JMIR Publications

ABSTRACT

Evidence-based medicine (EBM) emerged in the 1980-1990’s and emphasizes the integration of the best research evidence with clinical expertise and patient values. Over the past 25+ years, health informatics approaches developed and implemented by the Health Information Research Unit (HiRU) at McMaster University to support EBM have evolved greatly. Early on, digital health informatics took the form of teaching clinicians how to search MEDLINE with modems and phone lines. Searching and retrieval of published articles was transformed—with PubMed playing a pivotal role—providing an electronic platform to access clinically relevant studies, systematic reviews, and clinical practice guidelines. In the early 2000s, HiRU introduced the Clinical Queries, validated search filters derived from a curated, gold standard, human appraised datasets (HEDGES) to enhance precision of searches, allowing clinicians to narrow down their queries based on study design, population, and outcomes. Currently, almost 1M articles are added to PubMed annually. To filter through this volume of heterogenous publications for clinically important articles, HiRU and other researchers have been applying classical machine learning, deep learning, and, increasingly, large language models (LLMs). These approaches are built upon the foundation of gold standard annotated datasets and humans-in-the loop for active machine learning. We explore the evolution of health informatics in supporting evidence search and retrieval over the past 25+ years within HiRU, including the evolving roles of LLMs and responsible AI as we continue to facilitate the dissemination of knowledge, enabling clinicians to integrate the best available evidence into their clinical practice.