November 2025
Value in Health
An AI-assisted workflow demonstrated a 50% relative reduction in the time required to extract tabular HEOR data for systematic reviews without compromising data completeness.
Fewer than one-in-six screened medical articles can lawfully feed commercial AI pipelines under their native licenses, a barrier that processing XMLs from services like RightFind can overcome to more than double the coverage.
October 2025
The journal of Allergy and Clinical Immunology
This network meta-analysis of 31 studies clarifies the comparative efficacy and safety of intranasal treatments for allergic rhinitis in children, showing which medications perform best for seasonal versus perennial AR.
September 2025
Open-weight large language models demonstrate performance comparable to, and at times surpassing, leading proprietary systems in the context of biomedical question-answering.
January 2025
BMJ Evidence-Based Medicine
An analysis of using AI and Large Language Models for evidence synthesis, based on insights from the third Methods Forum of Cochrane Germany.
November 2024
Cochrane Colloquium Abstracts
Preliminary results from a comparative study suggest that AI-assisted full-text screening can expedite the systematic review process, demonstrating a time saving of 7 seconds per study without compromising the quality of reviewer decisions.
Achieving 100% specificity and 95% sensitivity, the SuperDeduper module demonstrates a safe and superior method for removing reference duplicates compared to six established literature review tools.
Evidence-Based Toxicology Journal
This protocol evaluates if integrating a semi-automated, machine learning-driven tool for data extraction can significantly enhance the efficiency and user experience in systematic human health chemical assessments.
June 2024
This AI-supported systematic review establishes a comprehensive dataset, which is crucial for future economic studies and identifying global research priorities.
January 2024
The European Journal of Allergy and Clinical Immunology
This systematic review found that patients with allergic rhinitis (AR) generally consider nasal symptoms, particularly congestion, to be the most important outcomes, leading to lower utilities with increased disease severity.
November 2023
AI tool demonstrates an average workload saving of 45.4% for systematic literature review screening while achieving 100% sensitivity in seven of eight updates.
September 2023
Integration of a machine learning classifier into a living systematic review for VTE outcomes in COVID-19 patients proved reliable and delivered an efficiency gain equivalent to 214 working days of manual screening.
June 2023
The AI-assisted single screening approach reduced the systematic review screening workload by 43% while successfully identifying 100% of the relevant studies, saving over 19 hours of screening time.
June 2022
German Medical Science
The successful application of a machine learning classifier to a high-volume living systematic review reduced the screening workload by 76% while maintaining 95% sensitivity for continuous, up-to-date evidence synthesis.
January 2022
Environment International
The Dextr semi-automated extraction tool substantially reduced the median data extraction time for environmental health studies by over 50% (933 s vs. 436 s per study) compared to a manual workflow, while maintaining a comparable precision rate.
June 2020
The Lancet
This systematic review and meta-analysis provides quantitative evidence supporting physical distancing of one meter or more, face mask use, and eye protection as effective measures to prevent the person-to-person transmission of SARS-CoV-2 and COVID-19.
January 2019
Arxiv
This work presents the deep learning, BI-LSTM-CRF sequence tagging model, featuring interleaved LSTM layers, that earned the first-place ranking at the 2018 Text Analysis Conference for systematic review data extraction.
Flexible plans designed for teams and projects of every size.