Lifetime Data Analysis

, Volume 24, Issue 3, pp 425–442 | Cite as

The competing risks Cox model with auxiliary case covariates under weaker missing-at-random cause of failure

  • Daniel NevoEmail author
  • Reiko Nishihara
  • Shuji Ogino
  • Molin Wang


In the analysis of time-to-event data with multiple causes using a competing risks Cox model, often the cause of failure is unknown for some of the cases. The probability of a missing cause is typically assumed to be independent of the cause given the time of the event and covariates measured before the event occurred. In practice, however, the underlying missing-at-random assumption does not necessarily hold. Motivated by colorectal cancer molecular pathological epidemiology analysis, we develop a method to conduct valid analysis when additional auxiliary variables are available for cases only. We consider a weaker missing-at-random assumption, with missing pattern depending on the observed quantities, which include the auxiliary covariates. We use an informative likelihood approach that will yield consistent estimates even when the underlying model for missing cause of failure is misspecified. The superiority of our method over naive methods in finite samples is demonstrated by simulation study results. We illustrate the use of our method in an analysis of colorectal cancer data from the Nurses’ Health Study cohort, where, apparently, the traditional missing-at-random assumption fails to hold.


Competing risks Masked cause of failure Missing-at-random Subtype analysis 



We thank two anonymous reviewers and the associate editor for insightful comments and suggestions that improved the paper.

Supplementary material

10985_2017_9401_MOESM1_ESM.pdf (317 kb)
Supplementary material 1 (pdf 317 KB)


  1. Bakoyannis G, Siannis F, Touloumi G (2010) Modelling competing risks data with missing cause of failure. Stat Med 29(30):3172–3185MathSciNetCrossRefGoogle Scholar
  2. Basu S, Sen A, Banerjee M (2003) Bayesian analysis of competing risks with partially masked cause of failure. J R Stat Soc Ser C (Appl Stat) 52(1):77–93MathSciNetCrossRefzbMATHGoogle Scholar
  3. Begg CB, Zabor EC, Bernstein JL, Bernstein L, Press MF, Seshan VE (2013) A conceptual and methodological framework for investigating etiologic heterogeneity. Stat Med 32(29):5039–5052MathSciNetCrossRefGoogle Scholar
  4. Campbell PT, Jacobs ET, Ulrich CM, Figueiredo JC, Poynter JN, McLaughlin JR, Haile RW, Jacobs EJ, Newcomb PA, Potter JD et al (2010) Case–control study of overweight, obesity, and colorectal cancer risk, overall and by tumor microsatellite instability status. J Natl Cancer Inst 102(6):391–400Google Scholar
  5. Cancer Genome Atlas Network (2012) Comprehensive molecular characterization of human colon and rectal cancer. Nature 487(7407):330–337Google Scholar
  6. Chatterjee N, Sinha S, Diver WR, Feigelson HS (2010) Analysis of cohort studies with multivariate and partially observed disease classification data. Biometrika 97(3):683–698MathSciNetCrossRefzbMATHGoogle Scholar
  7. Colussi D, Brandi G, Bazzoli F, Ricciardiello L (2013) Molecular pathways involved in colorectal cancer: implications for disease behavior and prevention. Int J Mol Sci 14(8):16,365–16,385CrossRefGoogle Scholar
  8. Cox DR (1972) Regression models and life-tables. J R Stat Soc Ser B (Methodol) 34(2):187–220MathSciNetzbMATHGoogle Scholar
  9. Craiu RV, Duchesne T (2004) Inference based on the EM algorithm for the competing risks model with masked causes of failure. Biometrika 91(3):543–558MathSciNetCrossRefzbMATHGoogle Scholar
  10. Dewanji A (1992) A note on a test for competing risks with missing failure type. Biometrika 79(4):855–857CrossRefGoogle Scholar
  11. Flehinger BJ, Reiser B, Yashchin E (2002) Parametric modeling for survival with competing risks and masked failure causes. Lifetime Data Anal 8(2):177–203MathSciNetCrossRefzbMATHGoogle Scholar
  12. Gao G, Tsiatis AA (2005) Semiparametric estimators for the regression coefficients in the linear transformation competing risks model with missing cause of failure. Biometrika 92(4):875–891MathSciNetCrossRefzbMATHGoogle Scholar
  13. Goetghebeur E, Ryan L (1995) Analysis of competing risks survival data when some failure types are missing. Biometrika 82(4):821–833MathSciNetCrossRefzbMATHGoogle Scholar
  14. Hughes LA, Williamson EJ, van Engeland M, Jenkins MA, Giles GG, Hopper JL, Southey MC, Young JP, Buchanan DD, Walsh MD et al (2012) Body size and risk for colorectal cancers showing BRAF mutations or microsatellite instability: a pooled analysis. Int J Epidemiol 41(4):1060–1072Google Scholar
  15. Hyun S, Lee J, Sun Y (2012) Proportional hazards model for competing risks data with missing cause of failure. J Stat Plan Inference 142(7):1767–1779MathSciNetCrossRefzbMATHGoogle Scholar
  16. Kalbfleisch JD, Prentice RL (2011) The statistical analysis of failure time data, vol 360. Wiley, New YorkzbMATHGoogle Scholar
  17. Kocarnik JM, Shiovitz S, Phipps AI (2015) Molecular phenotypes of colorectal cancer and potential clinical applications. Gastroenterol Rep 3(4):269–276Google Scholar
  18. Kuipers EJ, Grady WM, Lieberman D, Seufferlein T, Sung JJ, Boelens PG, van de Velde CJH, Watanabe T (2015) Colorectal cancer. Nat Rev Dis Primers 1:15,065 EP. doi: 10.1038/nrdp.2015.65 CrossRefGoogle Scholar
  19. Lu W, Liang Y (2008) Analysis of competing risks data with missing cause of failure under additive hazards model. Stat Sin 18(1):219–234Google Scholar
  20. Lu K, Tsiatis AA (2001) Multiple imputation methods for estimating regression coefficients in the competing risks model with missing cause of failure. Biometrics 57(4):1191–1197MathSciNetCrossRefzbMATHGoogle Scholar
  21. Lu K, Tsiatis AA (2005) Comparison between two partial likelihood approaches for the competing risks model with missing cause of failure. Lifetime Data Anal 11(1):29–40MathSciNetCrossRefzbMATHGoogle Scholar
  22. Moreno-Betancur M, Latouche A (2013) Regression modeling of the cumulative incidence function with missing causes of failure using pseudo-values. Stat Med 32(18):3206–3223MathSciNetCrossRefGoogle Scholar
  23. Nevo D, Zucker DM, Tamimi RM, Wang M (2016) Accounting for measurement error in biomarker data and misclassification of subtypes in the analysis of tumor data. Stat Med 35(30):5686–5700MathSciNetCrossRefGoogle Scholar
  24. Ogino S, Stampfer M (2010) Lifestyle factors and microsatellite instability in colorectal cancer: the evolving field of molecular pathological epidemiology. J Natl Cancer Inst 102(6):365–367CrossRefGoogle Scholar
  25. Ogino S, Chan AT, Fuchs CS, Giovannucci E (2011) Molecular pathological epidemiology of colorectal neoplasia: an emerging transdisciplinary and interdisciplinary field. Gut 60(3):397–411CrossRefGoogle Scholar
  26. Ogino S, Nishihara R, VanderWeele TJ, Wang M, Nishi A, Lochhead P, Qian ZR, Zhang X, Wu K, Nan H et al (2016) The role of molecular pathological epidemiology in the study of neoplastic and non-neoplastic diseases in the era of precision medicine. Epidemiology 27(4):602–611Google Scholar
  27. Prentice RL, Kalbfleisch JD, Peterson AV Jr, Flournoy N, Farewell V, Breslow N (1978) The analysis of failure times in the presence of competing risks. Biometrics 43(4):541–554CrossRefzbMATHGoogle Scholar
  28. Sen A, Banerjee M, Li Y, Noone AM (2010) A Bayesian approach to competing risks analysis with masked cause of death. Stat Med 29(16):1681–1695MathSciNetCrossRefGoogle Scholar
  29. Wang M, Kuchiba A, Ogino S (2015) A meta-regression method for studying etiological heterogeneity across disease subtypes classified by multiple biomarkers. Am J Epidemiol 182(3):263–270Google Scholar
  30. Wang M, Spiegelman D, Kuchiba A, Lochhead P, Kim S, Chan AT, Poole EM, Tamimi R, Tworoger SS, Giovannucci E, Rosner B, Ogino S (2016) Statistical methods for studying disease subtype heterogeneity. Stat Med 35(5):782–800. doi: 10.1002/sim.6793 MathSciNetCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2017

Authors and Affiliations

  • Daniel Nevo
    • 1
    Email author
  • Reiko Nishihara
    • 2
  • Shuji Ogino
    • 3
    • 4
    • 5
  • Molin Wang
    • 1
  1. 1.Departments of Biostatistics and EpidemiologyHarvard T.H. Chan School of Public HealthBostonUSA
  2. 2.Departments of Epidemiology and NutritionHarvard T.H. Chan School of Public HealthBostonUSA
  3. 3.Department of EpidemiologyHarvard T.H. Chan School of Public HealthBostonUSA
  4. 4.Department of Medical OncologyDana-Farber Cancer InstituteBostonUSA
  5. 5.Division of MPE Molecular Pathological Epidemiology, Department of PathologyBrigham and Womens Hospital and Harvard Medical SchoolBostonUSA

Personalised recommendations