மருத்துவ பரிசோதனைகளின் இதழ்
திறந்த அணுகல்

சுருக்கம்

Risk of Bias in Clinical Trials Reported for Foods with Functional Claims in Japan: A Cross-Sectional Study on Research Quality

Hiroharu Kamioka*, Hideki Origasa, Jun Kitayuguchi, Takahiro Yoshizaki, Mikiko Shimada, Yasuyo Wada, Hiromi Takano-Ohmuro, Kiichiro Tsutani

Background: The Foods with Function Claims (FFC) notification system was introduced in Japan in April 2015. We hypothesized that there would be risk of bias (RoB) specific to health food interventions. The purpose of this cross-sectional study was to clarify RoB and related factors of clinical trials (CTs) reported as the scientific basis of efficacy in the FFC.

Methods: All 103 articles based on CTs published on the Consumer Affairs Agency website from 1 July 2018 to 30 June 2021 were reviewed. We evaluated 14 items, the highest RoB: 14 points (pts), as well as related items including first author characteristics, journal name, year published, journal impact factor, article language, and name of clinical trial registration.

Results: The RoB score was 5.7 ± 2.5 pts. In general, there was a remarkable lack of execution and/or description of the intention-to-treatment analysis (81.6%), compliance (68.0%), and multiple outcome tests (67.0%). There was no significant difference (p=0.051) in RoB score between the published year categories of 2015-2017 (6.5 ± 2.4 pts) and 2018-2021 (5.5 ± 2.4 pts). There was also no significant difference (p=0.247) in RoB score between English (5.5 ± 2.6 pts) and Japanese (6.0 ± 2.3 pts) language publications, and no significant difference (p=0.740) between for-profit (5.7 ± 2.4 pts) and academia (6.0 ± 2.8 pts) in authors’ organization. A significant correlation (p=0.099) between IF and RoB score was not observed with Spearman’s rank correlation coefficient; r = -0.163. Conclusions: Four common biases in most CTs reported in the FFC were randomization, deviations from intended interventions, measurement of outcome, and selective reporting. In particular, RoB including lack of ITT analysis, unknown compliance, and multiple outcome tests seriously damaged the study quality.

Review registration: The study was registered as UMIN 000046267 by the University Hospital Medical Information Network Clinical Trials Registry (UMIN-CTR)* in Japan (refer: https://upload.umin.ac.jp/cgi-open-bin/ctr/ctr_view. cgi?recptno=R000052795).