A sensitivity analysis for causal parameters in structural proportional hazards models.

Goetghebeur, Els, and Loeys, Tom. "A sensitivity analysis for causal parameters in structural proportional hazards models.." SORT 27.1 (2003): 31-40. <http://eudml.org/doc/40436>.

@article{Goetghebeur2003,
abstract = {Deviations from assigned treatment occur often in clinical trials. In such a setting, the traditional intent-to-treat analysis does not measure biological efficacy but rather programmatic effectiveness. For all-or-nothing compliance situation, Loeys and Goetghebeur (2003) recently proposed a Structural Proportional Hazards method. It allows for casual estimation in the complier subpopulation provided the exclusion restriction holds: randomization per se has no effect unless exposure has changed. The asumption is typically made with structural models for noncompliance but questioned when the trial is not blinded. In this paper we extend the structural PH model to allow for an effect of randomization per se. This enables analyzing sensitivity of conclusions to deviations from the exclusion restriction. In a colo-rectal cancer trial we find the causal estimator for the effect of an arterial device implantation to be remarkably insensitive to such deviations.},
author = {Goetghebeur, Els, Loeys, Tom},
journal = {SORT},
keywords = {Ensayo clínico; Bioestadística; Análisis de supervivencia; Estimación; Inferencia causal; Análisis de datos censurados; Análisis de sensibilidad; causal inference; compliance; exclusion restriction; randomized clinical trials},
language = {eng},
number = {1},
pages = {31-40},
title = {A sensitivity analysis for causal parameters in structural proportional hazards models.},
url = {http://eudml.org/doc/40436},
volume = {27},
year = {2003},
}

TY - JOUR
AU - Goetghebeur, Els
AU - Loeys, Tom
TI - A sensitivity analysis for causal parameters in structural proportional hazards models.
JO - SORT
PY - 2003
VL - 27
IS - 1
SP - 31
EP - 40
AB - Deviations from assigned treatment occur often in clinical trials. In such a setting, the traditional intent-to-treat analysis does not measure biological efficacy but rather programmatic effectiveness. For all-or-nothing compliance situation, Loeys and Goetghebeur (2003) recently proposed a Structural Proportional Hazards method. It allows for casual estimation in the complier subpopulation provided the exclusion restriction holds: randomization per se has no effect unless exposure has changed. The asumption is typically made with structural models for noncompliance but questioned when the trial is not blinded. In this paper we extend the structural PH model to allow for an effect of randomization per se. This enables analyzing sensitivity of conclusions to deviations from the exclusion restriction. In a colo-rectal cancer trial we find the causal estimator for the effect of an arterial device implantation to be remarkably insensitive to such deviations.
LA - eng
KW - Ensayo clínico; Bioestadística; Análisis de supervivencia; Estimación; Inferencia causal; Análisis de datos censurados; Análisis de sensibilidad; causal inference; compliance; exclusion restriction; randomized clinical trials
UR - http://eudml.org/doc/40436
ER -