Performance of the Self-Controlled Case Series With Active Comparators for Drug Safety Signal Detection Using the...

Area of Science:

• Pharmacovigilance and Drug Safety
• Real-World Data (RWD) Analysis
• Epidemiological Research Methods

Background:

• The self-controlled case series (SCCS) is a key method for detecting drug safety signals using real-world data (RWD).
• Incorporating active comparators may enhance SCCS performance by addressing confounding by indication.
• The French National Health Data System (SNDS) is a large administrative claims database with underutilized potential for drug safety research.

Purpose of the Study:

• To evaluate the performance of the SCCS method for drug safety signal detection within the French SNDS database.
• To compare the effectiveness of SCCS with and without the inclusion of active comparators.

Main Methods:

• Applied the SCCS method to macrolide and fluoroquinolone antibiotics, using amoxicillin as the active comparator.
• Selected 7 drugs and 30 outcomes, creating a reference set of 104 positive and 58 negative controls.
• Utilized a 2-year observation period with a 30-day risk window, measuring diagnostic performance (sensitivity, specificity) against product labels.

Main Results:

• SCCS without active comparators achieved a sensitivity of 0.89 and specificity of 0.43 (for pairs with satisfactory power).
• Inclusion of active comparators increased specificity to 0.91 but decreased sensitivity to 0.52.
• The French National Health Data System (SNDS) proved to be a valuable data source for signal detection, especially for hospital-recorded outcomes.

Conclusions:

• The SCCS method demonstrated satisfactory performance for drug safety signal detection in the SNDS database.
• Active comparators enhanced overall performance by increasing specificity but significantly reduced sensitivity.
• The SNDS is a viable resource for drug safety signal detection, particularly when using a well-designed reference set for the SCCS method.