arXiv subjects

Helene C. W. Rytgaard

Publications and source records attributed to Helene C. W. Rytgaard.

11 recordsLinked to original sources

arXiv2026

On the limitations of causal inference with current-treatment Cox models

Cox models with time-varying treatments often include only the current treatment level. Translating such a model into causally meaningful intervention-specific survival probabilities relies on the Markov property: that the hazard is independent of treatment history conditional on current treatment. For the Markov property to not be population-specific, it needs to hold also conditional on any unmeasured prognostic heterogeneity (frailty). Using a discrete-time argument, earlier work concluded that the Markov property can hold both conditionally and marginally only in the absence of a treatment effect or an effect of the unmeasured heterogeneity. We broaden this argument by developing a continuous-time framework that encompasses both proposed extensions of the Kaplan-Meier curve and current-treatment marginal structural Cox models, and sharpen it by characterizing precisely the conditions under which the Markov property can hold both conditionally and marginally. Specifically, we show that it requires the absence of treatment-frailty interaction on the additive hazard scale. As frailty is inherently unmeasured, such a no-interaction assumption cannot be verified. Consequently, causal interpretation of a current-treatment marginal structural Cox model rests on a strong and unverifiable structural assumption. Illustrating this point, we construct a setting in which a Cox model is correctly specified conditionally on current treatment only, although the marginal Markov property fails. Transforming the fitted model to the survival scale does not recover the true intervention-specific survival probabilities. Thus, moving from the hazard scale to the survival scale is not in itself sufficient to obtain a causal interpretation in the time-varying treatment setting.

stat.ME↗

arXiv2026

Estimation of treatment effects in presence of differential use of post-randomization concomitant medication with time-to-event outcomes

In placebo-controlled randomized trials, the post-randomization use of concomitant medications may be higher in the placebo arm than in the treatment arm. This may dilute the full benefits of the randomized drug as estimated by the intention-to-treat analysis. We focus on cardiovascular outcomes trials in type-2 diabetes patients of glucose-lowering treatments where patients in the placebo arm are more likely to add other glucose-lowering agents with established cardio-protective properties. As a supplement to the intention-to-treat analysis, we propose a class of estimands within a causal framework that isolates the specific impact of the treatment being studied from that of concomitant treatment use. These estimands are defined under time-dependent treatment interventions to balance exposure to additional medications across intervention arms. We advocate for specific stochastic interventions to achieve this balance while minimizing positivity violations, which arise when certain treatment combinations or characteristics are not sufficiently represented in the data. We employ targeted minimum loss-based estimation (TMLE) to optimize the estimation procedure for our estimands while allowing for flexible adjustments for time-dependent covariates from follow-up visits. Finally, we demonstrate the application of the methods through a simulation study and a real-world example from the LEADER cardiovascular outcomes trial, which assessed cardiovascular risk for liraglutide versus placebo.

stat.ME↗

arXiv2026

Nonparametric efficient estimation of the longitudinal front-door functional

The front-door criterion is an identification strategy for the intervention-specific mean outcome in settings where the standard back-door criterion fails due to unmeasured exposure-outcome confounders, but an intermediate variable exists that completely mediates the effect of exposure on the outcome and is not affected by unmeasured confounding. The front-door criterion has been extended to the longitudinal setting, where exposure and mediator vary over time. However, with the exception of a simple plug-in estimator, no suitable estimation techniques have been proposed. In this work, we derive nonparametric efficient estimators of the longitudinal front-door functional. The estimators accommodate high-dimensional mediators, are multiply robust, and allow for the use of data-adaptive methods for estimating nuisance functions while still providing valid inference. The theoretical properties of the estimators are illustrated in a simulation study, and we apply the estimators to a trial of peanut allergy in infants.

stat.ME↗

arXiv2025

Improving precision of cumulative incidence estimates in randomized controlled trials with external controls

Augmenting the control arm in clinical trials with external data can improve statistical power for demonstrating treatment effects. In many time-to-event outcome trials, participants are subject to truncation by death. Direct application of methods for competing risks analysis on the joint data may introduce bias, for example, due to covariate shifts between the populations. In this work, we consider transportability of the conditional cause-specific hazard of the event of interest under the control treatment. Under this assumption, we derive semiparametric efficiency bounds of causal cumulative incidences. This allows for quantification of the theoretical efficiency gain from incorporating the external controls. We propose triply robust estimators that can achieve the efficiency bounds, where the trial controls and external controls are made comparable through time-specific weights in a martingale integral. We conducted a simulation study to show the precision gain of the proposed fusion estimators compared to their counterparts without utilizing external controls. As a real data application, we used two cardiovascular outcome trials conducted to assess the safety of glucagon-like peptide-1 agonists. Incorporating the external controls from one trial into the other, we observed a decrease in the standard error of the treatment effects on adverse non-fatal cardiovascular events with all-cause death as the competing risk.

stat.ME↗

arXiv2025

Proximal indirect comparison

We consider the problem of indirect comparison, where a treatment arm of interest is absent by design in one randomized controlled trial but available in the other. The former is the target trial, and the latter is the source trial. The identifiability of the target population average treatment effect often relies on conditional transportability assumptions. However, it is a common concern whether all relevant effect modifiers are measured and controlled for. We give a new proximal identification result in the presence of shifted, unobserved effect modifiers based on proxies: an adjustment proxy in both trials and an additional reweighting proxy in the source trial. We propose an estimator which is doubly-robust against misspecifications of the so-called bridge functions and asymptotically normal under mild consistency of estimators for the bridge functions. We use two weight management trials as a context to illustrate selection of proxies and apply our method to compare the weight loss effect of active treatments from these trials.

stat.ME↗

arXiv2025

Nonparametric efficient causal estimation of the intervention-specific expected number of recurrent events with continuous-time targeted maximum likelihood and highly adaptive lasso estimation

Longitudinal settings involving outcome, competing risks and censoring events occurring and recurring in continuous time are common in medical research, but are often analyzed with methods that do not allow for taking post-baseline information into account. In this work, we define statistical and causal target parameters via the g-computation formula by carrying out interventions directly on the product integral representing the observed data distribution in a continuous-time counting process model framework. In recurrent events settings our target parameter identifies the expected number of recurrent events also in settings where the censoring mechanism or post-baseline treatment decisions depend on past information of post-baseline covariates such as the recurrent event process. We propose a flexible estimation procedure based on targeted maximum likelihood estimation coupled with highly adaptive lasso estimation to provide a novel approach for double robust and nonparametric inference for the considered target parameter. We illustrate the methods in a simulation study.

stat.ME↗

arXiv2025

concrete: Targeted Estimation of Survival and Competing Risks in Continuous Time

This article introduces the R package concrete, which implements a recently developed targeted maximum likelihood estimator (TMLE) for the cause-specific absolute risks of time-to-event outcomes measured in continuous time. Cross-validated Super Learner machine learning ensembles are used to estimate propensity scores and conditional cause-specific hazards, which are then targeted to produce robust and efficient plug-in estimates of the effects of static or dynamic interventions on a binary treatment given at baseline quantified as risk differences or risk ratios. Influence curve-based asymptotic inference is provided for TMLE estimates and simultaneous confidence bands can be computed for target estimands spanning multiple multiple times or events. In this paper we review the one-step continuous-time TMLE methodology as it is situated in an overarching causal inference workflow, describe its implementation, and demonstrate the use of the package on the PBC dataset.

stat.CO↗

arXiv2024

Estimating conditional hazard functions and densities with the highly-adaptive lasso

We consider estimation of conditional hazard functions and densities over the class of multivariate càdlàg functions with uniformly bounded sectional variation norm when data are either fully observed or subject to right-censoring. We demonstrate that the empirical risk minimizer is either not well-defined or not consistent for estimation of conditional hazard functions and densities. Under a smoothness assumption about the data-generating distribution, a highly-adaptive lasso estimator based on a particular data-adaptive sieve achieves the same convergence rate as has been shown to hold for the empirical risk minimizer in settings where the latter is well-defined. We use this result to study a highly-adaptive lasso estimator of a conditional hazard function based on right-censored data. We also propose a new conditional density estimator and derive its convergence rate. Finally, we show that the result is of interest also for settings where the empirical risk minimizer is well-defined, because the highly-adaptive lasso depends on a much smaller number of basis function than the empirical risk minimizer.

math.ST↗

arXiv2023

Nonparametric estimation of the interventional disparity indirect effect among the exposed

In situations with non-manipulable exposures, interventions can be targeted to shift the distribution of intermediate variables between exposure groups to define interventional disparity indirect effects. In this work, we present a theoretical study of identification and nonparametric estimation of the interventional disparity indirect effect among the exposed. The targeted estimand is intended for applications examining the outcome risk among an exposed population for which the risk is expected to be reduced if the distribution of a mediating variable was changed by a (hypothetical) policy or health intervention that targets the exposed population specifically. We derive the nonparametric efficient influence function, study its double robustness properties and present a targeted minimum loss-based estimation (TMLE) procedure. All theoretical results and algorithms are provided for both uncensored and right-censored survival outcomes. With offset in the ongoing discussion of the interpretation of non-manipulable exposures, we discuss relevant interpretations of the estimand under different sets of assumptions of no unmeasured confounding and provide a comparison of our estimand to other related estimands within the framework of interventional (disparity) effects. Small-sample performance and double robustness properties of our estimation procedure are investigated and illustrated in a simulation study.

stat.ME↗

arXiv2021

One-step TMLE for targeting cause-specific absolute risks and survival curves

This paper considers one-step targeted maximum likelihood estimation method for general competing risks and survival analysis settings where event times take place on the positive real line R+ and are subject to right-censoring. Our interest is overall in the effects of baseline treatment decisions, static, dynamic or stochastic, possibly confounded by pre-treatment covariates. We point out two overall contributions of our work. First, our method can be used to obtain simultaneous inference across all absolute risks in competing risks settings. Second, we present a practical result for achieving inference for the full survival curve, or a full absolute risk curve, across time by targeting over a fine enough grid of points. The one-step procedure is based on a one-dimensional universal least favorable submodel for each cause-specific hazard that can be implemented in recursive steps along a corresponding universal least favorable submodel. We present a theorem for conditions to achieve weak convergence of the estimator for an infinite-dimensional target parameter. Our empirical study demonstrates the use of the methods.

stat.ME↗

arXiv2021

Ranking of average treatment effects with generalized random forests for time-to-event outcomes

In this paper we present a data-adaptive estimation procedure for estimation of average treatment effects in a time-to-event setting based on generalized random forests. In these kinds of settings, the definition of causal effect parameters are complicated by competing risks; here we distinguish between treatment effects on the crude and the net probabilities, respectively. To handle right-censoring, and to switch between crude and net probabilities, we propose a two-step procedure for estimation, applying inverse probability weighting to construct time-point specific weighted outcomes as input for the forest. The forest adaptively handles confounding of the treatment assigned by applying a splitting rule that targets a causal parameter. We demonstrate that our method is effective for a causal search through a list of treatments to be ranked according to the magnitude of their effect. We further apply our method to a dataset from the Danish health registries where it is of interest to discover drugs with an unexpected protective effect against relapse of severe depression.

stat.ME↗