Improving the efficiency of identifying practice ready evidence and generating evidence summaries

EBSCO improves the dissemination of high-quality, clinically relevant content to healthcare providers through DynaMed, their online clinical textbook that is constantly being updated as new research is published. This process includes many experts and has many steps. McMaster University’s Health Information Research Unit’s (HiRU) literature surveillance program filters medical research studies and passes them to EBSCO. Clinical experts review the studies to determine how relevant they are to current practice across 1000’s of clinical topics (i.e., related to diagnosis, treatment, monitoring of health conditions) and triage them into add now, add later, or do not add. Studies to be added are meticulously evaluated by highly trained EBSCO staff who then generate a summary of the content for the authors who update the textbook chapters. The process is time-consuming and there are some steps that could be more efficient with the use of artificial intelligence (AI) approaches. In this project, student interns from McMaster will apply AI to improve 3 processes in this pathway: improving how well articles are triaged with the aim of having fewer ‘add later’ that are subsequently dropped; automating a standard quality assessment of the studies; and testing the reliability and trustworthiness of summaries that are generated by AI. Collectively, the projects will improve the efficiency of the article selection to summary preparation stages of the literature surveillance system.

Faculty Supervisor:

Ashirbani Saha;Cynthia Lokker

Student:

Partner:

EBSCO Health

Discipline:

Computer science

Sector:

Information and cultural industries; Retail trade

University:

McMaster University

Program: