Article type
Year
Abstract
Background:Outcomes reporting bias is a prevalent and easily overlooked problem in clinical research. Since 1992, the Outcome Measures for Rheumatology Clinical Trial (OMERACT) initiative to identify and promote core outcome set in the field of rheumatic disease and with remarkable achievements. At the same time, the number of completed COS studies is rapidly increasing. As of November 2019, the COMET database contains 1327 references of planned, ongoing and completed work.
Objectives:The aim of our study was to assess inconsistencies between protocols and full-text publications in developing core outcome sets (COS), and to help improve the methodology.
Methods:Protocols and subsequent full-text publications were retrieved by searching electronic bibliographic databases. We summarized the changes of general and methodology characteristics by comparing the protocols with full-text publications and represented the information using frequency and proportion.
Results:A total of 24 protocols and 32 corresponding full-text publications were identified from databases. All included studies involved 14 different study topics. The earliest protocol was published in 2013. The gap of published years was one to four years between the full-text publications and the protocols. As for general characteristics, the first-authors of nine studies (37.5%) and the correspondent-authors of eight studies (33.3%) have changed. As for methodological characteristics, 11 studies (45.8%), literatures reviews were the only sources of the initial list of outcomes. And six (25.0%) study’s initial list of outcomes identified from literatures reviews and interviews, four (16.7%) study’s initial list of outcomes identify from literatures reviews and interviews and consultation. Four studies (16.7%) added study types of the literature reviews in full-text publications. Six studies (25.0%) added surveyed participants in full-text publications, and 11 studies (45.8%) participants were found inconsistent in consensus meeting.
Conclusions:Key methods reported in protocols and publications of core outcome sets were inconsistent. Methodologist should combine medical editors and scientific journals to encourage the registration of protocol before developing of COS.
Patient or healthcare consumer involvement:none.
Objectives:The aim of our study was to assess inconsistencies between protocols and full-text publications in developing core outcome sets (COS), and to help improve the methodology.
Methods:Protocols and subsequent full-text publications were retrieved by searching electronic bibliographic databases. We summarized the changes of general and methodology characteristics by comparing the protocols with full-text publications and represented the information using frequency and proportion.
Results:A total of 24 protocols and 32 corresponding full-text publications were identified from databases. All included studies involved 14 different study topics. The earliest protocol was published in 2013. The gap of published years was one to four years between the full-text publications and the protocols. As for general characteristics, the first-authors of nine studies (37.5%) and the correspondent-authors of eight studies (33.3%) have changed. As for methodological characteristics, 11 studies (45.8%), literatures reviews were the only sources of the initial list of outcomes. And six (25.0%) study’s initial list of outcomes identified from literatures reviews and interviews, four (16.7%) study’s initial list of outcomes identify from literatures reviews and interviews and consultation. Four studies (16.7%) added study types of the literature reviews in full-text publications. Six studies (25.0%) added surveyed participants in full-text publications, and 11 studies (45.8%) participants were found inconsistent in consensus meeting.
Conclusions:Key methods reported in protocols and publications of core outcome sets were inconsistent. Methodologist should combine medical editors and scientific journals to encourage the registration of protocol before developing of COS.
Patient or healthcare consumer involvement:none.