Improving Search for Evidence-based Practice using Information Extraction.

Area of Science:

• Health Informatics
• Information Science

Background:

• Current evidence-based practice (EBP) resources often lack explicit presentation of crucial study data, hindering effective research article retrieval.
• Key information such as patient demographics, study design, and results are frequently omitted or difficult to access.

Purpose of the Study:

• To develop and evaluate a novel method for automatically extracting essential data from research articles.
• To enhance the usability of research findings for evidence-based practice by improving data accessibility.

Main Methods:

• A two-step supervised soft classification approach was employed to extract specific data points from research articles.
• This method automates the extraction process, reducing the labor intensity associated with manual data annotation.

Main Results:

• The proposed classification method demonstrates a more flexible and less labor-intensive alternative to manual data extraction.
• The extracted data can be utilized to improve information seeking support systems for evidence selection.

Conclusions:

• Automated data extraction from research articles is feasible and beneficial for evidence-based practice.
• This approach facilitates better utilization of research evidence in clinical decision-making and information retrieval systems.