Lifetime Data Analysis

, 16:45 | Cite as

Adjusting for time-varying confounding in the subdistribution analysis of a competing risk

  • Maarten Bekaert
  • Stijn VansteelandtEmail author
  • Karl Mertens


Despite decades of research in the medical literature, assessment of the attributable mortality due to nosocomial infections in the intensive care unit (ICU) remains controversial, with different studies describing effect estimates ranging from being neutral to extremely risk increasing. Interpretation of study results is further hindered by inappropriate adjustment (a) for censoring of the survival time by discharge from the ICU, and (b) for time-dependent confounders on the causal path from infection to mortality. In previous work (Vansteelandt et al. Biostatistics 10:46–59), we have accommodated this through inverse probability of treatment and censoring weighting. Because censoring due to discharge from the ICU is so intimately connected with a patient’s health condition, the ensuing inverse weighting analyses suffer from influential weights and rely heavily on the assumption that one has measured all common risk factors of ICU discharge and mortality. In this paper, we consider ICU discharge as a competing risk in the sense that we aim to infer the risk of ‘ICU mortality’ over time that would be observed if nosocomial infections could be prevented for the entire study population. For this purpose we develop marginal structural subdistribution hazard models with accompanying estimation methods. In contrast to subdistribution hazard models with time-varying covariates, the proposed approach (a) can accommodate high-dimensional confounders, (b) avoids regression adjustment for post-infection measurements and thereby so-called collider-stratification bias, and (c) results in a well-defined model for the cumulative incidence function. The methods are used to quantify the causal effect of nosocomial pneumonia on ICU mortality using data from the National Surveillance Study of Nosocomial Infections in ICU’s (Belgium).


Causal inference Competing risk ICU Inverse weighting Nosocomial infection Time-dependent confounding 


  1. Andersen P, Borgan O, Gill R, Keiding N (1993) Statistical models based on counting processes, Springer series in statistics New York. Springer, NYGoogle Scholar
  2. Barnett A, Graves N (2008) Competing risks models and time-dependent covariates. Crit Care 12: 143CrossRefGoogle Scholar
  3. Bercault N, Boulain T (2001) Mortality rate attributable to ventilator-associated nosocomial pneumonia in an adult intensive care unit: a prospective case-control study. Crit Care 29: 2303–2309CrossRefGoogle Scholar
  4. Beyersmann J, Schumacher M (2008) Time-dependent covariates in the proportional subdistribution hazards model for competing risks. Biostatistics 9: 765–776CrossRefGoogle Scholar
  5. Beyersmann J, Gastmeier P, Grundmann H, Bärwolff S, Geffers C, Behnke M, Rüden H, Schumacher M (2006) Use of multistate models to assess prolongation of intensive care unit stay due to nosocomial infection. Inf Control Hosp Epidemiol 27: 493–499CrossRefGoogle Scholar
  6. Beyersmann J, Wolkewitz M, Schumacher M (2008) The impact of time-dependent bias in proportional hazards modelling. Stat Med 27: 6439–6454CrossRefGoogle Scholar
  7. Chamberlain G (1987) Asymptotic efficiency in estimation with conditional moment restrictions. J Econom 34: 305–334zbMATHCrossRefMathSciNetGoogle Scholar
  8. Davis K (2006) Ventilator-associated pneumonia: a review. J Intensive Care Med 21: 211–226CrossRefGoogle Scholar
  9. Fagon J, Chastre J, Hance A, Montravers P, Novara A, Gibert C (1993) Nosocomial pneumonia in ventilated patient: a cohort study evaluating attributable mortality and hospital stay. Am J Med 94: 281–299CrossRefGoogle Scholar
  10. Fine J, Gray R (1999) A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 94: 496–509zbMATHCrossRefMathSciNetGoogle Scholar
  11. Girou E, Stephan F, Novara A, Safar M, Fagon J (1998) Risk factors and outcome of nosocomial infections: results of a matched case-control study of ICU patients. Am J Respir Crit Care Med 157: 1151–1158Google Scholar
  12. Goetgeluk S, Vansteelandt S, Goetghebeur E (2008) Estimation of controlled direct effects. J R Stat Soc Ser B Methodol 70: 1049–1066CrossRefGoogle Scholar
  13. Greenland S (2003) Quantifying biases in causal models: classical confounding vs collider-stratification bias. Epidemiology 14: 300–306CrossRefGoogle Scholar
  14. Hernán M, Robins J (2006) Estimating causal effects from epidemiological data. J Epidemiol Commun Health 60: 578–586CrossRefGoogle Scholar
  15. Hernán M, Brumback B, Robins J (2000) Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology 11: 561–570CrossRefGoogle Scholar
  16. Hernán M, Diaz SH, Robins J (2004) A structural approach to selection bias. Epidemiology 15: 615–625CrossRefGoogle Scholar
  17. Heyland D, Cook D, Griffith L (1999) The attributable morbidity and mortality of ventilator-associated pneumonia in the critically ill patient. Am J Respir Crit Care Med 159: 1249–1256Google Scholar
  18. Kalbfleisch J, Prentice R (2002) The statistical analysis of failure time data, 2nd edn. John Wiley & Sons, HobokenzbMATHGoogle Scholar
  19. Keiding N, Filiberti M, Esbjerg S, Robins J, Jacobsen N (1999) The graft versus leukemia effect after bone marrow transplantation: a case study using structural nested failure time models. Biometrics 55: 23–28zbMATHCrossRefGoogle Scholar
  20. Kollef M (2003) The importance of appropriate initial antibiotic therapy for hospital-acquired infections. Am J Med 115: 582–584CrossRefGoogle Scholar
  21. Latouche A, Porcher R, Chevret S (2005) A note on including time-dependent covariate in regression model for competing risks data. Biometric J 47: 807–814CrossRefMathSciNetGoogle Scholar
  22. Papazian L, Bregeon F, Thirion X (1996) Effect of ventilator-associated pneumonia on mortality and morbidity. Am J Respir Crit Care Med 154: 91–97Google Scholar
  23. Pepe M, Mori M (1993) Kaplan–Meier, marginal or conditional probability curves in summarizing competing risks failure time data?.  Stat Med 12: 737–751CrossRefGoogle Scholar
  24. Prentice R, Kalbfleisch J, Peterson A, Flournoy N, Farewell V, Breslow N (1978) The analysis of failure times in the presence of competing risks. Biometrics 34: 541–554zbMATHCrossRefGoogle Scholar
  25. Resche-Rigon M, Azoulay E, Chevret S (2006) Evaluating mortality in intensive care units: contribution of competing risks analyses. Crit Care 10: R5CrossRefGoogle Scholar
  26. Robins J (1986) A new approach to causal inference in mortality studies with sustained exposure periods–application to control of the healthy worker survivor effect. Math Model 7: 1393–1512zbMATHCrossRefMathSciNetGoogle Scholar
  27. Robins J, Rotnitzky A (1992) Recovery of information and adjustment for dependent censoring using surrogate markers. AIDS Epidemiol Methodol Issue 29: 297–331Google Scholar
  28. Robins J, Hernán M, Brumback B (2000) Marginal structural models and causal inference in epidemiology. Epidemiology 11: 550–560CrossRefGoogle Scholar
  29. Rubin D (1978) Bayesian inference for causal effects: the role of randomization. Ann Stat 6: 34–58zbMATHCrossRefGoogle Scholar
  30. Safdar N, Dezfulian C, Collard H, Saint S (2005) Clinical and economic consequences of ventilator-associated pneumonia: a systematic review. Crit Care Med 33: 2184–2193CrossRefGoogle Scholar
  31. Schulgen G, Schumacher M (1996) Estimation of prolongation of hospital stay attributable to nosocomial infections: new approaches based on multistate models. Lifetime Data Anal 2: 219–240zbMATHCrossRefGoogle Scholar
  32. Schumacher M, Wangler M, Wolkewitz M, Beyersmann J (2007) Attributable mortality due to nosocomial infections: a simple and useful application of multistate models. Method Inf Med 46: 595–600Google Scholar
  33. Suetens C, Jans B, Carsauw H (1999) Nosocomial infections in intensive care: results from the belgian national surveillance—1996 to 1998. Arch Public Health 57: 221–231Google Scholar
  34. Suetens C, Savey A, Lepape A, Morales I, Carlet J, Fabry J (2003) Surveillance of nosocomial infections in intensive care units: towards a consensual approach in europe. Réanimation 12: 205–213CrossRefGoogle Scholar
  35. Timsit J, Chevret S, Valcke J (1996) Mortality of nosocomial pneumonia in ventilated patients: influence of diagnostic tools. Am J Respir Crit Care Med 154: 116–123Google Scholar
  36. Vallés J, Pobo A, García-Esquirol O, Mariscal D, Real J, Fernández R (2007) Excess ICU mortality attributable to ventilator-associated pneumonia: the role of early vs late onset. Int Care Med 33: 1363–1368CrossRefGoogle Scholar
  37. van der Laan M, Robins J (2003) Unified methods for censored longitudinal data and causality. Springer, New YorkzbMATHGoogle Scholar
  38. VanderWeele T, Vansteelandt S (2009) Conceptual issues concerning mediation, interventions and composition. Stat Interface (in press)Google Scholar
  39. Van Walraven C, Davis D, Forster A, Wells G (2004) Time-dependent bias was common in survival analyses publisched in leading clinical journals. J Clin Epidemiol 57: 672–682CrossRefGoogle Scholar
  40. Vansteelandt S, Mertens K, Suetens C, Goetghebeur E (2009) Marginal structural models for partial exposure regimes. Biostatistics 10: 46–59CrossRefGoogle Scholar
  41. Vrijens F, Hulstaert F, Gordts B, De Laet C, Devriese S, Van de Sande S, Huybrechts M, Peeters G (2009) Infecties in België, deel II: impact op Mortaliteit en Kosten. Health Services Research (HSR). Tech rep Federaal Kenniscentrum voor de Gezondheidszorg (KCE), BrusselGoogle Scholar
  42. Wolkewitz M, Vonberg R, Grundmann H, Beyersmann J, Gastmeier P, Barwolff S, Geffers C, Behnke M, Ruden H, Schumacher M (2008) Risk factors for the development of nosocomial pneumonia and mortality on intensive care units: application of competing risks models. Crit Care R44Google Scholar
  43. Wolkewitz M, Beyersmann J, Gastmeier P, Schumacher M (2009) Modeling the effect of time-dependent exposure on intensive care unit mortality. Int Care Med 35: 826–832CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  • Maarten Bekaert
    • 1
  • Stijn Vansteelandt
    • 1
    Email author
  • Karl Mertens
    • 2
  1. 1.Department of Applied Mathematics and Computer ScienceGhent UniversityGhentBelgium
  2. 2.Epidemiology UnitScientific Institute of Public HealthBrusselsBelgium

Personalised recommendations