It’s been awhile since I’ve written about FDA warning letters, so I thought I would take a look at the letters issued throughout the year and see if anything egregious (or at least interesting) came up.
I was surprised to learn that the FDA issued letters to six separate clinical investigators, and I thought the reasons why might interest you. Here are the areas in which they fell short, along with some noteworthy highlights.
5/6 violated 21 CFR 312.60, General Responsibilities of Investigators, for failing to follow the protocol. The biggest issue was ignoring inclusion/exclusion criteria; all five letters cited enrolling patients who should have been excluded. The second biggest issue was skipping procedures that were required for patient safety, the worst of which was not performing chest x-rays on subjects enrolled in a lung-related trial when they reported worsening symptoms. There was also a citation for skipping an early termination visit and not indicating whether any adverse events were involved.
This guide analyzes how artificial intelligence – including machine learning – can be used by pharmaceutical and medical device companies to improve the clinical data review and cleansing process.
2/6 violated 21 CFR 312.62(a), Disposition of Drug, for not maintaining adequate records of study drug. One investigator couldn’t account for 25 units of unused study drug; he claimed the units were destroyed, but had no evidence or documentation of the destruction. The other investigator had a number of discrepancies in her records about how much drug was dispensed, taken, and returned to the site.
2/6 violated 21 CFR 312.62(c), Record Retention, for not keeping records for the required two years following either drug approval or overall study termination. One investigator shredded or otherwise destroyed (he thinks) all records related to a study two years after the sponsor terminated his site, clearly not understanding when the two year retention period actually begins. The other investigator “misplaced” 15 ECG tracings for 7 subjects and the completed C-SSRS for 1 subject.
1/6 violated 21 CFR 312.62(b), Case Histories, for not maintaining adequate and accurate records of observations and data. This particular investigator had files for several subjects that contained a critical form that had not been completed, along with a recorded diagnosis that cited the content of the (blank) form as the basis for the diagnosis.
So, what can we learn from all of this? The most obvious takeaway is to not work with these specific investigators! But, beyond that, here’s what I see:
- The process of determining eligibility needs an overhaul. It seems like it would be safest for patients if the human element were removed altogether, and the automated process were calibrated to err on the side of caution/exclusion. After a patient is excluded, the investigator could review the exclusion reasons and decide whether to lobby to have the patient included.
- Subject recruitment needs to be easier/faster. Perhaps if more tools were in place to support subject recruitment, investigators would feel less pressure to enroll patients who don’t meet eligibility criteria. These tools could involve the use of mobile apps, social media, and even artificial intelligence combing giant databases of patient data to find exactly the right fit.
We discuss these and other common issues in our guide, 7 Ways to Fix Your Relationship with Clinical Sites. If any of the content in this post is hitting home for you, it might just be worth a look.