Model Selection for G-Estimation of Dynamic Treatment Regimes

Wallace, Michael P.; Moodie, Erica E. M.; Stephens, David A.

doi:10.1111/biom.13104

Cited by 14 publications

(25 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our analysis follows (Wallace et al, 2019) As in (Wallace et al, 2019), the treatment-free models are linear in (q 1 , s 1 , p 1 ) at stage 1 and (a 1 , q 2 , s 2 , p 2 ) at stage 2. Linear blip models with covariates (q 1 , s 1 , p 1 ) at stage 1, and…”

Section: Simulations Evaluating Performance In Multi-stage Settingmentioning

confidence: 99%

See 1 more Smart Citation

Variable Selection in Regression-based Estimation of Dynamic Treatment Regimes

Bian

Moodie

Shortreed

et al. 2021

Preprint

View full text Add to dashboard Cite

Dynamic treatment regimes (DTRs) consist of a sequence of decision rules, one per stage of intervention, that finds effective treatments for individual patients according to patient information history. DTRs can be estimated from models which include the interaction between treatment and a small number of covariates which are often chosen a priori. However, with increasingly large and complex data being collected, it is difficult to know which prognostic factors might be relevant in the treatment rule. Therefore, a more data-driven approach of selecting these covariates might improve the estimated decision rules and simplify models to make them easier to interpret. We propose a variable selection method for DTR estimation using penalized dynamic weighted least squares. Our method has the strong heredity property, that is, an interaction term can be included in the model only if the corresponding main terms have also been selected. Through simulations, we show our method has both the double robustness property and the oracle property, and the newly proposed methods compare favorably with other variable selection approaches.

show abstract

Section: Simulations Evaluating Performance In Multi-stage Settingmentioning

confidence: 99%

“…(a 1 , q 2 , s 2 , p 2 ) at stage 2 were considered. Noted in (Wallace et al, 2019), a 1 and p 2 were not included in the blip models to avoid the multicollinearity; this is not necessary in pdWOLS and hence our model specifications differ.…”

Section: Simulations Evaluating Performance In Multi-stage Settingmentioning

confidence: 99%

Variable Selection in Regression-based Estimation of Dynamic Treatment Regimes

Bian

Moodie

Shortreed

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…For example, a Cox model could be assumed for a survival outcome. Except for a few instances (e.g., [3,21]), other outcome types such as counts or binary outcomes have rarely been considered. We are unaware of any analyses or methods that have aimed to optimize a binary outcome over more than two stages of intervention.…”

Section: Further Extensionsmentioning

confidence: 99%

Precision medicine: Statistical methods for estimating adaptive treatment strategies

Moodie

Krakow

2020

Bone Marrow Transplant

Self Cite

View full text Add to dashboard Cite

The beauty of science is that all the important things are unpredictable. Freeman Dyson In the typescript which follows, Moodie and Krakow tackle the topical issue of precision medicine and statistical methods for estimating adaptive treatment strategies. This may be the most difficult typescript in our series so far for non-statisticians to understand. It even has equations! But please bear with the authors and give it a chance. One needs not to understand the equations to get the thrust of the strategy. Precision medicine as we discuss elsewhere, is misnamed. In statistics and mathematics precision refers to getting the same answer again and again. It does not mean getting the correct answer, the term for which is accuracy, not precision. However, precision is the current buzz word so there's no point trying to get this straight. When we think about precision we need to consider two elements, reproducibility and replicability. Reproducibility means you give me your data and computer code and I come to the same conclusion you did. Replicability is another matter. I try to replicate your experiment and hopefully reach the same conclusion. In medicine, replicability is obviously more important than reproducibility but things which cannot be reproduced are unlikely to be replicated. As the authors discuss, one can think about precision medicine as one does a family vacation. A best vacation depends on several covariates: where you live, your prior travel experiences, advice from family and friends, online reviews, Wikitravel, cost, your travel budget, if you have kids and many other co-variates. Consequently, there is unlikely to be a best vacation for everyone. Yours might be a week at the Ritz Carlton Cancun with dinner at Careyes and ours, a week at the Pfister Hotel in Milwaukee with dinner at Mader's German Restaurant (bring simvastatin). Similarly, it is unlikely there is a best therapy of acute myeloid leukemia, a best donor, a best conditioning regimen, a best posttransplant immune suppressive regimen etc. and certainly no best combination of these co-variates for your patient. The question Moodie and Krakow tackle is how we can determine the best therapy or combination of therapies for someone receiving a haematopoietic cell transplant. Although the default answer is typically: randomized clinical trials are the gold standard, these inform us of the outcome of a cohort of subjects, not individuals. In many instances, although a new therapy may be shown to be better than an old one in a controlled randomized trial the benefit is not uniformly distributed. Some subjects in the experimental cohort may do worse with the new therapy compared with controls, others better. The question is who are the winners and losers? We cannot do a controlled randomized trial of one person. Moodie and Krakow discuss statistical tools to help us sort this out. Again, please do not be put off by the equations; forgetaboutit. The overriding message is not so complex, and important. We are always standing by on twitter @BMTStats t...

show abstract

“…Although most methodologies target continuous outcomes, there are some existing studies regarding non-continuous outcomes. Some theoretical developments in the DTR literature have focused on discrete-outcome settings, including Qlearning with discrete outcomes (Bernoulli and Poisson) utilities ([Moodie et al, 2014]), an extension of G-estimation to the case of non-additive treatment effects for discrete outcomes ([Wallace et al, 2019]), and the extension of dWOLS to time-to-event data with survival outcomes subject to right-censoring ( [Simoneau et al, 2020]).…”

Section: Introductionmentioning

confidence: 99%

Doubly-Robust Dynamic Treatment Regimen Estimation for Binary Outcomes

Jiang¹,

Wallace²,

Thompson³

2022

Preprint

View full text Add to dashboard Cite

In precision medicine, Dynamic Treatment Regimes (DTRs) are treatment protocols that adapt over time in response to a patient's observed characteristics. A DTR is a set of decision functions that takes an individual patient's information as arguments and outputs an action to be taken. Building on observed data, the aim is to identify the DTR that optimizes expected patient outcomes. Multiple methods have been proposed for optimal DTR estimation with continuous outcomes. However, optimal DTR estimation with binary outcomes is more complicated and has received comparatively little attention. Solving a system of weighted generalized estimating equations, we propose a new balancing weight criterion to overcome the misspecification of generalized linear models' nuisance components. We construct binary pseudo-outcomes, and develop a doubly-robust and easy-to-use method to estimate an optimal DTR with binary outcomes. We also outline the underlying theory, which relies on the balancing property of the weights; provide simulation studies that verify the double-robustness of our method; and illustrate the method in studying the effects of e-cigarette usage on smoking cessation, using observational data from the Population Assessment of Tobacco and Health (PATH) study.

show abstract

Model Selection for G-Estimation of Dynamic Treatment Regimes

Cited by 14 publications

References 28 publications

Variable Selection in Regression-based Estimation of Dynamic Treatment Regimes

Variable Selection in Regression-based Estimation of Dynamic Treatment Regimes

Precision medicine: Statistical methods for estimating adaptive treatment strategies

Doubly-Robust Dynamic Treatment Regimen Estimation for Binary Outcomes

Contact Info

Product

Resources

About