurred, have the same subsequent risk as subjects still at risk. This assumption is almost universally violated in clinical data. Doing so inflates the estimated cumulative incidence of the event of interest. Competing risks methods directly model the joint distribution of event types, yielding unbiased cumulative incidence estimates.

Should I report Fine-Gray subdistribution hazard ratios or cause-specific hazard ratios?

Both serve distinct purposes and ideally both should be reported. The Fine-Gray SHR reflects the effect of an exposure on the observable cumulative incidence — the clinically and prognostically relevant quantity. The cause-specific hazard ratio reflects the instantaneous effect on event occurrence among those currently event-free, which is more relevant for understanding etiology. When they differ substantially, it signals that the exposure also affects the competing event rate.

Does matching eliminate the need for multivariable adjustment in the competing risks model?

Not entirely. Matching achieves balance on measured covariates in expectation, but residual imbalance can persist in any given matched sample due to chance. Including the matched covariates or the propensity score as additional regressors in the Fine-Gray model (doubly-robust approach) can further reduce residual confounding and is often recommended alongside matching.

What software can I use?

In R, the cmprsk package provides the crr() function for Fine-Gray models, and the crrSC package extends this to matched data with robust variance. The survival package supports cause-specific hazard models. The MatchIt package facilitates propensity-score matching. In Stata, the stcrreg command fits Fine-Gray models. SAS users can use PROC PHREG with the EVENTCODE option for competing risks.

How large a sample is needed?

There is no single rule, but a common guideline adapted from standard survival analysis is at least 10–15 events of the primary type per covariate included in the model. For matched designs, the effective sample size is determined by the number of matched pairs, not the total pre-match cohort size. Very small numbers of competing events can also destabilise the CIF estimates.

Sources

  1. Fine, J. P., & Gray, R. J. (1999). A proportional hazards model for the subdistribution of a competing risk. Journal of the American Statistical Association, 94(446), 496–509. DOI: 10.1080/01621459.1999.10474144 ↗
  2. Austin, P. C., Lee, D. S., & Fine, J. P. (2016). Introduction to the analysis of survival data in the presence of competing risks. Circulation, 133(6), 601–609. DOI: 10.1161/CIRCULATIONAHA.115.017719 ↗

How to cite this page

ScholarGate. (2026, June 3). Matched Competing Risks Survival Analysis. ScholarGate. https://scholargate.app/en/epidemiology/matched-competing-risks-analysis

Related methods

Cox proportional hazardsInverse Probability WeightingKaplan-Meier EstimatorPropensity Score Matching

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

  • Cox proportional hazardsEpidemiology↔ compare
  • Inverse Probability WeightingCausal inference↔ compare
  • Kaplan-Meier EstimatorStatistics↔ compare
  • Propensity Score MatchingResearch Statistics↔ compare
Compare side by side →

Similar methods

Risk-adjusted competing risks analysisProspective Competing Risks AnalysisRetrospective competing risks analysisCompeting Risks AnalysisFine-Gray Competing Risks ModelMulticenter Competing Risks AnalysisMeta-analytic competing risks analysisAdaptive Competing Risks Analysis

Related reference concepts

Competing RisksSurvival Analysis and Time-to-Event MethodsCox Regression ModelsCensoring and Follow-Up DataHazard RatioProportional Hazards Assumption

Spotted an issue on this page? Report or suggest a fix →

ScholarGate — Matched Competing Risks Analysis (Matched Competing Risks Survival Analysis). Retrieved 2026-07-21 from https://scholargate.app/en/epidemiology/matched-competing-risks-analysis · Dataset: https://doi.org/10.5281/zenodo.20539026
Quick facts
Originator
Fine & Gray (subdistribution hazard model); Austin, Lee & Fine (matched competing risks framework)
Year
1999 (Fine-Gray model); extended to matched designs ~2010s
Type
Observational survival analysis with matching and competing events
DataType
Time-to-event data with multiple mutually exclusive event types; matched pairs from observational cohorts
Subfamily
Clinical / epidemiology
Related methods
Cox proportional hazardsInverse Probability WeightingKaplan-Meier EstimatorPropensity Score Matching
ScholarGate

A content-first reference library for research methods — what each one is, how it works, and where it comes from.

Open data (CC-BY)

Explore

  • Library
  • Search the library…
  • Browse by field
  • Fields
  • Journey
  • Compare
  • Which method?

Reference

  • Subjects
  • Atlas
  • Glossary
  • Methodology
  • Philosophy

Your tools

  • Bookshelf
  • Desk
  • Chat

Company

  • About
  • Pricing
  • Contact
  • Suggest a method

Entries are compiled from published sources for reference. Verifying the accuracy and suitability of any information for your own use remains your responsibility.

© 2026 ScholarGate · A research-method reference library
  • Privacy
  • Cookies
  • Terms
  • Delete account