Adherence to PRISMA-A and reporting was suboptimal in meta-analysis abstracts on drug efficacy for tumors: a literature survey

OBJECTIVE: To assess the reporting of meta-analysis abstracts on drug efficacy for tumors in terms of adherence to PRISMA-A and identify the potential factors associated with adherence to PRISMA-A. STUDY DESIGN AND SETTING: A total of 3211 eligible meta-analysis abstracts were assessed using a checklist adapted from the PRISMA-A statement. Adherence to PRISMA-A was analyzed by the total PRISMA-A score and adherence rate (AR). The independent samples t-test was performed to compare the difference of the total scores between two groups with different characteristics, and the analysis of variance (ANOVA) or Kruskal-Wallis test was used among multiple groups. The Pearson's correlation coefficient was used to measure the correlation between the word count and the total PRISMA-A score. RESULTS: The mean total score was 8.11(±1.76) and the adherence rate (AR) was 57.94%. The items with lower AR were funding (AR=0.93%), registration (AR=3.86%), and risk of bias (AR=7.85%). Meta-analyses published after the release of PRISMA-A showed better adherence to PRISMA-A. Compared to unstructured abstracts, structured abstracts had a higher AR for each item in PRISMA-A. There was a positive correlation between the word count of abstract and the total PRISMA-A score (r = 0.358, P < 0.001). CONCLUSION: Adherence to PRISMA-A was suboptimal in meta-analysis abstracts on drug efficacy for tumors, despite the improvement after the release of PRISMA-A. Various measures should be implemented to improve compliance with PRISMA-A and enhance the reporting of meta-analysis abstracts, including journal endorsement of PRISMA-A, requirement of stricter adherence to PRISMA-A, relaxation of abstract word limits, etc.