Data integrity (DI) has been a problem for as long as I remember, and my memory goes way back. The first real exposure I had to DI issues dates to the early days of Hatch-Waxman, passage of which had led to the generic drug scandal of the late 1980s. I could write a book about all that went on both at the FDA and in the industry, as well as the chilling impact that it had for years on trust and the review of applications. The FDA’s reaction to systemic problems internally placed the Agency in a very uncomfortable position, and externally placed a significant portion of the generic drug industry on the brink of collapse. So, with the publicity of that event back in the late 80s and early 90s, why do we still see a continuing number of issues with DI today?
During the virtual Small Business Industry Assistance Generic Drug Forum held April 26-27, 2022, there were seven separate presentations on DI. The speakers highlighted various issues related to DI and provided case studies to outline the types of issues that the Agency has uncovered. Presentations covered DI issues in bioequivalence studies, pharmacology/toxicology studies, clinical studies, and analytical studies, as well as DI in ANDA documentation. Think about it, seven presentations in one seminar about DI, which took up almost half of the second day of the forum. That tells me that DI issues are alive and well in the industry and that the FDA has again turned its focus on identifying and weeding out DI events, and fixing the problems that DI issues present to the Agency.
Just to give you a general sense of the content discussed, here is a summary from one general slide about the types of DI issues seen in the various categories of applications and development disciplines.
- Unethical conduct for inclusion of study subjects
- Multiple peaks with large fluctuations from bioanalysis assigned to improper study conduct
- Failure to report AEs and SAEs, including deaths
- Multiple subjects dosed prior to screening them for eligibility in the study
BE Studies (Bioanalysis and Statistics):
- Discrepancies and manipulated values in statistical analysis
- No investigation for frequent run failures, large IS variability, consecutive QC failures, incorrectly rejected runs
- Sample substitution and deliberate misinformation
- Same data for different species/study
- Repeat data unjustified
- Biological implausible results
- Missing information, missing data, incomplete methods or results
- Fabricated data
- Uncontrolled documentation
- Invalidated OOS results without justification
Some of the case studies brought back chilling memories of the Generic Drug Scandal days; that chill should remain and shake up firms to look at the areas of DI that could impact their organizations. If you have a chance, please look at the slides from the Generic Drug Forum as some great information was relayed to the industry. You may be able to download the slides here. Be proactive not reactive, especially when it comes to DI issues.