OBJECTIVES: Systematic literature reviews (SLRs) summarize all available evidence using rigorous methodologies to inform regulatory and HTA decisions, guideline development and clinical practice. With the exponential growth of medical literature, the volume of data included in SLRs have increased tremendously. Further, advanced analysis methods such as network meta-analysis are heavily technical, possibly involving multiple networks, which may hinder interpretation. The density and complexity of static information in SLRs causes challenges in the uptake of data from SLRs. Failure to optimally use high-quality research evidence results in inefficiencies and potential inaccuracies in interpretation. This research aims to develop data visualizations to improve readability of SLRs.

METHODS: We carried out a targeted literature review on gastric cancer and gastroesophageal junction cancer in Medline, CDSR, EconLit, and grey literature, identifying 1,544 unique references to inform our initiative. Of those, 47 were ultimately included. Based on these data we developed and tested data visualizations suitable for use with data extractions from SLRs.

RESULTS: Data visualizations have the potential to enhance communication of complex data and key findings of SLRs to stakeholders. Our dynamic data visualization tool allows non-technical users to interact with the data outside the static structure of SLR reports. Users are able to select preset data cuts to visually explore the heterogeneity across included studies and customize the graphical presentations of the key findings. The tool is fully bespoke to maximize interpretability.

CONCLUSIONS: The current reporting paradigm for SLRs has hindered the dynamic, efficient consumption of evidence due to the increasing volumes and complexity of data. Visualizations will help policy makers and other stakeholders to understand and effectively communicate the key value messages from SLRs that can have important impacts on clinical practice.