A systematic approach to 'cleaning' of drug name records data in the FAERS database: a case report

Data 'cleaning', also known as data 'cleansing', or data 'curation' is about identifying and rectifying errors in data. The objective of this report is to present a data cleaning and standardisation process for the drug name files in the U.S. Food and Drug Administration adverse event reporting system database, FAERS. Drug name data was cleaned and standardised using a combination of data cleaning tools and manual correction techniques. Data files were organised into frequency intervals and a strategy of cleaning using iteration and programming scripts in the MySQL Workbench was employed. The download of the FAERS quarterly reports for the time periods ranging from Q1 2004 to Q3 2016 resulted in 32,736,657 DRUG file records. Records contained a variety of errors, such as misspellings, abbreviations and non-descript or ambiguous names. Upon completion of the process, standardisation of greater than 95% of the drug name data in the FAERS database was achieved. With large datasets such as FAERS, a cleaning process is necessary to rectify data that may be incomplete or inaccurate due to input errors, in order to improve the quality and validity of information.