BACKGROUND
Figures are an important form of expressing results commonly found in medical papers and make data easy to read and compare. The quality of graphs in original papers has improved in western medical journals. However, some figures fail to correctly express the results of a paper. Additionally, graph quality and application has not been assessed in medical journals outside western countries.
OBJECTIVE
To determine the frequency and types of data graphs used in Chinese academic medical journals and evaluate the quality of graphs used in original medical papers.
METHODS
A total of 783 papers were surveyed from the medical journals of five colleges and universities in Shanghai from 2011 to 2015. A cross-sectional study was used to analyse the applied status and graph quality. The evaluation criteria of graphs mainly included graph type, visual clarity, completeness, and special standards.
RESULTS
Most authors prefer to use simple charts, and bar charts with 95% CI were the most widely used. More than 60% of charts have problems with visual clarity, completeness, and special standards. Of 841 incorrect graphs, 10 (0.58%) graphs had three combined problems of graph characteristics, and 292 (34.72%) graphs had any two combined problems of graph characteristics. For detailed errors, the absence of variance description was the most substantial problem, especially in 2014 and in some academic medical journals.
CONCLUSIONS
Graphs are less commonly applied in the five university journals. However, the quality of papers using graphs was not properly controlled. Editors and journal quality management should strengthen the quality control of charts in papers. Authors should also avoid error bias and distorting their conclusions.
The Trouble with Medical Journals
by Richard Smith: an alternative view
