Check your sources

Do you know where your safety data come from? The answer is not as simple as you may think.

There are two sources of safety data in clinical studies – the clinical database and the safety database. Many writers and clinical developers have tied themselves in knots confusing these two sources.

The clinical database collects adverse events (AEs) and other clinical assessment data (eg, vital signs, laboratory results, ECG parameters) specific to each study protocol. Data are entered by the clinical sites using case report forms (CRFs) which contain specific and predefined fields. These data fields may include checkboxes or drop-down lists of pre-determined values or free-text fields. The clinical database is the primary source for all programmed tables, listings and figures supporting all regulatory documents and publications. The clinical database is the official database of record for each clinical study.

The safety database is designed to collect ongoing safety data for a specific medicinal compound or device inclusive of all studies across all stages of development. As opposed to the study-specific clinical database, the safety database collects significant safety information (eg, serious adverse events) for reporting to regulatory authorities in an ongoing manner to ensure patient safety and is the key database for pharmacovigilance activities. The scope and reporting processes are specified in protocols but involve industry-standard forms on which safety events are reported narratively as free-text. These forms can be updated serially as more data become available or in response to queries from the sponsor. The narratives may be sparse or voluminous, concise or messy, depending upon how the study site collect and describe the information. Much of the data in the safety database remains outside the clinical database, such as detailed information from hospital visits performed outside the predefined study visits.     

In a perfect world, the common data elements between clinical database and the safety database match perfectly; however, since the data are added at different times by different people, discordances are inevitable despite attempts to reconcile. Learning to recognize these two data sources, including their purposes and limitations, will help you to proactively identify where the discrepancies might lie, and which data to include in your documents.

One simple tip – when the clinical database and the safety database disagree on a key data element, the clinical database data should be used.  

Dr. Virginia Kelly

Co-Founder and CMO, Acumen

Virginia Kelly

CMO and Founder of Acumen Medical Communications, Virginia is a board-certified hematologist/oncologist, Virginia provides strategic medical and clinical development support.

https://acumenmedcom.com
Previous
Previous

NDA/BLA Submission Tips

Next
Next

How to Write CSR Conclusions