Published On: 10/9/2024

Advancing RWE: Target RWE's Staging and Clean Room Committee Approaches Set New Benchmark

Target RWE publication demonstrates the effectiveness of advanced real-world evidence analyses – showcasing the company's innovative analytical expertise

DURHAM, N.C., Oct. 9, 2024 /PRNewswire/ -- Target RWE, a leading provider of real-world evidence (RWE) solutions, today announced a study in Pharmacoepidemiology and Drug Safety which introduces novel methodological approaches aimed at enhancing the integrity and reproducibility of real-world data analyses.

The research, titled "Staging and clean room: Constructs designed to facilitate transparency and reduce bias in comparative analyses of real-world data," outlines two methodological innovations developed by Target RWE to increase transparency and reduce the risk of study bias.

• Staging: Target RWE applies a series of checkpoints at key study stages to review preliminary results and identify potential concerns. Early identification of concerns, such as the presence of bias, can be addressed before proceeding with critical next analyses.

• Clean room committee: In coordination with the staging process, an independent review committee (clean room committee) shielded from knowing how their recommendations might affect results, can review potential concerns during staging and provide direction to the study team.

"As the use of real-world data in healthcare decision-making continues to grow, we must develop robust processes to ensure the validity and transparency of our analyses," said M. Alan Brookhart, PhD, Target RWE Scientific Advisor and co-author of the study. "The use of a staged analysis approach overseen by a clean room committee are two principled approaches for increasing confidence in the findings from non-randomized studies."

The staging and clean room process is another advancement at Target RWE, in addition to the application of proprietary advanced analytics and visualizations on the causalStudio™ software platform, to conduct the most robust analyses possible.

"Our software, causalStudio™, has been designed to use cutting edge, principled epidemiology methods to solve complex, clinical questions and generate meaningful real-world evidence for regulators, payers, and clinicians," said Target RWE Chief Scientific Officer Jennifer Christian, PharmD, PhD, FISPE. "By leveraging our advanced analytical capabilities, partners can confidently address complications in the data, conduct causal inference analyses, and analyze various data types - ultimately driving more informed decision-making in clinical research and drug development."

causalStudio™ addresses real-world data challenges with causal inference methods to produce meaningful and valid real-world evidence. Additional examples of advanced RWE studies and methods supported by causalStudio™ include:

• Negative Control Outcome Studies - to assess comparability of potential treatment groups. Treatment use in the real world is affected by many factors associated with a patient's prognosis that can lead to bias when comparing treatment groups in observational settings. It is important to understand whether patients initiating a particular treatment can be compared to patients initiating similar medications after controlling for measured confounding factors. This information is critical to determine if there may be unmeasured confounders and whether or not comparative safety and/or comparative effectiveness studies should be conducted.

• Clone Censor Weighting – enables the evaluation of optimal treatment regimens at critical disease management decision time points. The approach involves 'cloning' patients into cohorts based on pre-defined treatment sequences, allowing for direct comparisons of treatment protocols and easy clinical interpretation. Patients deviating from the study sequence are artificially censored and reweighted to reflect the original population; ensuring that the observed treatment effects are not confounded by factors related to patient adherence or selection bias.

• Predictive Modeling with Machine Learning – backed by expert-driven AI, we have proven experience using predictive modeling with claims data, including the development of empirically derived variables, variable screening, variable importance measures, Harrell C index, inverse probability of censoring weighting, regularized Cox regression, and survival trees.

causalStudio™ is comprised of two unique components: causalRisk™ - a software package designed to simplify the complex process of causal inference studies, and causalPHR™ - an intuitive and interactive tool to visualize, stratify, filter, and publish analytical results. Both causalRisk™ and causalPHR™ are data agnostic and developed by statisticians and developers at Target RWE.

Building on the success of causalStudio™, Target RWE will be rolling out new analytical modules in the future to continue empowering researchers with cutting-edge tools, particularly to assist users with clone-censor-weighting analyses, estimation of counterfactual effects for repeated measures of categorical or continuous data, such as longitudinal laboratory values, and nested trial emulation.

To learn more about causalStudio™ and how it has been used in the real world click here or contact our team at info@targetrwe.com.