Disproportionality Analysis

The CIOMS Working Group VIII report on Practical Aspects of Signal Detection in Pharmacovigilance (2010) suggests that disproportionality analysis complements traditional signal detection approaches. It says the interpretation of such analyses "should take place within the context of other safety data derived from relevant sources." FAERS data (published quarterly by the FDA) has been referred to in many articles as one of the best sources to perform this analysis because it is the broadest, most complete, and most widely used data set for safety analysis. It is a challenging task for every company to map the drugs and adverse events every quarter, considering the dictionary upgrades, follow-up versions, and tools required to perform all these activities.

Two key challenges in doing robust disproportional analysis are:

1. Comparator selection: Do you want to compare your drug to all other drugs in the FAERS database or only to a specific subset of drugs? Once the drugs and adverse events are identified and mapped, defining whether to use all FAERS data for comparison or a specific subset of drugs or class of drugs adds additional complexity. Usually, WHO UMC's Drug Dictionary and MedDRA are used in the mapping process. Taking advantage of the WHO Drug Dictionary's ATC to group the drugs gives us an added advantage to analyze the drugs within their own therapeutic class. As drugs are metabolized by different organs of our body, comparing the adverse events within a drug’s therapeutic class helps to focus the analysis on the most relevant patient populations.

2. Data preparation: The two key data elements for a disproportional analysis of FAERS data are the drug and the adverse event. FAERS data includes the drug’s trade name and active ingredient as well as the MedDRA preferred term (PT) for the adverse event. The drug and the adverse events are mapped to the latest version of the WHO Drug Dictionary and MedDRA Dictionary respectively. The MedDRA mapping is straightforward, but the WHO Drug Dictionary mapping can be very complicated for a number of reasons including missing active ingredients, variations in trade name (e.g. some with and some without dosage, formulation, route of administration, and company names). To address some of these challenges we map drugs to the active moiety (which is common across multiple active ingredients, trade names, salts, esters, etc). Using this approach we have been able to map more than 99% of drugs and 100% of adverse events within the FAERS database. With this mapping complete, you can leverage the inherent hierarchical structure within the WHO Drug Dictionary and MedDRA to calculate disproportional statistics at each ATC level.

AERion Analytics DiscoverAE platform is a sophisticated tool built specifically to perform disproportional analysis at any WHO Drug Dictionary ATC level. The FAERS data is curated and updated quarterly with the latest versions of MedDRA and WHO Drug Dictionaries. It leverages TIBCO Spotfire's world-class visualization features to present the data so safety scientists can start their analysis immediately. The tabulations, treemaps, and charts help safety scientists quickly and effectively move from data to insights.