Skip to main content


A new approach to dealing with the analysis of cutpoints in time to event analyses, designed to correct for the over-inflation both of significance levels and of survival differences

Article metrics

  • 516 Accesses


Continuous variables, such as biomarkers, are often analysed by comparing survival or other event times by dividing the data above and below a particular cutpoint. The cutpoints are often defined pragmatically, using medians, quartiles or values chosen from other reported analyses. Choosing the most significant cutpoint leads to an over inflation of significance levels and effect sizes. A more statistically sound approach is to consider transformations of these variables. However, clinicians often wish to use different treatments based on the values of such biomarkers, and therefore need appropriate cutpoints on which to base these choices.


To develop more rigourous and statistically sound methods for the choice and analysis of cutpoints, as described.


Simulations were developed to evaluate whether the effect of the biomarker on the event time was really a threshold effect, which would be optimally described by a cutpoint. Further simulations were used to adjust the significance level to allow for optimising the cutpoint. A method was developed for then adjusting the resulting survival curves comparing values above and below the cutpoint for the fact that the cutpoint had been optimised.


Examples will be given from actual clinical data/trials to demonstrate the application of these methods, with associated issues.


Methods have been developed to deal with the real-life and very common need to choose and analyse cutpoints in clinical trial data, which can potentially avoid studies leading to a plethora of different cutpoint analyses that become very difficult to reconcile and interpret.

Author information

Correspondence to Walter Gregory.

Rights and permissions

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and Permissions

About this article


  • Public Health
  • Clinical Trial
  • Continuous Variable
  • Survival Curve
  • Event Time


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate. Please note that comments may be removed without notice if they are flagged by another user or do not comply with our community guidelines.