The McMaster Health Information Research Unit: Over a Quarter-Century of Health Informatics Supporting Evidence-Based Medicine.
Level 5 - mechanism / opinion, no new human data
Narrative review and historical viewpoint paper without primary empirical data (graded by design analogy)
PubMed 39083765 · doi:10.2196/58764
What was done
This viewpoint and historical article chronicles the evolution of health informatics methods developed at the McMaster University Health Information Research Unit (HiRU) over more than 25 years to support evidence-based medicine. It describes the progression from early MEDLINE dial-up searching and the development of the Hedges-derived Clinical Queries search filters to the recent application of classical machine learning, deep learning, and large language models (LLMs) with human-in-the-loop active learning.
What was found
The abstract reports no experimental comparative data or statistical outcomes. It contextualizes the development of informatics tools, noting the founding of HiRU in 1985, the release of validated Clinical Queries filters in the early 2000s, and the challenge of filtering the approximately 1 million articles added to PubMed each year using modern machine learning and LLM architectures.
Why it matters
The paper outlines the methodological history and ongoing technological transition in evidence-based medicine retrieval systems, documenting how automated filtering tools have evolved to handle rapidly expanding biomedical literature volumes.
Limits
This is a narrative viewpoint and historical commentary rather than an empirical trial or systematic review. The abstract provides no quantitative benchmark metrics, accuracy evaluations, or direct performance comparisons of the various informatics retrieval tools.
Cited by
- supports Approximately one million new publications are indexed in peer-reviewed journals on PubMed each year.