Skip to main content

Collinearity and multivariable analysis: response to comments by Claret et al.

Dear Editor,

We are grateful to Claret et al. [1] for their interest in our article [2]. The collinearity between variables was checked by inspection of the correlation between them, by looking at the correlation matrix of the estimated parameters, and by looking at the change in parameter estimates and their estimated standard errors. None of the correlations exceeded 0.24 and no instability of parameter estimates or significant increase in standard errors was observed during the analyses, so there was no indication of collinearity. After removing the age component from the conditional model, the odds ratio was 0.962 (0.955–0.968) versus 0.960 (0.953–0.966) for the intermediate time-to-death group and 0.958 (0.946–0.969) versus 0.957 (0.945–0.968) for the late time-to-death group. Accordingly, there are no concerns about the statistical aspects of our retained model. The coefficient estimates are stable, and the conclusions drawn are appropriate.

References

  1. Claret PG, Bobbia X, de La Coussaye JE (2016) Collinearity and multivariable analysis. Intensive Care Med. doi:10.1007/s00134-016-4528-8

    PubMed  Google Scholar 

  2. Martin-Loeches I, Wunderink RG, Nanchal R, Lefrant JY, Kapadia F, Sakr Y, Vincent JL, on behalf of the ICON Investigators (2016) Determinants of time to death in hospital in critically ill patients around the world. Intensive Care Med 42(9):1454–1460. doi:10.1007/s00134-016-4479-0

    Article  PubMed  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Jean-Louis Vincent.

Ethics declarations

Conflicts of interest

None.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Martin-Loeches, I., Njimi, H. & Vincent, JL. Collinearity and multivariable analysis: response to comments by Claret et al.. Intensive Care Med 42, 1835 (2016). https://doi.org/10.1007/s00134-016-4529-7

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00134-016-4529-7

Keywords

  • Public Health
  • Standard Error
  • Parameter Estimate
  • Emergency Medicine
  • Correlation Matrix