Assumption-Lean Cox Regression
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
Assumption-Lean Cox Regression. / Vansteelandt, Stijn; Dukes, Oliver; Van Lancker, Kelly; Martinussen, Torben.
In: Journal of the American Statistical Association, Vol. 119, No. 545, 2024, p. 475-484.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Assumption-Lean Cox Regression
AU - Vansteelandt, Stijn
AU - Dukes, Oliver
AU - Van Lancker, Kelly
AU - Martinussen, Torben
PY - 2024
Y1 - 2024
N2 - Inference for the conditional association between an exposure and a time-to-event endpoint, given covariates, is routinely based on partial likelihood estimators for hazard ratios indexing Cox proportional hazards models. This approach is flexible and makes testing straightforward, but is nonetheless not entirely satisfactory. First, there is no good understanding of what it infers when the model is misspecified. Second, it is common to employ variable selection procedures when deciding which model to use. However, the bias and uncertainty that imperfect variable selection adds to the analysis is rarely acknowledged, rendering standard inferences biased and overly optimistic. To remedy this, we propose a nonparametric estimand which reduces to the main exposure effect parameter in a (partially linear) Cox model when that model is correct, but continues to capture the (conditional) association of interest in a well understood way, even when this model is misspecified in an arbitrary manner. We achieve an assumption-lean inference for this estimand based on its influence function under the nonparametric model. This has the further advantage that it makes the proposed approach amenable to the use of data-adaptive procedures (e.g., variable selection, machine learning), which we find to work well in simulation studies and a data analysis. for this article are available online.
AB - Inference for the conditional association between an exposure and a time-to-event endpoint, given covariates, is routinely based on partial likelihood estimators for hazard ratios indexing Cox proportional hazards models. This approach is flexible and makes testing straightforward, but is nonetheless not entirely satisfactory. First, there is no good understanding of what it infers when the model is misspecified. Second, it is common to employ variable selection procedures when deciding which model to use. However, the bias and uncertainty that imperfect variable selection adds to the analysis is rarely acknowledged, rendering standard inferences biased and overly optimistic. To remedy this, we propose a nonparametric estimand which reduces to the main exposure effect parameter in a (partially linear) Cox model when that model is correct, but continues to capture the (conditional) association of interest in a well understood way, even when this model is misspecified in an arbitrary manner. We achieve an assumption-lean inference for this estimand based on its influence function under the nonparametric model. This has the further advantage that it makes the proposed approach amenable to the use of data-adaptive procedures (e.g., variable selection, machine learning), which we find to work well in simulation studies and a data analysis. for this article are available online.
KW - Conditional treatment effect
KW - Debiased machine learning
KW - Estimand
KW - Hazard ratio
KW - Model misspecification
KW - Post-selection inference
KW - CAUSAL INFERENCE
KW - MODELS
KW - HAZARDS
U2 - 10.1080/01621459.2022.2126362
DO - 10.1080/01621459.2022.2126362
M3 - Journal article
VL - 119
SP - 475
EP - 484
JO - Journal of the American Statistical Association
JF - Journal of the American Statistical Association
SN - 0162-1459
IS - 545
ER -
ID: 325327418