Competing event regression on the relative subdistribution and cumulative-incidence scales

In competing risks settings, covariate effects and group comparisons are usually assessed one event at a time—through log-rank or Cox tests on the cause-specific hazards, or Gray’s test or Fine–Gray regression on a cumulative incidence function (CIF). This can obscure a clinically important quantity: the ratio between the event of interest and the competing event, since groups may differ little on the individual events yet differ sharply in their ratio. The generalized competing event (GCE) framework makes this ratio the object of inference; on the cause-specific scale the hazard ratio ω ⁺ ( t ) = λ ₁ ( t ) / λ ₂ ( t ) is estimated efficiently from a single stacked (Lunn–McNeil) model. We extend the framework to two scales that describe realized incidence. The subdistribution hazard ratio is estimated by a stacked, risk-set-weighted extension of the Lunn–McNeil construction; the cumulative-incidence ratio ρ ( t ) = F ₁ ( t ) /F ₂ ( t )—the odds that a subject’s realized event by time t is the event of interest—by jackknife pseudo-observation regression of the Aalen–Johansen estimator. We relate the three contrasts: ρ equals ω ⁺ exactly under proportional cause-specific hazards, and equals only in the small-time limit under proportional subdistribution hazards, drifting toward 1 thereafter. The orthogonality that makes ω ⁺ efficient is lost on both cumulative-incidence scales— through overlapping weighted risk sets and shared censoring weights, ρ through the shared all-cause survivor—so each carries a covariance term that must be handled and that bounds efficiency relative to the hazard-scale test. We derive the corresponding variances, study operating characteristics by simulation, illustrate on hypothetical prostate and head-and-neck cohorts, and provide an implementation in the gcemod R package.