SAME PAGE: Bridging the gaps between high-quality clinical research evidence for COVID‑19 and clinicians, guideline developers, and policy makers
Due to worldwide interest in COVID-19, over 300 research articles are published daily. However, only a small fraction of these articles are of high enough quality to be used to change how COVID‑19 is managed. COVID‑19 Premium Literature Service (COVID-19+ EA) is a university-based service that uses computers to search for COVID‑19 artricles nightly, and then uses trained human researchers to sort the high-quality articles from the low-quality articles. The titles of these articles are currently posted on a free website that is used by clinicians, guideline developers, public health officers, and is open to the public. The purpose of this grant is to achieve 2 objectives. (1) To find more efficient ways to select the articles that our human staff must read. The high number of daily articles makes manually checking them for quality very time-consuming. To help with this task, we have trained a computer to search for words and phrases that more frequently appear in high-quality articles (called machine-learning, a form of artificial intelligence). Early results are promising, but we need additional testing against our human readers to be sure the computer doesn’t discard any high-quality articles by mistake. (2) To add 2 new services to COVID-19+ EA. One new service will be a focused search for articles on COVID‑19 vaccines. These articles will be checked for quality and kept in a separate collection. The other new service will pull out data on sex, age, and race from all high-quality articles and label the ones that report results separately based on these features. These new collections will help researchers identify patterns that may not have been obvious in any single study alone. By supporting and improving COVID-19+ EA, this grant will help make the SAME high-quality evidence accessible to everyone in a form that is tailored to their needs. In turn, this evidence will be used to help prevent, diagnose, and treat COVID‑19 in Canadians.
