FDA Flags Serious Data Falsification at a Chinese Medical Device Testing Facility

Accurate data is crucial for the FDA to evaluate the safety of medical devices properly. If the data is flawed or falsified, it raises risks for patients, including serious health issues like toxic reactions or organ failure. This problem could ultimately compromise public health, as the FDA relies on trustworthy data to assess risks associated with medical devices.

The FDA has raised serious concerns regarding the data integrity of a testing facility in China due to multiple studies conducted at the facility exhibiting signs of manipulated or fabricated data. This includes cytotoxicity studies showing identical results across different dates, sensitization studies with matching guinea pig weights across different groups, and a large animal safety study displaying implausible results in bleeding and weight gain in Bama pigs. The FDA concluded that obtaining such identical results is statistically improbable and that the data trends observed do not align with normal biological responses in healthy animals.

In an earlier letter released late last year, the FDA requested a detailed response from the testing facility within 30 days. They sought explanations for the data inconsistencies found, reasons why the facility failed to identify the anomalies, insights into their processes that allowed such issues to occur, information on other potentially affected studies, and justifications for why the identified data issues do not cast doubt on all past and future submissions to the FDA.

The facility responded on January 29, 2025, stating that they have initiated an internal investigation to find the root causes of the unreliable data and to understand why their quality controls failed. They emphasized their commitment to addressing the issues raised by the FDA and planned to involve third-party experts in their investigation.

Despite this commitment, the FDA maintained that the response did not adequately address their concerns. The facility did not provide explanations regarding the data anomalies or how their systems allowed for such discrepancies. They also failed to offer any insights into whether other studies conducted at their facility had similar data issues or any analysis of the implications of those studies. Furthermore, there was no effort to argue against the FDA’s claims that data had been copied or falsified.

The FDA concluded that the testing facility’s response did not alleviate concerns regarding the data generated. They stressed that due to these ongoing issues, the reliability and validity of all data produced by the facility remain questionable. They indicated that until the company resolves these issues satisfactorily, the agency would reject all study data from the facility.

The FDA also noted that the facility’s intentions to continue serving existing customers did not address their concerns, as the same issues could affect all studies conducted by the facility. The agency reminded them that they might conduct a future inspection to verify any corrective measures taken.

Lastly, although the facility requested a meeting with the FDA to discuss their progress, the FDA declined, stating that a meeting would not be useful until the facility had adequately addressed the concerns outlined in the GCL. The FDA remains firm that the unresolved issues undermine the trustworthiness of the data provided by the testing facility.

Such data integrity issues would be impossible for the sponsors of such studies to catch. The FDA is better placed to catch such fraud as it sees data from a diversity of applications to see troubling trends, as happened here.