The FDA’s Adverse Event Reporting System, which allows healthcare professionals, patients, and others to submit reports on adverse events, contains over 8.5 million entries, a massive number complicated by the fact that any drug may be listed under an average of 16 different names.
Postmarketing pharmacovigilance is critical in the effort to track and ultimately reduce the number of adverse events caused by drug treatment, but a new research article published in eLife indicates that the primary vehicle for such tracking in the United States has significant shortcomings.
The FDA’s Adverse Event Reporting System (FAERS), the primary source for US post-marketing pharmacovigilance, may be a “solid frame for reporting,” as the article’s authors say, but the database is “largely uncurated, unstandardized, and [lacking] a method for linking drugs to the chemical structures to their active ingredients, increasing noise and artefactual trends.”
FAERS, which allows healthcare professionals, patients, and others to submit reports on adverse events (AEs), contains over 8.5 million entries, a massive number complicated by the fact that any drug may be listed under an average of 16 different names, the authors report. Such redundancies, they write, can introduce false associations of AEs with specific product names.
Furthermore, duplicate reports can increase the apparent significance of an AE association for a drug, especially when the total number of reports is low; the researchers found that approximately 1% of entries in the database represent redundant reports. The fact that over half of all reports in the database are submitted by users who do not identify themselves as medical professionals (they may be patients or their attorneys) contributes to these redundancies, and can introduce errors. The authors point out that some reporters are confused about the difference between diseases that drugs were intended to treat versus AEs, and cite an example in which diabetes was listed as a side effect of a drug that was indicated to treat diabetes. In other cases, there exists a lack of clarity concerning AEs and patient outcomes.
The authors provide recommendations to help disentangle the FAERS database, including the following:
The article queries whether tracking fewer names for drugs, rather than more names, may improve pharmacovigilance, especially when AEs are predominantly reported by nonmedical professionals. The issue is of particular interest in the biosimilars context, as the FDA requires each biosimilar to bear a 4-letter suffix appended to its nonproprietary name. While many stakeholders believe that this system will allow for better tracking of AEs by biosimilar product, the FAERS findings suggests that additional differentiation among names could create more confusion and “noise” in reports.
While patient reports may be a primary cause of confusion or misreported AEs in FAERS, healthcare providers, too, may have difficulty in reporting events that correspond to the appropriate biosimilar suffix. As Allan Gibofsky, MD, told The Center for Biosimilars®, “We’re going to write a prescription for infliximab, and we may not always know what the 4-letter suffix is that the patient is getting. “When you prescribe aspirin,” Gibofsky went on, “you don’t know if the patient is getting Bayer or St. Joseph’s or Costco or Walmart. You know that they’re getting aspirin, but you don’t know what brand they’re getting. Similarly, we have a way of tracking which brand of infliximab the patient may be getting from the last 4 letters of the suffix. But, at the time that we’re writing for it, we may not know what’s actually being received.”
While healthcare providers may be conflicted about the proliferation of suffixes for biosimilar products, CMS has begun to differentiate biosimilars further, if only in its billing practices. In August, CMS announced a new modifier for biosimilar infliximab products; Merck and Samsung’s Renflexis, a newly launched biosimilar, will now be reported using an additional modifier, ZC, appended to its billing code. Pfizer and Hospira’s Inflectra will be reported with its existing modifier, ZB. This change follows a period of public comment in which some stakeholders said that grouping all biosimilars under a single billing code could, among other problems, introduce problems with properly tracking AEs to particular products.