Evaluating the use of bootstrapping in cohort studies conducted with 1:1 propensity score matching—A plasmode simulation study

2019 ◽  
Vol 28 (6) ◽  
pp. 879-886 ◽  
Author(s):  
Rishi J. Desai ◽  
Richard Wyss ◽  
Younathan Abdia ◽  
Sengwee Toh ◽  
Margaret Johnson ◽  
...  
Keyword(s):  
Propensity Score ◽  
Cohort Studies ◽  
Propensity Score Matching ◽  
Simulation Study

scholarly journals Comparative Effectiveness of Supplementation with a Juice Powder Concentrate: The Challenge of Multiplicity and Propensity Score Matching (P06-073-19)

2019 ◽  
Vol 3 (Supplement_1) ◽  
Author(s):  
Dario Gregori ◽  
Ileana Baldi ◽  
Giulia Lorenzoni ◽  
Melina Tsiountsioura ◽  
Manfred Lamprecht
Keyword(s):  
Propensity Score ◽  
Vitamin C ◽  
Cohort Studies ◽  
Propensity Score Matching ◽  
Comparative Effectiveness ◽  
Specific Effect ◽  
Operating Characteristics ◽  
Primary Endpoints ◽  
Tnf Α ◽  
Multiple Primary

Abstract Objectives There is a need for real-world evaluation of dietary supplementation in complex scenarios. We have designed an observational study intended to assess the comparative effectiveness of supplementation with a juice powder concentrate foreseeing multiple exposure groups, multiple primary endpoints and prospective use of propensity score matching. Methods Three different exposure groups (Control, “Capsules” and “Shake”) and three primary endpoints have been identified (TNF-α, Homocysteine and Vitamin C). This study design presents a multiplicity problem with two sources of multiplicity. The two supplementation-control comparisons (Capsules vs Control and Shake vs Control) and the between-supplementation comparison (Capsules vs Shake) will serve as the first source of multiplicity. In addition to that, three multiple primary endpoints are tested in each comparison. A formal power analysis is carried out through simulations assuming: i) specific effect sizes per comparison per endpoint; ii) t-test with parallel gatekeeping with truncated Holm test as the multiplicity adjustment procedure; iii) control of the familywise error rate (FWER) at a pre-specified 0.05 level. For a proper comparative analysis of effectiveness, we foresee the use of matching on a propensity score (PS) prospectively. PSs are typically applied in retrospective cohort studies. However we initially developed a Random Forest-based PS model on historical data. The PSs predicted by this tool are used to match patients on an ongoing basis to evaluate the comparative effectiveness of supplementation. Results With a sample size of 20 subjects per exposure group, the disjunctive power for testing each primary endpoint (i.e., the probability of establishing a significant effect in Capsules vs. Control or Shake vs. Control or Capsules vs. Shake) is 78%, 51%, 97% for TNF-α, Homocysteine and Vitamin C, respectively. Conclusions The increasing availability of computational resources and methods allows researchers to conduct comparative effectiveness cohort studies that require prospective data collection, by transporting Pss to new patients. Moreover, simulations allow examining the operating characteristics of complex testing frameworks. Funding Sources ZETA RESEARCH S.r.l.

scholarly journals How to Apply Multiple Imputation in Propensity Score Matching with Partially Observed Confounders: A Simulation Study and Practical Recommendations

2021 ◽  
Vol 19 (1) ◽  
Author(s):  
Albee Ling ◽  
Maria Montez-Rath ◽  
Maya Mathur ◽  
Kris Kapphahn ◽  
Manisha Desai
Keyword(s):  
Missing Data ◽  
Propensity Score ◽  
Multiple Imputation ◽  
Propensity Score Matching ◽  
Simulation Study ◽  
Observational Studies ◽  
Auxiliary Variables ◽  
Imputation Model ◽  
Partially Observed ◽  
Practical Recommendations

Propensity score matching (PSM) has been widely used to mitigate confounding in observational studies, although complications arise when the covariates used to estimate the PS are only partially observed. Multiple imputation (MI) is a potential solution for handling missing covariates in the estimation of the PS. However, it is not clear how to best apply MI strategies in the context of PSM. We conducted a simulation study to compare the performances of popular non-MI missing data methods and various MI-based strategies under different missing data mechanisms. We found that commonly applied missing data methods resulted in biased and inefficient estimates, and we observed large variation in performance across MI-based strategies. Based on our findings, we recommend 1) estimating the PS after applying MI to impute missing confounders; 2) conducting PSM within each imputed dataset followed by averaging the treatment effects to arrive at one summarized finding; 3) a bootstrapped-based variance to account for uncertainty of PS estimation, matching, and imputation; and 4) inclusion of key auxiliary variables in the imputation model.

scholarly journals Greedy caliper propensity score matching can yield variable estimates of the treatment‐outcome association ‐ a simulation study

2021 ◽  
Author(s):  
Joris J. Komen ◽  
Svetlana V. Belitser ◽  
Richard Wyss ◽  
Sebastian Schneeweiss ◽  
Anne C. Taams ◽  
...  
Keyword(s):  
Treatment Outcome ◽  
Propensity Score ◽  
Propensity Score Matching ◽  
Simulation Study

One-to-many propensity score matching in cohort studies

2012 ◽  
Vol 21 ◽  
pp. 69-80 ◽  
Author(s):  
Jeremy A. Rassen ◽  
Abhi A. Shelat ◽  
Jessica Myers ◽  
Robert J. Glynn ◽  
Kenneth J. Rothman ◽  
...  
Keyword(s):  
Propensity Score ◽  
Cohort Studies ◽  
Propensity Score Matching

The Balance Super Learner: A robust adaptation of the Super Learner to improve estimation of the average treatment effect in the treated based on propensity score matching

2016 ◽  
Vol 27 (8) ◽  
pp. 2504-2518 ◽  
Author(s):  
Romain Pirracchio ◽  
Marco Carone
Keyword(s):  
Propensity Score ◽  
Propensity Score Matching ◽  
Simulation Study ◽  
Model Misspecification ◽  
Average Treatment Effect ◽  
Machine Learning Algorithms ◽  
Propensity Score Model ◽  
Covariate Balance ◽  
Ensemble Machine Learning ◽  
Super Learner

Consistency of the propensity score estimators rely on correct specification of the propensity score model. The propensity score is frequently estimated using a main effect logistic regression. It has recently been shown that the use of ensemble machine learning algorithms, such as the Super Learner, could improve covariate balance and reduce bias in a meaningful manner in the case of serious model misspecification for treatment assignment. However, the loss functions normally used by the Super Learner may not be appropriate for propensity score estimation since the goal in this problem is not to optimize propensity score prediction but rather to achieve the best possible balance in the covariate distribution between treatment groups. In a simulation study, we evaluated the benefit of a modification of the Super Learner by propensity score estimation geared toward achieving covariate balance between the treated and untreated after matching on the propensity score. Our simulation study included six different scenarios characterized by various degrees of deviation from the usual main term logistic model for the true propensity score and outcome as well as the presence (or not) of instrumental variables. Our results suggest that the use of this adapted Super Learner to estimate the propensity score can further improve the robustness of propensity score matching estimators.

Sign in / Sign up

Export Citation Format

Share Document