Heterogeneous use of registry data for participant identification and primary outcome ascertainment is found in registry-based randomized controlled trials: a scoping review

Registry-based randomized controlled trials (RRCTs) have potential to address limitations of traditional clinical trials. To describe their current use, information on planned and published RRCTs was identified and synthesized. A scoping review of published RRCT protocols and reports was conducted. Articles published between 2010 and 2021 identified from electronic database searching, a recent review of RRCTs, and targeted searching for recent RRCT protocols (2018-2021) were screened. Data on trial data sources, types of primary outcomes, and how these primary outcomes were described, selected, and reported were extracted. Ninety RRCT articles (77 reports; 13 protocols) were included. Forty nine (54%) used or planned to rely on registry data for their trial, 26 (29%) used both registry and additional data, and 15 (17%) used the registry solely for recruitment. Primary outcomes were routinely collected from the registry for 66 articles (73%). Only 28 articles (31%) described any methods to promote outcome data quality during or after data collection. Core outcome sets were not used in any of the trials. With improvements in registry design, outcome selection, measurement, and reporting, future RRCTs may deliver on promises of efficient, high-quality trials that address clinically relevant questions.