I enjoy reading FDA Warning Letters. They’re insightful and often entertaining. Sometimes, though, they’re kind of unbelievable. And, sometimes, they’re downright nauseating.
Take, for example, this letter issued on February 19, 2016 to a principal investigator (PI) based in New York. FDA inspectors found him to be in violation of 21 CFR Part 312.62(b), related to maintaining complete and accurate case histories for clinical subjects, and 21 CFR Part 312.60, related to executing the study in accordance with the protocol.
In the case histories for the subjects enrolled at his site, the PI attributed the performance of study procedures to his sub-investigator (Sub-I)…when the Sub-I was not the one who performed them. Sometimes, the PI hand-wrote notes that said the Sub-I conducted the procedures, when they were actually performed by the PI himself or by other study employees (49 instances). Other times, the Sub-I signed off as having performed the procedures when someone else actually performed them (15 instances).
In his response to the Form FDA 483, the PI’s lawyer noted that 1) the investigational plan didn’t specify who was responsible for performing the study procedures, 2) after the study staff collected the information to complete the study records, the PI and the Sub-I carefully reviewed and discussed them, and 3) the records were not “final documents.”
I don’t know about you, but I’m not sure how any of that justifies falsifying records.
That’s the unbelievable part. Here comes the nauseating part: The lawyer went on to say that, while the PI disagreed with the FDA that he had done something wrong, if he actually had, it wasn’t his fault. His employer should be held responsible.
The FDA reminded the PI that, when he signed the Form FDA 1572 for the study, he agreed to become responsible for the execution of that study at his site.
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.
When determining whether patients were eligible to participate in the study, the PI ignored multiple inclusion/exclusion criteria.
For example, the protocol specified that a specific psychiatric disorder had to be the primary diagnosis for the patient, and that the primary diagnosis had to be confirmed by completing the Structured Clinical Interview for DSM-IV, Research Version, Patient Edition (SCID-RV/P). For all but one of the subjects enrolled in the site, the SCID-RV/P section of the screening visit questionnaire was not completed.
That’s kind of like testing a cancer drug on patients who might not actually have cancer. It’s a miracle! Your cancer is cured! It’s almost as if you never had it! Oh, wait…
On top of that, for the subjects for whom he did complete the SCID-RV/P (one in part and one in full), he used the wrong version.
In response, the PI’s lawyer indicated that the study staff used information other than the required SCID-RV/P to confirm the primary diagnosis, including the fact that they knew all of the subjects from previous studies. (warning: nauseating part) The lawyer went on to say that the PI didn’t know the SCID-RV/P assessments weren’t completed properly, and that his study staff or the sponsor monitors should have told him.
The FDA reminded the PI (again) that, when he signed the Form FDA 1572 for the study, he agreed to become responsible for the execution of that study at his site.
But, wait – there’s more! Another failure to adhere to the inclusion/exclusion criteria: The protocol specified that patients who previously participated in a clinical study could only enroll in this study if there were at least a 30-day gap or a 5-half-life gap (whichever was longer) between the end of the previous trial and the start of this one. The PI enrolled a patient who had been involved in a previous study just 15 days prior, and had been dosed with an investigational drug just 16 days prior.
In response, the PI’s lawyer said that, even though the subject didn’t meet the criteria, he/she was monitored closely and didn’t experience any adverse events in either study.
Once again, I don’t really see how the lack of adverse events justifies ignoring eligibility criteria.
But, wait – there’s more! The protocol also required a particular assessment to be performed 24 hours after the dose at Visit 2, which the PI failed to do for at least three subjects. In response, the PI’s lawyer stated that, because the assessment was required twice in a short time period (4 hours post-dose and 24 hours post-dose for Visit 2), the staff and subjects “…were instructed to rate these timeframes [sic] in their assessments.” Not only does that not make sense, it doesn’t justify deviating from the protocol.
The FDA reminded the PI and his lawyer that, “The eligibility criteria for each clinical investigation are designed to optimize the interpretability of the collected data and to minimize foreseeable harm to enrolled subjects. Enrollment of subjects who do not meet the eligibility criteria jeopardizes subject safety and welfare, and raises concerns about the validity and integrity of the data collected at your site.”
On a final note, the FDA wrote, “Your lack of supervision and oversight of the clinical studies raises significant concerns about the adequacy of your protection of study subjects enrolled at your site in the studies mentioned above, and also raises concerns about the integrity of the data generated at your site.”
Sponsors, don’t let this happen to you!