Standardisation of the FAERS database: a systematic approach to manually recoding drug name variants.

PURPOSE

The US Food and Drug Administration Adverse Event Reporting System (FAERS), one of the world's largest spontaneous reporting systems, is difficult to use because of report duplication and a lack of standardisation in the recording of drug names. Unresolved data quality issues may distort statistical analyses, rendering the results difficult to interpret when detecting and monitoring adverse effects of pharmaceutical products. The aim of this study was to develop and implement a data cleaning protocol to identify and resolve drug nomenclature issues. The key 'data treatment' plan involved standardising drug names held in the FAERS database.

METHODS

Four million five hundred and six thousand five hundred and seventy-seven. Individual Safety Reports submitted to the FAERS between 1 January 2003 and 31 August 2012 were included for this study. OpenRefine was used to standardise drug name variants in the database such that they were consistent with international non-proprietary nomenclature defined by the World Health Organisation Anatomical Therapeutic Chemical classification. Drug variants where generic constituents could not be confidently determined, undecipherable drug names and non-medicinal products were retained verbatim.

RESULTS

After the standardisation process, more than 16 611 916 drug entries were cleaned to their relevant international non-proprietary name. The cleaned drug table comprised 71 858 drug name variants and includes both standardised and original terms. Ninety-nine per cent of drug names was standardised using this method.

CONCLUSIONS

The millions of reports enclosed in the FAERS contain valuable information that is of interest to pharmacovigilance, toxicology and post-marketing surveillance researchers. With the standardisation of the drug nomenclature, the database can be better utilised by research groups around the world.