Vertical modelling: Analysis of competing risks data with missing causes of failure

Abstract

We propose vertical modelling as a natural approach to the problem of analysis of competing risks data when failure types are missing for some individuals. Under a natural missing-at-random assumption for these missing failure types, we use the observed data likelihood to estimate its parameters and show that the all-cause hazard and the relative hazards appearing in vertical modelling are indeed key quantities of this likelihood. This fact has practical implications in that it suggests vertical modelling as a simple and attractive method of analysis in competing risks with missing causes of failure; all individuals are used in estimating the all-cause hazard and only those with non-missing cause of failure for relative hazards. The relative hazards also appear in a multiple imputation approach to the same problem proposed by Lu and Tsiatis and in the EM algorithm. We compare the vertical modelling approach with the method of Goetghebeur and Ryan for a breast cancer data set, highlighting the different aspects they contribute to the data analysis.

Keywords

cause-specific hazard competing risks missing at random

Get full access to this article

View all access options for this article.

References

Park

. Parameter estimation of incomplete data in competing risks using the EM algorithm. IEEE Trans Reliab 2005; 54: 282–290.

Goetghebeur

Ryan

. Analysis of competing risks survival data when some failures types are missing. Biometrika 1995; 82: 821–833.

Tsiatis

. Multiple imputation methods for estimating regression coefficients in the competing risks model with missing causes of failure. Biometrics 2001; 57: 1191–1197.

Craiu

Duchesne

. Inference based on the EM algorithm for competing risks model with masked causes of failure. Biometrika 2004; 91: 543–558.

Liang

. Analysis of competing risks data with missing cause of failure under additive hazards model. Stat Sin 2008; 18: 219–234.

Nicolaie

van Houwelingen

Putter

. Vertical modelling: a pattern mixture approach for competing risks modelling. Stat Med 2010; 29: 1190–1205.

Cummings

Gray

Davis

Tormey

Harris

Falkson

. Tamoxifen versus placebo: double-blind adjuvant trial in elderly women with stage II breast cancer. Natl Canc Inst Monogr 1986; 1: 119–123.

Putter

Fiocco

Geskus

. Tutorial in biostatistics: Competing risks and multistate models. Stat Med 2007; 26: 2389–2430.

Rubin

. Inference and missing data. Biometrika 1976; 63: 581–592.

10.

Little

RJA

Rubin

. Theory of inference based on the likelihood function. In: Statistical analysis with missing data, New York: John Wiley & Sons, 1987, pp. 88–92.

11.

de Wreede

Fiocco

Putter

. The mstate package for estimation and prediction in non- and semi-parametric multi-state and competing risks models. Comput Methods Programs Biomed 2010; 99: 261–274.

12.

Van Houwelingen

. Dynamic prediction by landmarking in event history analysis. Scand J Stat 2007; 34: 70–85.

13.

Klein

Moeschberger

. Semiparametric proportional hazards regression with fixed covariates. In: Survival analysis: techniques for censored and truncated data, New York: Springer, 2003, pp. 257–258.

14.

Louis

. Finding the observed information matrix when using the EM algorithm. J R Stat Soc B 1982; 44(2): 226–233.