Healthcare AI Project - Evidence Synthesis Workflows

Brief and project description

AI-assisted tools (ML and NLP) for abstract screening in systematic literature reviews to reduce manual workload and time in evidence synthesis across healthcare topics.

Quantified impact

55–64% fewer abstracts needing human review in dual-screening scenarios

Strategy

Deploy ML/NLP classifiers to predict abstract eligibility and support/replace second reviewers.
Use WSS@95% (work saved over sampling at 95% recall) to quantify workload reduction

Design

Comparative studies of human-only vs AI-assisted screening.
25 included studies; 17 reported >50% time reduction, with 5–10× decreases in abstract review workload

Client

Pharmaceutical/health technology assessment and outcomes research teams (e.g., Frontiers in Pharmacology)

Project Info

Brief and project description

Strategy

Design

Client

Subscribe for latest AI updates & insights

Resources

AI Services