In the Q3 2014 FDA Adverse Event Reporting System (FAERS) data file, FDA quietly added a new field that links a patient and case report to a literature reference. Or in other words, there is additional information on the adverse event (AE) report that previously was not available. These literature references can provide beneficial context to the patient case report, validate or refute a causal relationship to an adverse event, and often times bring to the surface data that is otherwise undiscoverable.In order to understand how these literature references can be used, it is important to understand exactly what they are. A case containing a literature reference is often the result of an AE being identified by the manufacturer during a literature review, and as required by FDA regulation, the manufacturer must document that AE and report it to FAERS.
For instance, in March of this year, our proprietary methodology triggered an active RxSignal for Hepatitis B reactivation for Olysio and Daklinza, and Watchlist RxSignal for Viekira Pak, Harvoni, and Sovaldi, direct acting antiviral medicines that treat Hepatitis C. These signals were further enhanced by five literature references for Olysio warning about this potential risk and showing serious consequences for the patients referenced such as liver transplantation. The FDA acted on this risk on October 6th, by requiring a black boxed warning detailing this hepatitis B reactivation risk for all direct acting antiviral Hepatitis C medicines.
In addition to enhancing our process of finding unlabeled risks, literature references also allow an astute FAERS observer (at Advera Health, we call them pharmacovigilantes) to identify an AE that was not specifically reported in a clinicalstudy because it fell under a 5% occurrence threshold. This reporting gap was explained in a previous blog post Uncovering Hidden Data Points with an Evidence Aggregation Platform. Our team of analysts have gone through throusands of raw literature references in FAERS and have found their matching PubMed and ClinicalTrials.gov reference. With this extra step of analysis, FAERS cases can be tied to clinical studies and can provide a more robust picture of the adverse events reported in those trials. . These key data points can be used by various healthcare decision makers. Managed care organizations can use these specific case reports to dig deeper into issues that they are seeing in their own data or to help validate an Active RxSignal that is seen on the Evidex platform. Health systems and hospitals can further support evidence of a drug induced condition. Pharmaceutical companies can keep closer tabs on their competition and generate new ideas on product differentiation and evidence generation strategies.
That said, none of these data are actionable without a way to properly access and contextualize the cases themselves. To support this unmet need, Advera Health has released a new, one page summary of all of the evidence that is known around a drug-AE pair.
As you can see in this screen shot, the drug-AE analysis shows all of the various data points that the Evidex platform provides, aggregated into one snapshot, including real world data from FAERS and clinical trial results data. Literature references are clearly documented and sourced back, with links to the PubMed abstract.
So, for the case of Invokana and Ketoacidosis, you can easily determine these key takeaways:
- Advera Health’s RxSignal was triggered in 2014, over a year before the FDA added the AE to the label of the SGLT-2’s.
- The reporting rate in FAERS is 0.0978%, very close to the 0.1227% rate that was seen across all Invokana clinical trials vs. a control rate of 0.0197%.
- 73% of patients that reported ketoacidosis as an AE with Invokana, were hospitalized as a result of the AE
- There are 16 distinct cases, across 18 patients that are linked to a literature reference. Those references include manuscripts published in the journals Diabetes Care, Endocrine Practice, Hospital Pharmacy, The American Journal of Emergency Medicine, the Journal of Diabetes, and even the Pakistan Journal of Medical Sciences.
With this amount of data easily accessible, researching a drug-AE pair is a much simpler task. When paired with powerful signaling analytics like RxSignal and harm cost algorithms like RxCost, these data become truly actionable.
See more on researching drug-AE pairs with powerful signaling analytics in our comprehensive guide to drug safety data.
Bob Kyle, Chief Product Officer