Medical clinical practice guidelines serve as roadmaps for clinicians to provide step-by-step instructions on the best ways to treat patients. However, creating a reliable medical guideline is a lengthy process. Screening thousands of articles from literature searches is one of the most time-consuming steps in the development of this process.
Limited previous research has shown that two artificial intelligence (AI) tools, DistillerSR and EPPI-Reviewer, can reduce the time needed to identify relevant articles during the article-screening process. However, there is no study to compare these two AI tools in one cancer-related article. So, this study aims to rigorously explore the effectiveness of these two tools across different types of cancer guidelines, both individually and in comparison to each other.
The Program in Evidence-Based Care (PEBC) at the Department of Oncology in McMaster University develops cancer-related guidelines. This study will test DistillerSR and EPPI-Reviewer in 10 previously published PEBC guidelines, with a focus on the five most common cancer types: lung cancer, breast cancer, prostate cancer, colorectal cancer, and bladder cancer.
If findings show that the AI approach can save time without sacrificing quality, the PEBC staff would be able to produce high-quality guidelines more efficiently. This could also free up resources to create additional medical guidelines for cancer care. As a result, clinicians in Ontario would have access to more medical guidelines when treating patients with cancer within the same timeframe.
