Menu Back to Poster-Presentations-Details

P236: Considerations for Multiple Enrollment Subject Data for Analysis and Reporting

Poster Presenter

Jon Godshall

Senior Director, Statical Data Sciences and Analytics
Pfizer Inc
United States

Objectives

This poster outlines key concepts for consideration in creation of data sets and analysis planning for studies that include participants who have violated the protocol by knowingly or unknowingly enrolling in the same trial multiple times at one or more study sites, referred as “multiple-enrollers".

Method

For this topic, we reviewed FDA guidance, industry examples and our own study experience.

Results

Multi-enrolling subjects in a clinical trial present unique challenges in the data and analysis of the trial. Unlike other multiple study entry issues for re-screens, extension rollovers, etc., multi-enrollers present the complication of potentially overlapping study data for the same person after initiating administration of investigational product. Data collection tools must allow for identification of multi-enrollers and if identified during their active participations, disruption of normal data collection can occur. Although CDISC guidelines allows for both the unique subject identifying number across the study (USUBJID) as well as subject numbers (SUBJID’s) per enrollment attempt , the parallel collection of data from the multiple enrollments must be accounted for and defined clearly in analysis plans. Assignment of treatment arm for reporting is complicated with the possibility of different treatment randomization assignments across enrollments for the same individual. ICH principles, as well as the objectives of the study, different study endpoints, and per-protocol population definition must be factored into the decisions on how to report the data, especially if combining with data from participants who were not multiple enrollers. Options, including sensitivity analysis, do exist and some will be featured on the poster.

Conclusion

These multiple-enrolling participants present challenges to the typical CDISC dataset structures and methods for summarizing subjects as unique study entities in analysis tables, different than for re-screened participants, extension study/rollover participants, and transfer participants. No one method serves the purpose of solving the analytical challenges presented by multi-enrollers across every clinical trial. The different scenarios with which multi-enrollers can present across diverse types of clinical trials is in itself seemingly infinite. However, some guidance, examples, and options exist for featuring the data accurately in both data sets and outputs. If the number of multi-enrollers is significant in a study, alternatives to simply excluding from analyses. The poster will explore some of the options for complying with data set guidance and representing these subjects’ study data in tables, figures, and listings.