Trial Tokenization Accelerating Innovation in SEPRA – A Pragmatic Randomized Trial

OBJECTIVES:

SEmaglutide once-weekly randomized PRAgmatic Trial (SEPRA) (NCT03596450) is an ongoing, randomized, pragmatic trial comparing the effects of once-weekly subcutaneous semaglutide with other treatments for type 2 diabetes in routine clinical practice in the US. Traditionally, endpoints such as health care resource utilization and adherence are patient-reported in randomized trials. SEPRA utilized tokenization and linkage of study participants’ trial data to their claims data to support evaluation of supportive secondary endpoints. The objective of this project was to evaluate the overlap of participants with available claims data and compare baseline characteristics with the overall trial population.

METHODS:

Upon completion of informed consent, HealthVerity’s privacy-preserving record linkage technology de-identified participants' personally identifiable data with a unique, persistent token. These tokens were then matched within HealthVerity’s real-world data ecosystem to evaluate the number of participants with available claims data between July 1, 2017 and October 31, 2020. Baseline characteristics were then measured in the 1-year pre-randomization period and compared against the total trial population.

RESULTS:

Of the 1,278 participants enrolled in SEPRA, 49.5% had either medical or pharmacy claims data, 32% had both medical and pharmacy, 35% had medical, and 37% had pharmacy claims data. Mean age of participants with claims data was 55.0 years (SD: 10.2) (vs. 57.4 years [SD: 11.1] for the overall trial population) and 45.2% were female (vs. 45.8%). Most participants with claims data lived in the US South (46.1%), followed by the Midwest (23.2%), West (14.5%), Northeast (10.7%), and unknown (5.4%). Similar trends were apparent in the overall trial population.

CONCLUSIONS: Approximately half of the participants enrolled in SEPRA had available claims data and their baseline demographics were similar to the overall trial population. The use of this innovative approach within SEPRA provides an example for future research studies wherein multiple data sources are linked to evaluate patient outcomes through tokenization.