Dataset Information

Lasso adjustments of treatment effect estimates in randomized experiments.

ABSTRACT: We provide a principled way for investigators to analyze randomized experiments when the number of covariates is large. Investigators often use linear multivariate regression to analyze randomized experiments instead of simply reporting the difference of means between treatment and control groups. Their aim is to reduce the variance of the estimated treatment effect by adjusting for covariates. If there are a large number of covariates relative to the number of observations, regression may perform poorly because of overfitting. In such cases, the least absolute shrinkage and selection operator (Lasso) may be helpful. We study the resulting Lasso-based treatment effect estimator under the Neyman-Rubin model of randomized experiments. We present theoretical conditions that guarantee that the estimator is more efficient than the simple difference-of-means estimator, and we provide a conservative estimator of the asymptotic variance, which can yield tighter confidence intervals than the difference-of-means estimator. Simulation and data examples show that Lasso-based adjustment can be advantageous even when the number of covariates is less than the number of observations. Specifically, a variant using Lasso for selection and ordinary least squares (OLS) for estimation performs particularly well, and it chooses a smoothing parameter based on combined performance of Lasso and OLS.

SUBMITTER: Bloniarz A

PROVIDER: S-EPMC4941428 | biostudies-literature | 2016 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Lasso adjustments of treatment effect estimates in randomized experiments.

Bloniarz Adam A Liu Hanzhong H Zhang Cun-Hui CH Sekhon Jasjeet S JS Yu Bin B

Proceedings of the National Academy of Sciences of the United States of America 20160701 27

We provide a principled way for investigators to analyze randomized experiments when the number of covariates is large. Investigators often use linear multivariate regression to analyze randomized experiments instead of simply reporting the difference of means between treatment and control groups. Their aim is to reduce the variance of the estimated treatment effect by adjusting for covariates. If there are a large number of covariates relative to the number of observations, regression may perfo ...[more]

PMID: 27382153

Similar Datasets

Project description:BackgroundPatient-centred care requires evidence of treatment effects across many outcomes. Outcomes can be beneficial (e.g. increased survival or cure rates) or detrimental (e.g. adverse events, pain associated with treatment, treatment costs, time required for treatment). Treatment effects may also be heterogeneous across outcomes and across patients. Randomized controlled trials are usually insufficient to supply evidence across outcomes. Observational data analysis is an alternative, with the caveat that the treatments observed are choices. Real-world treatment choice often involves complex assessment of expected effects across the array of outcomes. Failure to account for this complexity when interpreting treatment effect estimates could lead to clinical and policy mistakes.ObjectiveOur objective was to assess the properties of treatment effect estimates based on choice when treatments have heterogeneous effects on both beneficial and detrimental outcomes across patients.MethodsSimulation methods were used to highlight the sensitivity of treatment effect estimates to the distributions of treatment effects across patients across outcomes. Scenarios with alternative correlations between benefit and detriment treatment effects across patients were used. Regression and instrumental variable estimators were applied to the simulated data for both outcomes.ResultsTrue treatment effect parameters are sensitive to the relationships of treatment effectiveness across outcomes in each study population. In each simulation scenario, treatment effect estimate interpretations for each outcome are aligned with results shown previously in single outcome models, but these estimates vary across simulated populations with the correlations of treatment effects across patients across outcomes.ConclusionsIf estimator assumptions are valid, estimates across outcomes can be used to assess the optimality of treatment rates in a study population. However, because true treatment effect parameters are sensitive to correlations of treatment effects across outcomes, decision makers should be cautious about generalizing estimates to other populations.

Project description:ObjectiveTo assess the influence of trial sample size on treatment effect estimates within meta-analyses.DesignMeta-epidemiological study.Data sources93 meta-analyses (735 randomised controlled trials) assessing therapeutic interventions with binary outcomes, published in the 10 leading journals of each medical subject category of the Journal Citation Reports or in the Cochrane Database of Systematic Reviews.Data extractionSample size, outcome data, and risk of bias extracted from each trial.Data synthesisTrials within each meta-analysis were sorted by their sample size: using quarters within each meta-analysis (from quarter 1 with 25% of the smallest trials, to quarter 4 with 25% of the largest trials), and using size groups across meta-analyses (ranging from <50 to ≥ 1000 patients). Treatment effects were compared within each meta-analysis between quarters or between size groups by average ratios of odds ratios (where a ratio of odds ratios less than 1 indicates larger effects in smaller trials).ResultsTreatment effect estimates were significantly larger in smaller trials, regardless of sample size. Compared with quarter 4 (which included the largest trials), treatment effects were, on average, 32% larger in trials in quarter 1 (which included the smallest trials; ratio of odds ratios 0.68, 95% confidence interval 0.57 to 0.82), 17% larger in trials in quarter 2 (0.83, 0.75 to 0.91), and 12% larger in trials in quarter 3 (0.88, 0.82 to 0.95). Similar results were obtained when comparing treatment effect estimates between different size groups. Compared with trials of 1000 patients or more, treatment effects were, on average, 48% larger in trials with fewer than 50 patients (0.52, 0.41 to 0.66) and 10% larger in trials with 500-999 patients (0.90, 0.82 to 1.00).ConclusionsTreatment effect estimates differed within meta-analyses solely based on trial sample size, with stronger effect estimates seen in small to moderately sized trials than in the largest trials.

Dataset Information

Lasso adjustments of treatment effect estimates in randomized experiments.

Publications

Lasso adjustments of treatment effect estimates in randomized experiments.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets