Graphic data representation in anaesthesiological journals: A proposed methodology for assessment of appropriateness

Few authors have addressed the topic of graphic data presentation. The purpose of our study was to combine several guidelines in order to evaluate three anaesthesiology journals listed in Index Medicus (Australian, American and Italian) in terms of the appropriateness and the quality of presentation of graphs. Our analysis was based on concepts expressed by Cox and Tufte. We calculated the optimization of the amount of information in each graph using two parameters: Data Density Index (DDI) and Data Ink Ratio (DIR). The correctness and clearness of each component of the graph (scale, title, axes, legends and abbreviations) was evaluated on the basis of a binary score. We analysed 300 exploratory plots, quantitative graphs and summaries of statistical analysis. About 50% of papers had more than three graphs. Mean scores were 3.22 for the Italian journal, 3.47 for the American journal and 3.82 for the Australian journal. Tufte parameters were calculated on 42 scatterplots: DDI was 5.4+/-13.9 and DIR was 0.7+/-0.1. The criteria applied in our study appear sufficiently sensitive to differentiate the quality of graphs.