Related Experiment Video
Updated: Jul 10, 2026
06:55
Inverse Probability of Treatment Weighting (Propensity Score) using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
Causal logistic models for non-compliance under randomized treatment with univariate binary response
Thomas R Ten Have1, Marshall Joffe, Mark Cary
1Department of Biostatistics and Epidemiology, University of Pennsylvania School of Medicine, Blockley Hall, 6th FLR, 423 Guardian Dr., Philadelphia, PA 19104-6021, USA. ttenhave@cceb.upenn.edu
Statistics in Medicine
|April 11, 2003
Summary
We introduce a new method for estimating causal effects in trials with treatment non-compliance. This approach helps assess unmeasured confounding by comparing marginal and conditional log-odds ratios.
Area of Science:
- Biostatistics
- Epidemiology
- Clinical Trials
Background:
- Treatment non-compliance is a common issue in clinical trials.
- Existing methods for causal inference with non-compliance have limitations.
- Assessing unmeasured confounding due to non-adherence is crucial.
Purpose of the Study:
- To propose a novel method for estimating the marginal causal log-odds ratio for binary outcomes.
- To provide a marginal alternative to existing conditional causal logistic approaches.
- To enable assessment of unmeasured confounding by comparing marginal and conditional estimates.
Main Methods:
- Extension of Robins' G-estimation approach for logistic models.
- Development of a marginal causal estimation method.
- Simulation studies comparing marginal and conditional approaches under varying confounding levels.
Main Results:
- The proposed marginal method performs well in terms of bias under different confounding levels.
- Simulations show differences in bias and confidence interval coverage between marginal and conditional estimates under stronger confounding.
- Comparison of marginal and conditional estimators aids in evaluating confounding magnitude.
Conclusions:
- The proposed marginal causal estimation method offers a valuable tool for analyzing trials with treatment non-compliance.
- Comparing marginal and conditional estimates can reveal the impact of unmeasured confounding.
- This approach enhances causal inference in the presence of non-adherence.
Related Concept Videos
01:13Randomized Experiments
The randomization process involves assigning study participants randomly to experimental or control groups based on their probability of being equally assigned. Randomization is meant to eliminate selection bias and balance known and unknown confounding factors so that the control group is similar to the treatment group as much as possible. A computer program and a random number generator can be used to assign participants to groups in a way that minimizes bias.
Simple randomization
Simple...
Simple randomization
Simple...
01:23Mechanistic Models: Compartment Models in Individual and Population Analysis
Mechanistic models are utilized in individual analysis using single-source data, but imperfections arise due to data collection errors, preventing perfect prediction of observed data. The mathematical equation involves known values (Xi), observed concentrations (Ci), measurement errors (εi), model parameters (ϕj), and the related function (ƒi) for i number of values. Different least-squares metrics quantify differences between predicted and observed values. The ordinary least squares (OLS)...
01:29Mechanistic Models: Compartment Models in Algorithms for Numerical Problem Solving
Mechanistic models play a crucial role in algorithms for numerical problem-solving, particularly in nonlinear mixed effects modeling (NMEM). These models aim to minimize specific objective functions by evaluating various parameter estimates, leading to the development of systematic algorithms. In some cases, linearization techniques approximate the model using linear equations.
In individual population analyses, different algorithms are employed, such as Cauchy's method, which uses a...
In individual population analyses, different algorithms are employed, such as Cauchy's method, which uses a...
01:21Dose Response Curve: Conventional Versus Nonmonotonic
The correlation between a drug's dosage and its impact on a biological system is a cornerstone of pharmacology and toxicology. Conventional dose–response curves, which include graded and quantal relationships, are key to this understanding. Graded dose–response curves depict the spectrum of a biological reaction to different doses within an individual, indicating that as the drug dosage increases, so does the intensity of the response. On the other hand, quantal dose–response relationships...
01:20Comparing the Survival Analysis of Two or More Groups
Survival analysis is a cornerstone of medical research, used to evaluate the time until an event of interest occurs, such as death, disease recurrence, or recovery. Unlike standard statistical methods, survival analysis is particularly adept at handling censored data—instances where the event has not occurred for some participants by the end of the study or remains unobserved. To address these unique challenges, specialized techniques like the Kaplan-Meier estimator, log-rank test, and Cox...
01:09Censoring Survival Data
Survival analysis is a statistical method used to analyze time-to-event data, often employed in fields such as medicine, engineering, and social sciences. One of the key challenges in survival analysis is dealing with incomplete data, a phenomenon known as "censoring." Censoring occurs when the event of interest (such as death, relapse, or system failure) has not occurred for some individuals by the end of the study period or is otherwise unobservable, and it might have many different reasons...