Real world evidence (RWE) - Are we (RWE) ready?

Area of Science:

• Health research methodology
• Clinical epidemiology
• Data science in healthcare

Background:

• Real-world evidence (RWE) complements randomized controlled trials (RCTs) but has inherent limitations.
• Both RWE and RCTs require careful design for valid interpretation and extrapolation.
• Discordance between RWE and RCTs typically favors RCTs due to their rigorous methodology.

Purpose of the Study:

• To highlight the importance of proper real-world study design.
• To discuss the strengths and limitations of RWE compared to RCTs.
• To explore the application of RWE in comparative effectiveness and personalized medicine.

Main Methods:

• Discussion of study design principles for RWE.
• Comparison of RWE methodologies (e.g., propensity score matching) with RCT standards.
• Exploration of machine learning applications in electronic healthcare databases.

Main Results:

• RCTs are considered the gold standard due to randomization, prospective, and double-blind designs.
• Real-world studies, even prospective ones with matching, may not fully address confounding factors.
• Machine learning enables identification of patient characteristics for optimal drug response.

Conclusions:

• Rigorous design is crucial for generating reliable real-world evidence.
• While RCTs hold a higher evidence hierarchy, RWE is valuable for specific research areas.
• Advancements in data mining and machine learning enhance the utility of RWE for personalized medicine.