Just recently, on January 12, 2022, the co-owner of a clinical research site, Olga Torres, pleaded guilty to a charge of obstructing an FDA investigation into falsified clinical trial data. Her firm, Unlimited Medical Research, was among several hired to conduct a clinical trial to assess the safety and efficacy of an asthma medication in children.
This case is alarming and a cause for concern, but it is, unfortunately, not unprecedented, occurring in a wide range of disciplines. In cases involving medical science, and specifically drug or treatment approval, the consequences of failure to detect such fraud can, in the worst cases, lead to damage to public health, and more generally to a reduction in public trust in the scientific establishment.
As a point of hope, there is evidence that editorial oversight is improving, leading to increased detection and retraction of offending research. Moreover, researchers are developing improved methods for the detection of falsified data using computational and/or unsupervised techniques. Beyond this, however, the worry lies in that even once detected, with the offending papers retracted, fraudulent or inaccurate science can continue to influence the course of research, as in the phenomenon of zombie papers, where retracted papers, and more importantly, papers that cite those retracted papers, continue to be cited at high frequency.
These problems are not new to the broader field of scientific research. And, it seems clear now that attention needs to be paid to the prevention of fraudulent research in addition to its detection and subsequent retraction. Incentive structures may need to be changed to achieve this, in order to stymy the persistent effects of bad science from beyond the grave.
Click here for the Japanese version.