Reporting completeness of nutrition and diet-related randomised controlled trials protocols

Abstract

Background and AimsThere is a need to consolidate reporting guidance for nutrition randomised controlled trial (RCT) protocols. The reporting completeness in nutrition RCT protocols and study characteristics associated with adherence to SPIRIT and TIDieR reporting guidelines are unknown. We, therefore, assessed reporting completeness and its potential predictors in a random sample of published diet and nutrition-related RCT protocols. MethodsWe conducted a meta-research study of 200 nutrition and diet-related RCT protocols published in 2019 and 2021 (aiming to consider periods before and after the start of the COVID pandemic). Data extraction included bibliometric information, general study characteristics, compliance with 122 questions corresponding to items and subitems in the SPIRIT and TIDieR checklists combined, and mention to these reporting guidelines in the publications. We calculated the proportion of protocols reporting each item and the frequency of items reported for each protocol. We investigated associations between selected publication aspects and reporting completeness using linear regression analysis. ResultsThe majority of protocols included adults and elderly as their study population (n=73; 36.5%), supplementation as intervention (n=96; 48.0%), placebo as comparator (n=89; 44.5%), and evaluated clinical status as the outcome (n=80; 40.0%). Most protocols described a parallel RCT (n=188; 94%) with a superiority framework (n=141; 70.5%). Overall reporting completeness was 52.0% (SD=10.8%). Adherence to SPIRIT items ranged from 0% (n=0) (data collection methods) to 98.5% (n=197) (eligibility criteria). Adherence to TIDieR items ranged from 5.5% (n=11) (materials used in the intervention) to 98.5% (n=197) (description of the intervention). The multivariable regression analysis suggests that a higher number of authors [β=0.53 (95%CI: 0.28-0.78)], most recent published protocols [β=3.19 (95%CI: 0.24-6.14)], request of reporting guideline checklist during the submission process by the journal [β=6.50 (95%CI: 2.56-10.43)] and mention of SPIRIT by the authors [β = 5.15 (95%CI: 2.44-7.86)] are related to higher reporting completeness scores. ConclusionsReporting completeness in a random sample of 200 diet or nutrition-related RCT protocols was low. Number of authors, year of publication, self-reported adherence to SPIRIT, and journals’ endorsement of reporting guidelines seem to be positively associated with reporting completeness in nutrition and diet-related RCT protocols.