Sequence symmetry analysis in pharmacovigilance and pharmacoepidemiologic studies

Area of Science:

• Pharmacovigilance and Pharmacoepidemiology
• Computational Methods in Health Research

Background:

• Sequence symmetry analysis (SSA) is a validated method for detecting adverse drug events (ADEs).
• It leverages computerized claims data for pharmacovigilance and identifying unexpected medication side effects.
• SSA has demonstrated moderate sensitivity, high specificity, and robust performance in validation studies.

Purpose of the Study:

• To present the conceptual framework of Sequence Symmetry Analysis (SSA).
• To discuss the advantages and potential pitfalls of employing SSA in practical applications.
• To provide insights into interpreting SSA study results for ADE detection.

Main Methods:

• SSA analyzes sequences of medication initiations.
• A potential adverse effect of drug A is indicated if drug B is more frequently initiated after drug A than before.
• The method utilizes minimal datasets and is computationally efficient.

Main Results:

• SSA inherently controls for time-constant confounders.
• Potential biases include time-varying confounders and time trends in exposure or outcome events.
• Adjustments for trend effects can be made by modeling expected sequence ratios.

Conclusions:

• SSA is an efficient pharmacovigilance tool for detecting adverse drug events.
• Careful consideration of potential biases, such as time-varying confounders and trends, is crucial for valid interpretation.
• Minimizing false positive or negative results requires diligent assessment of study limitations.