Living systematic reviews: 2. Combining human and machine effort.

Area of Science:

• Evidence synthesis methodologies
• Health informatics
• Systematic review processes

Background:

• Living systematic reviews require sustainable and feasible approaches.
• Traditional systematic reviews face challenges in keeping pace with new research.
• Integrating human intelligence with machine automation offers potential solutions.

Purpose of the Study:

• To explore new approaches for evidence synthesis using human-machine collaboration.
• To enhance the feasibility and sustainability of living systematic reviews.
• To identify current applications and future research needs for human-machine technologies in systematic reviewing.

Main Methods:

• Developing workflows where human effort and machine automation mutually reinforce each other.
• Leveraging contributions from online communities (crowdsourcing) and traditional experts.
• Utilizing automation for tasks like searching, eligibility screening, data extraction, and risk of bias assessment.

Main Results:

• Human-machine collaboration can significantly enhance systematic review productivity.
• Automation assists in various systematic review tasks, freeing up human resources for complex judgments.
• New workflows demonstrate potential for more effective and efficient evidence synthesis.

Conclusions:

• Hybrid human-machine approaches are crucial for the advancement of living systematic reviews.
• These enabling technologies are applicable to both living and standard systematic review methods.
• Further research and development are needed to optimize human-machine integration in evidence synthesis.