Propensity score weighting for covariate adjustment in randomized clinical trials

Zeng, Shuxi; Li, Fan; Wang, Rui

doi:10.1002/sim.8805

Cited by 42 publications

(78 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Equation (3.5) shows that such a covariate adjustment approach in RCT leads to variance reduction for pairwise comparisons, extending the results developed in Zeng et al (2020) to multiple treatments and censored survival outcomes.…”

Section: Draw From the Sample Spacementioning

confidence: 67%

Propensity Score Weighting Analysis of Survival Outcomes Using Pseudo-Observations

Zeng¹,

2024

STAT SINICA

Self Cite

View full text Add to dashboard Cite

Survival outcomes are common in comparative effectiveness studies and require unique handling because they are usually incompletely observed due to right-censoring.A "once for all" approach for causal inference with survival outcomes constructs pseudoobservations and allows standard methods such as propensity score weighting to proceed as if the outcomes are completely observed. For a general class of model-free causal estimands with survival outcomes on user-specified target populations, we develop corresponding propensity score weighting estimators based on the pseudo-observations and establish their asymptotic properties. In particular, utilizing the functional delta-method and the von Mises expansion, we derive a new closed-form variance of the weighting estimator that takes into account the uncertainty due to both pseudo-observation calculation and propensity score estimation. This allows valid and computationally efficient inference without resampling. We also prove the optimal efficiency property of the overlap weights within the class of balancing weights for survival outcomes. The proposed methods are applicable to both binary and multiple treatments. Extensive simulations are conducted to explore the

show abstract

Section: Draw From the Sample Spacementioning

confidence: 67%

Propensity Score Weighting Analysis of Survival Outcomes Using Pseudo-Observations

Zeng¹,

2024

STAT SINICA

Self Cite

View full text Add to dashboard Cite

show abstract

“…This smallest estimation error for each treatment allocation translates into larger efficiency over repeated experiments. Extending the work of Zeng et al 16 to subgroups, we can show that under some mild conditions, the class of propensity score weighting estimators is asymptotically equivalent to the ''ANCOVA-S'' estimatort ANCOVA r for estimating subgroup average treatment effect in RCTs.…”

Section: Propensity Score Weighting In Subgroup Analysismentioning

confidence: 63%

“…Overlap weighting has similar results, but showed improved finite sample performance over inverse probability weighting. 16 By similar arguments, we can show that the propensity score weighting estimator improves the precision of the unadjusted estimator,t UNADJ r , if the propensity score model adjusts for important prognostic covariates. Variance and confidence intervals of thet h r can be obtained either via M-estimation-based sandwich estimator 18,[31][32][33] or bootstrap resampling.…”

Section: Propensity Score Weighting In Subgroup Analysismentioning

confidence: 84%

“…In this case, the propensity score weighting methods preserve the definition of subgroup estimands averaged over pre-treatment covariates within the subgroup and may be preferred, for the same reason that weighting is preferred in estimating the overall average treatment effect in RCTs. 15,16,23…”

Section: Discussionmentioning

confidence: 99%

“…One limitation of ANCOVA is the potential for inviting a ''fishing expedition,'' that is, one may search for a model specification that provides the largest overall and/or subgroup treatment effects, and thus jeopardize the objectivity of statistical analysis. 14,15 Propensity score weighting-an approach that does not require access to the outcome data-is an attractive alternative approach to covariate adjustment; 16,23 it achieves the same goal of improving precision, but offers better transparency by encouraging pre-specification and diagnostic assessment of covariate balance.…”

Section: Propensity Score Weighting In Subgroup Analysismentioning

confidence: 99%

See 2 more Smart Citations

Covariate adjustment in subgroup analyses of randomized clinical trials: A propensity score approach

Yang

Thomas

2021

Clinical Trials

Self Cite

View full text Add to dashboard Cite

Background: Subgroup analyses are frequently conducted in randomized clinical trials to assess evidence of heterogeneous treatment effect across patient subpopulations. Although randomization balances covariates within subgroups in expectation, chance imbalance may be amplified in small subgroups and adversely impact the precision of subgroup analyses. Covariate adjustment in overall analysis of randomized clinical trial is often conducted, via either analysis of covariance or propensity score weighting, but covariate adjustment for subgroup analysis has been rarely discussed. In this article, we develop propensity score weighting methodology for covariate adjustment to improve the precision and power of subgroup analyses in randomized clinical trials. Methods: We extend the propensity score weighting methodology to subgroup analyses by fitting a logistic regression propensity model with pre-specified covariate–subgroup interactions. We show that, by construction, overlap weighting exactly balances the covariates with interaction terms in each subgroup. Extensive simulations were performed to compare the operating characteristics of unadjusted estimator, different propensity score weighting estimators and the analysis of covariance estimator. We apply these methods to the Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training trial to evaluate the effect of exercise training on 6-min walk test in several pre-specified subgroups. Results: Standard errors of the adjusted estimators are smaller than those of the unadjusted estimator. The propensity score weighting estimator is as efficient as analysis of covariance, and is often more efficient when subgroup sample size is small (e.g. <125), and/or when outcome model is misspecified. The weighting estimators with full-interaction propensity model consistently outperform the standard main-effect propensity model. Conclusion: Propensity score weighting is a transparent and objective method to adjust chance imbalance of important covariates in subgroup analyses of randomized clinical trials. It is crucial to include the full covariate–subgroup interactions in the propensity score model.

show abstract

Dual vs Single Cardioversion of Atrial Fibrillation in Patients With Obesity

Aymond,

Sanchez,

Castine

et al. 2024

JAMA Cardiol

View full text Add to dashboard Cite

ImportanceAtrial fibrillation and obesity are common, and both are increasing in prevalence. Obesity is associated with failure of cardioversion of atrial fibrillation using a standard single set of defibrillator pads, even at high output.ObjectiveTo compare the efficacy and safety of dual direct-current cardioversion (DCCV) using 2 sets of pads, with each pair simultaneously delivering 200 J, with traditional single 200-J DCCV using 1 set of pads in patients with obesity and atrial fibrillation.Design, Setting, and ParticipantsThis was a prospective, investigator-initiated, patient-blinded, randomized clinical trial spanning 3 years from August 2020 to 2023. As a multicenter trial, the setting included 3 sites in Louisiana. Eligibility criteria included body mass index (BMI) of 35 or higher (calculated as weight in kilograms divided by height in meters squared), age 18 years or older, and planned nonemergent electrical cardioversion for atrial fibrillation. Patients who met inclusion criteria were randomized 1:1. Exclusions occurred due to spontaneous cardioversion, instability, thrombus, or BMI below threshold.InterventionsDual DCCV vs single DCCV.Main Outcomes and MeasuresReturn to sinus rhythm, regardless of duration, immediately after the first cardioversion attempt of atrial fibrillation, adverse cardiovascular events, and chest discomfort after the procedure.ResultsOf 2079 sequential patients undergoing cardioversion, 276 met inclusion criteria and were approached for participation. Of these, 210 participants were randomized 1:1. After exclusions, 200 patients (median [IQR] age, 67.6 [60.1-72.4] years; 127 male [63.5%]) completed the study. The mean (SD) BMI was 41.2 (6.5). Cardioversion was successful more often with dual DCCV compared with single DCCV (97 of 99 patients [98%] vs 87 of 101 patients [86%]; P = .002). Dual cardioversion predicted success (odds ratio, 6.7; 95% CI, 3.3-13.6; P = .01). Patients in the single cardioversion cohort whose first attempt failed underwent dual cardioversion with all subsequent attempts (up to 3 total), all of which were successful: 12 of 14 after second cardioversion and 2 of 14 after third cardioversion. There was no difference in the rating of postprocedure chest discomfort (median in both groups = 0 of 10; P = .40). There were no cardiovascular complications.Conclusions and RelevanceIn patients with obesity (BMI ≥35) undergoing electrical cardioversion for atrial fibrillation, dual DCCV results in greater cardioversion success compared with single DCCV, without any increase in complications or patient discomfort.Trial RegistrationClinicalTrials.gov Identifier: NCT04539158

show abstract

Propensity score weighting for covariate adjustment in randomized clinical trials

Cited by 42 publications

References 55 publications

Propensity Score Weighting Analysis of Survival Outcomes Using Pseudo-Observations

Propensity Score Weighting Analysis of Survival Outcomes Using Pseudo-Observations

Covariate adjustment in subgroup analyses of randomized clinical trials: A propensity score approach

Dual vs Single Cardioversion of Atrial Fibrillation in Patients With Obesity

Contact Info

Product

Resources

About