Hodges-Lehmann Point Estimates of Treatment Effect in Observational Studies

1993 ◽  
Vol 88 (424) ◽  
pp. 1250-1253 ◽  
Author(s):  
Paul R. Rosenbaum
2019 ◽  
Vol 5 (2) ◽  
pp. 21-35
Author(s):  
Carlos Carvalho ◽  
Avi Feller ◽  
Jared Murray ◽  
Spencer Woody ◽  
David Yeager

Biometrika ◽  
2020 ◽  
Author(s):  
Oliver Dukes ◽  
Stijn Vansteelandt

Summary Eliminating the effect of confounding in observational studies typically involves fitting a model for an outcome adjusted for covariates. When, as often, these covariates are high-dimensional, this necessitates the use of sparse estimators, such as the lasso, or other regularization approaches. Naïve use of such estimators yields confidence intervals for the conditional treatment effect parameter that are not uniformly valid. Moreover, as the number of covariates grows with the sample size, correctly specifying a model for the outcome is nontrivial. In this article we deal with both of these concerns simultaneously, obtaining confidence intervals for conditional treatment effects that are uniformly valid, regardless of whether the outcome model is correct. This is done by incorporating an additional model for the treatment selection mechanism. When both models are correctly specified, we can weaken the standard conditions on model sparsity. Our procedure extends to multivariate treatment effect parameters and complex longitudinal settings.


Author(s):  
Ariel Linden ◽  
Maya B. Mathur ◽  
Tyler J. VanderWeele

In this article, we introduce the evalue package, which performs sensitivity analyses for unmeasured confounding in observational studies using the methodology proposed by VanderWeele and Ding (2017, Annals of Internal Medicine 167: 268–274). evalue reports E-values, defined as the minimum strength of association on the risk-ratio scale that an unmeasured confounder would need to have with both the treatment assignment and the outcome to fully explain away a specific treatment-outcome association, conditional on the measured covariates. evalue computes E-values for point estimates (and optionally, confidence limits) for several common outcome types, including risk and rate ratios, odds ratios with common or rare outcomes, hazard ratios with common or rare outcomes, standardized mean differences in outcomes, and risk differences.


Author(s):  
Edo Richard

Observational studies have taught us a lot about the origin of neurological and neuropsychiatric diseases. This chapter describes how we can translate this knowledge into action. Before engaging in a large RCT, several steps have to be taken. First, the potential for a treatment effect has to be compelling. The target population in the RCT has to resemble the population in which observational studies described an association. Second, the outcome of an RCT has to be chosen, and has to have clinical relevance or at least have the potential of clinical relevance in the future. Third, the right study design has to be decided on. Each research question will require a specific study design with accompanying sample size calculation. Lastly, specific ethical considerations have to be taken into account when designing and executing an intervention study. This chapter presents an overview of these issues.


2015 ◽  
Vol 105 (5) ◽  
pp. 476-480 ◽  
Author(s):  
Susan Athey ◽  
Guido Imbens

Researchers often report estimates and standard errors for the object of interest (such as a treatment effect) based on a single specification of a statistical model. We propose a systematic approach to assessing sensitivity to specification. We construct estimates of the object of interest for each of a large set of models. Our proposed robustness measure is the standard deviation of the point estimates over the set of models. Each member of the set is generated by splitting the sample into two subsamples based on covariate values, constructing separate parameter estimates for each subsample, and then combining the results.


2021 ◽  
Vol 80 (Suppl 1) ◽  
pp. 790.1-790
Author(s):  
P. Nash ◽  
P. Richette ◽  
L. Gossec ◽  
A. Marchesoni ◽  
C. T. Ritchlin ◽  
...  

Background:Approximately 40% of PsA patients (pts) on advanced therapy are on monotherapy.1,2 Upadacitinib (UPA) showed efficacy and safety in pts with active PsA in the Phase 3 SELECT-PsA 1 and SELECT-PsA 2 clinical trials.3,4Objectives:Assess efficacy and safety in subgroups of pts treated with UPA as monotherapy or in combination with non-biologic disease-modifying antirheumatic drugs (non-bDMARDs).Methods:The SELECT-PsA program enrolled pts with prior inadequate response (IR) or intolerance to ≥1 non-bDMARD (N=1705) and prior IR or intolerance to ≥1 bDMARD (N=642). Data from both trials was integrated for pts receiving placebo (PBO), UPA 15 mg once daily (QD) and UPA 30 mg QD. Stable background treatment of ≤2 non-bDMARDs was permitted, but not required. Analysis includes UPA monotherapy vs combination therapy for endpoints: ACR20/50/70 responses and change from baseline in pain and HAQ-DI (Wk 12); Static Investigator Global Assessment of Psoriasis of 0 or 1 and at least a 2-point improvement from baseline and PASI75/90/100 responses (Wk 16); proportion of pts achieving resolution of enthesitis, dactylitis, and minimal disease activity (Wk 24). Binary outcomes, using the Cochran-Mantel-Haenszel-method and continuous outcomes, using mixed-effects model, were analyzed for repeated measures in the subgroups of UPA monotherapy and combination therapy. Point estimates and 95% confidence intervals (CIs) of PBO subtracted treatment effect were calculated. Treatment-emergent adverse events (TEAEs) were analyzed.Results:Of 1916 pts, 574 (30%) received monotherapy and 1342 (70%) received combination therapy; 84% in combination therapy group received MTX +/- another non-bDMARD. Both UPA monotherapy and combination therapy led to improvements in efficacy vs PBO and across endpoints, for each dose, generally consistent point estimates of PBO subtracted treatment effect and associated overlapping CIs were observed (Figure 1). Generally, frequency of AEs and serious AEs, were comparable with UPA administered as monotherapy and combination therapy (Table 1). Frequency of AEs of serious infections and hepatic disorder were lower with monotherapy while frequency of AEs leading to discontinuation of study drug were lower with combination therapy. Most hepatic disorders were transient transaminase elevations.Conclusion:In the SELECT PsA trials, efficacy and safety of UPA was generally consistent when administered as monotherapy or when given in combination with non-bDMARDs. Results from this analysis support the use of UPA with or without concomitant non-bDMARDs.References:[1]Ianculescu I and Weisman MH, Clin Exp Rheumatol 2015; 33:S94–S97.[2]Mease PJ, et al. RMD Open 2015; 1:e0000181.[3]McInnes IB, et al. Ann Rheum Dis, 2020; 79:12.[4]Genovese MC, et al. Ann Rheum Dis, 2020; 79:139.Acknowledgements:AbbVie and the authors thank the patients, study sites, and investigators who participated in this clinical trial. AbbVie, Inc was the study sponsor, contributed to study design, data collection, analysis & interpretation, and to writing, reviewing, and approval of final version. No honoraria or payments were made for authorship. Medical writing support was provided by Ramona Vladea of AbbVie Inc.Disclosure of Interests:Peter Nash Speakers bureau: AbbVie, BMS, Roche, Pfizer, Janssen, Amgen, Sanofi-Aventis, UCB, Eli Lilly, Novartis, and Celgene, Consultant of: AbbVie, BMS, Roche, Pfizer, Janssen, Amgen, Sanofi-Aventis, UCB, Eli Lilly, Novartis, and Celgene, Grant/research support from: AbbVie, BMS, Roche, Pfizer, Janssen, Amgen, Sanofi-Aventis, UCB, Eli Lilly, Novartis, and Celgene, Pascal Richette Speakers bureau: AbbVie, Biogen, Janssen, BMS, Roche, Pfizer, Amgen, Sanofi-Aventis, UCB, Lilly, Novartis, and Celgene, Consultant of: AbbVie, Biogen, Janssen, BMS, Roche, Pfizer, Amgen, Sanofi-Aventis, UCB, Lilly, Novartis, and Celgene, Laure Gossec Speakers bureau: Abbvie, Amgen, Biogen, BMS, Celgene, Lilly, Novartis, Pfizer, Janssen, Sandoz, Sanofi-Aventis, UCB, Consultant of: Abbvie, Amgen, Biogen, BMS, Celgene, Lilly, Novartis, Pfizer, Janssen, Sandoz, Sanofi-Aventis, UCB, Grant/research support from: Abbvie, Amgen, Biogen, BMS, Celgene, Lilly, Novartis, Pfizer, Janssen, Sandoz, Sanofi-Aventis, UCB, Antonio Marchesoni Speakers bureau: AbbVie, BMS, Celgene, Eli-Lilly, Janssen, MSD, Novartis, Pfizer, and UCB, Consultant of: AbbVie, BMS, Celgene, Eli-Lilly, Janssen, MSD, Novartis, Pfizer, and UCB, Christopher T. Ritchlin Consultant of: AbbVie, Amgen, Bristol-Myers Squibb, Janssen, Novartis, UCB, Grant/research support from: UCB, Koji Kato Shareholder of: AbbVie, Employee of: AbbVie, Erin McDearmon-Blondell Shareholder of: AbbVie, Employee of: AbbVie, Elizabeth Lesser Shareholder of: AbbVie, Employee of: AbbVie, Reva McCaskill Shareholder of: AbbVie, Employee of: AbbVie, Dai Feng Shareholder of: AbbVie, Employee of: AbbVie, Jaclyn Anderson Shareholder of: AbbVie, Employee of: AbbVie, Eric Ruderman Consultant of: AbbVie, Amgen, Gilead, Janssen, Lilly, Novartis, and Pfizer.


Sign in / Sign up

Export Citation Format

Share Document