(191) Landscape Assessment of Clone-Censor-Weight Methodology Application in Real-World Data Studies: A Scoping Review

Background: Target trial emulation has become an increasingly popular framework to aid in identifying causal effects from real-world data (RWD) studies. A commonly applied analytic method within this framework is “clone-censor-weight” (CCW), which helps mitigate, if not eliminate, immortal time bias through the creation of a pseudo-population. A major strength of CCW is its versatility across a variety of RWD research questions – such as non-user comparisons and dynamic treatment strategies – but little is known about its application, dispersion, and reporting across such studies.

Objectives: To assess the use of CCW in addressing comparative research questions that use RWD.

Methods: A scoping literature review strategy was applied to articles between Jan 2016 and Aug 2023. Using PubMed, EMBASE, and citations of three prominent CCW papers, candidate articles were collected and assessed for inclusion independently by two reviewers. Included articles were RWD studies that used CCW to answer a comparative clinical research question. Data extraction focused on study characteristics, CCW details, and reporting mechanisms, which were used to inform descriptive statistics on the application and uptake of CCW.

Results: Of 2079 candidate articles, 131 qualified for full-text screening, with 39 for data extraction. Of those 39, most were published in 2022 or later (28 [72%] articles), with the most common therapeutic areas being oncology (13 [33%]) and infectious diseases (9 [23%]). The research questions included initiating an intervention (therapeutic or procedural) v no initiation (20 [51%]); comparing dynamic treatment strategies (8 [21%]); intervention continuation v discontinuation (6 [15%]); and time to intervention initiation (5 [13%]). The majority specified using a target trial emulation approach (30 [77%]) but reporting of relevant components varied. CCW was chosen as the primary analysis in 30 [77%] articles, while 4 [10%] applied it as a sensitivity analysis and 5 [13%] applied it as part of an explicit methods comparison for a sample comparative research question. All articles using CCW as a sensitivity analysis indicated that the CCW results were similar to the primary analysis.

Conclusions: CCW is becoming more prominent across many applications, particularly for addressing comparisons with difficult to define study entry points. Although familiarization of CCW is being explored through sensitivity analyses and methods assessments, many have applied CCW as a primary analysis because it was an optimal strategy for their question. Further synthesis, considerations, and takeaways are currently ongoing, with the intention to further inform how to best leverage and apply this method in future RWD studies.