Translating Clinical Questions by Physicians Into Searchable Queries: Analytical Survey Study

Area of Science:

• Medical Informatics
• Clinical Decision Support

Background:

• Accessing current best evidence for clinical questions is challenging for healthcare professionals.
• Federated search engines like MacPLUS FS aim to simplify evidence retrieval by searching multiple resources simultaneously.
• Clinicians need to effectively translate clinical questions into searchable queries.

Purpose of the Study:

• To analyze how clinicians formulate search queries using the population, intervention, comparison, and outcome (PICO) framework on a federated search engine.
• To identify the types of clinical questions (background vs. foreground) posed by clinicians.
• To determine which evidence resources clinicians access most frequently.

Main Methods:

• An analytical survey was conducted with 908 clinicians using MacPLUS FS during a randomized controlled trial.
• 1085 queries were extracted over 6 months and analyzed using the PICO framework.
• Queries were categorized as background or foreground, and accessed resources were documented.

Main Results:

• Of 695 structured queries, 56.5% were background and 43.5% were foreground questions, primarily concerning therapy.
• Queries typically included 3 terms, focusing on population and intervention, rarely on outcome or comparator.
• Clinicians accessed summaries (51.5%) most often, followed by preappraised (24.4%) and non-preappraised research (24.1%).

Conclusions:

• Real-world query data can inform educational programs to enhance clinicians' evidence retrieval skills.
• Findings can guide the development of more effective evidence-based clinical resources.
• Understanding query patterns is crucial for optimizing clinical decision support tools.