FDA Warns Investigator Over Trial Drug Accountability

A Bioresearch Monitoring Program inspection of a clinical trial site in Homestead, Florida has produced one of FDA's more pointed data-integrity findings this year. In a Warning Letter dated 27 July 2026, the agency's Center for Drug Evaluation and Research told clinical investigator Maria E. De La Torre Silva, M.D. that dosing logs, the site's own drug-accountability records, and its electronic data capture (EDC) system did not agree for six enrolled subjects — and that some investigational drug kits logged as dispensed and used came back with their tamper-evident seals still intact.

A BIMO inspection, not a for-cause raid

Nothing about the origin of this letter is unusual. FDA's Bioresearch Monitoring Program routinely inspects clinical investigator sites as part of its ordinary oversight of trials supporting marketing applications, independent of any complaint or reported problem. What makes this letter worth reading is how quickly a routine Good Clinical Practice inspection escalated into a question of whether the data collected at the site can be trusted at all. FDA's own framing draws that line explicitly: the letter treats the drug-accountability discrepancies not as a documentation gap to be corrected going forward, but as a reason to doubt “the validity and integrity of the data collected at the site.”

The overlapping-enrollment finding

FDA's primary citation, under 21 CFR 312.60, is that Dr. De La Torre Silva failed to ensure the investigation was conducted according to its approved investigational plan. The specific example FDA cites is stark: one subject was enrolled in two separate clinical trials at the same time, receiving investigational drug from both studies during overlapping periods. Concurrent enrollment defeats the basic premise of a controlled trial — that an observed effect can be attributed to a single investigational product under a single protocol — and it is the kind of finding that calls a site's entire enrollment-screening process into question, not just the one subject's record.

Sealed kits, logged as used

The accountability finding is where the letter turns from a protocol-conduct problem into a data-integrity one. For six enrolled subjects, FDA found that self-reported dosing did not match the site's own drug-accountability records, which in turn did not match what was entered into the trial's EDC system. Investigational product kits that the site's records showed as dispensed to, and used by, subjects were returned to the site with their tamper-evident seals still intact — physical evidence, in FDA's reading, that the kits were never opened at all. Three independent records disagreeing about the same doses is a monitoring failure; kits coming back sealed after being logged as consumed is evidence the underlying events may not have happened as documented.

What the sponsor did

FDA's letter states that the consequences were immediate and specific: the study at the site was closed, the sponsor agreed not to use the site's data in its submission, and funds paid to the site were returned. That is a sponsor treating the site's output as unsalvageable rather than correctable — a stronger response than a corrective action plan or re-monitoring, and consistent with how FDA frames the underlying finding. The letter does not name the sponsor or the trial.