The Database of Dietary Supplement Adverse Events (DDSAE)

The Database of Dietary Supplement Adverse Events (DDSAE) is a collection of dietary supplement - adverse event signals detected from the CFSAN Adverse Event Reporting System (CAERS). It contains adverse event signals detected for both dietary supplement products and ingredients. We used four signal detection methods:

1. Proportional Reporting Ratio (PRR)
2. Reporting Odds Ratio (ROR)
3. Bayesian Confidence Propogation Neural Network (BCPNN), in which the information component (IC) metric was estimated using 50,000 Monte Carlo simulations.
4. Gamma Poisson Shrinker (GPS)

Directory structure

The detected adverse events are within two subdirectories of data/:

• ingredient_adverse_events/: Detected dietary supplement ingredient - adverse event associations.
• product_adverse_events/: Detected dietary supplement product - adverse event associations.

Each of these subdirectories contains four CSV files containing the ranked pairs detected by each of the four methods.

File structure

All files contain the following columns:

• {ingredient,product}: The ingredient (under ingredient_adverse_events) or product (for product_adverse_events) in each detected association.
• adverse_event: The adverse event in each detected association. These are coded according to the Medical Data Dictionary for Regulatory Activities (MedDRA).
• count: The number of times this pair occurred in CAERS.

In addition to these three columns, each CSV file contains a fourth column corresponding to the detection metric associated with that signal detection method. The detection metrics are:

• PRR: PRR05, the lower bound of the 95% confidence interval. A signal was detected if the PRR05 >= 1.
• ROR: ROR05, the lower bound of the 95% confidence interval. A signal was detected if the ROR05 >= 1.
• BCPNN: IC025, the 2.5% quantile of the posterior distribution of IC. A signal was detected if the IC025 > 0.
• GPS: EB05, the lower bound of the 95% CI. A signal was detected if the EB05 >= 2.

Using the dataset

Use of this dataset is subject of the terms of the LICENSE.

If you use our dataset in your work, please cite the corresponding paper:

Vasilakes JA, Rizvi RF, Zhang J, Adam TJ, Zhang R. Detecting Signals of Dietary Supplement Adverse Events from the CFSAN Adverse Event Reporting System (CAERS). AMIA Jt Summits Transl Sci Proc. 2019;2019:258–266.