Competing risk syntax

[CEdward]

Does anyone know if Stata has syntas for competing risks survival analysis? The current syntax stcrreg is an option, but it is based on the Fine and Gray model. Is there syntax available for the cause-specific hazard?

[Andrea Discacciati]

You don't need a special command to estimate the cause-specific hazard function. Treat the competing event times as censored and use "standard" parametric / semi-parametric regression models (-streg-, -stpm2- from SSC, -stcox- (if you want to get cause-specific HRs leaving the baseline cause-specific hazard function unspecified))

[CEdward]

How would using stcox which is for standard Cox regression analysis also get you the cause-specific HRs?

[Steve Samuels]

Although it may seem counter-intuitive, Andrea's advice is absolutely correct. Read "The case for competing-risks regression" in the manual entry for stcrreg. It includes this paragraph:

When you have covariates, you can use stcox to perform regression on h1(t) by treating failures of type 2 as censored, on h2(t) by treating failures of type 1 as censored, or on h1(t) and h2(t) simultaneously by using the method of data duplication described by Lunn and McNeil (1995) and Cleves (1999). Because cause-specific hazards are identified by the data, all three of the above analyses are suitable for estimating how covariates affect the mechanism behind a given type of failure. For example, if you are interested in how smoking affects breast cancer in general terms (competing death notwithstanding), then a Cox model for h1(t) that treats death as censored is perfectly valid; see Pintilie (2007).