Screen Shot 2017-04-06 at 11.43.02 AM.pngIt has now been one full quarter since we launched Evidex Alerts, the service which keeps our clients on top of the latest Drug Label Changes, Regulatory Alerts and Clinical Safety Data in real-time.   Evidex Alerts have been a smashing success and our clients have openly thanked us making this information easier to follow.  One client in particular was amazed at how well we untangle the mess that is a drug label and pull out the important pieces of information that they need to know, when they need to know it.

While going through the process of tracking and deciphering these label changes the last few months, we have learned some interesting things about drug labels that I have summarized below. (Yes, a drug label can be interesting!) 

Evidex Predicts Label Changes

So far this year, we have created 27 Evidex Alerts covering label changes with new safety risks.   Of those 27 Evidex Alerts with new adverse events, disproportionality analysis predicted 26 of those added adverse events, or 96%.   Some of the active RxSignals for serious risks were triggered three years prior to being added to the label.

We always knew our analysis was predictive of label changes (with back-testing and methodology published last year in Drug Safety), but it is satisfying to see those predictions come true in real-time.

Clinical Trial Stats are Becoming More Prominent in Labeling

Historically, serious risks have often been reported on the warnings section of a label without getting into specifics on how often they occur.   For example, the label might say the risk is rare, or extremely rare, but not provide additional data on specific rates.   Through the course of the year, we have seen more and more labels provide specific data.

For instance, in a label change from February 22nd covering Revlimid and its new maintenance therapy indication for multiple myeloma, an estimated risk of secondary primary malignancies (SPM) (including Acute Myeloid Leukemia (AML)) was disclosed for the first time in the warnings section of the label.  In patients with newly diagnosed multiple myeloma, an increase in SPM was shown in 5.3% of patients receiving Revlimid plus melphalan compared with 1.3% of patients receiving melphalan alone. Instead of just disclosing that the risk was “increased”, the warnings section of the label details out that approximately 4% more of these patients in clinical trials experience these malignancies.  

This is extremely useful from both a patient safety perspective and a cost of care perspective.

Payers, prescribers, and patients now can estimate the specific likelihood of experiencing the adverse event. Furthermore, that likelihood can be translated into an estimate of downstream medical cost.  According to AHRQ’s HCUP survey, AML costs $50,620 to treat. If an additional 4% of patients experience this event when taking Revlimid in a specific patient group, that equates to $2,000 in downstream costs each time Revlimid is prescribed to a new patient.

This overall trend of more detailed information from clinical trial results fits well with the clinical evidence we are curating and providing analytics on.    Users of Evidex have access to established clinical trial data that on-demand puts into context any new adverse event that is reported in the real-world population.    This helps users better validate and evaluate a signal and highlights any increases in risk that are occurring in the real-world setting that were not showing up in controlled homogenous clinical trials.

Other AEs not publicly disclosed, but important to FDA

Prior to digging deep into label changes, it was my understanding that drug labels simply reflected what has already been publicly disclosed, either in trial registries such as or other published research.   What we have found however, is that often this is not the case.

In many studies, the non-serious adverse events from the trial are not reported below a certain threshold (for instance 5% or 10%), and they may not be reported if they didn’t occur at higher rate than the comparator in the study.      

An interesting example of this data disparity can be seen with the recent label change covering the expanded approval of Technivie for Hepatitis C patients with compensated cirrhosis, where certain adverse event rates were disclosed for the first time.

Prior to this label change, non-serious Adverse Events were reported in the Phase III AGATE-1 study with a cutoff at 10%. When this most recent label change occurred, the risks that occurred below the 10% cutoff were disclosed for the first time.  The AE additions included a notable risk of certain cardiac disorders with a reporting rate of 9%, just below the 10% reporting threshold from the clinical trial. Although the underlying compensated cirrhosis or combination with ribavirin could be playing a role in this cardiac risk, this is vital information that healthcare decision makers would have liked to have seen previously.

Luckily for Evidex clients, some of the cardiac disorders that were reported in this label change had an Active RxSignal status, warning of these risks prior to the label change.


Overall, the process of digging deep into labels and providing context from disparate data sources in Evidex Alerts has been fascinating. We’ll keep providing updates on our predictive value, highlighting specific AE rates found in the clinical evidence dataset, and uncovering hidden safety concerns that are not generally prominent. 

Contact us to learn more about Evidex Alerts or to start a conversation.


Bob Kyle, Chief Product Officer


Topics: FAERS, Clinical Evidence

Bob Kyle

Written by Bob Kyle