NAVIGATION
Clinical data management is generally understood to mean a process of implementing a reliable method for collecting, processing, managing, and conveying data in a manner that protects subjects’ rights, and insures data integrity. Historically, regulations intended to ensure the integrity of clinical data are contained within 21 CFR 312.
The ability to falsify data is far greater in clinical studies than in routine production, due to the nature of this data:
• access to it is far more open than post market data
• the interpretation of the meaning of data is common, as opposed to the clear
– measurements against specified limits taken during routine production
Falsification of data includes creating, altering, recording, or omitting data in such a way that the data no longer represents what actually occurred.
Examples of falsification of data include but are not limited to:
• creation of data points that were never obtained
• alteration of data points that were obtained by substituting different data
• recording or obtaining data from a specimen, sample or test whose origin is not
– accurately described or in a way that does not accurately reflect the data
• omission of data points that were obtained and ordinarily would have been recorded
Falsification of clinical data can have extremely dangerous consequences, including placing all subjects in that trial at possible safety risk. Falsification also jeopardizes the reliability of data submitted and/or published and undermines FDA’s mission to protect and promote the public health.
The criticality of clinical data integrity was well known to the regulating body and the regulations contained within 12 CFR 312.62, .64, .66 and .68 were developed in an attempt to mitigate that risk. The extent to which these regulations can protect the integrity of data is limited when compared to similar regulations for post market data.
Post market data management regulations and practices were designed to promote the integrity of data that can be specified and predicted. These types of regulations are designed to control and demonstrate integrity as well as accuracy. However, clinical data is generated during research phases, and therefore cannot be specified. This type of data is generated to inform, rather than to confirm.
The regulations concerning the management of clinical data were designed to focus on confirmation of accuracy, and not on integrity. It is critical to note that the regulators were aware of the inability to predict data at this point in research and therefore regulations designed to associate legal responsibility for the integrity of clinical data were developed and are contained within 21 CFR 312.70.
The scope of 312.70 is limited to Clinical Investigators, formally assigning legal responsibility to those individuals most directly responsible for the generation and protection of clinical data.
The authority granted in 312.70 however, could only be utilized (triggering penal enforcement actions) if falsification was discovered. The process of approving new drug applications and investigational applications includes a data review by the FDA. In post market situations, FDA can use past experiences with the manufacturer and the drug/device to target inspections in such a way as to pre-define high risk areas.
As every data set submitted during the investigational phase is new to the agency, utilizing historic experiences with the product was not possible, so another mechanism for determining high risk areas needed to be developed. Toward that end, regulators decided to determine risk by motivation of the researchers.
As such, the clinical regulations were supplemented by new regulations in 1999 that were intended to ensure that any financial interests of clinical investigators that might have the ability to affect the reliability of data were classified and fully disclosed to the government.
This expansion of the regulations was considered necessary as many forms of compensation provided to clinical investigators at that time created a situation that could incentivize the falsification of data.
These types of situations included:
• Compensation made to the clinical investigator in which the value of the
– compensation could be affected by the study outcome
• a proprietary interest by the investigator in the tested product, such as a patent
• a significant equity interest in the sponsor of the covered study
• significant payments by the sponsor of the covered study of other sorts, such as
– a grant to fund ongoing research, compensation in the form of equipment, or
– retainers for ongoing consultation or honoraria
The general response to this action (communicated during public meetings) from the manufacturing and clinical populations voiced support for efforts to reduce the likelihood of any financial arrangements that had the potential to bias the quality of clinical data.
Under this rule, if any clinical investigator was compensated in any way that could lessen their objectivity, the scrutiny with which their data sets were reviewed would be increased.
In a continuing effort to reduce situations that could incentivize the falsification of clinical data, the FDA is now focusing on the role of drug and device sponsors. Toward that end, the FDA has recently submitted a proposed rule that requires the study sponsor to report any suspected falsification of clinical data within 45 days of issue identification. The FDA is hoping, as the sponsor is in a unique position to internally monitor data collection, that regulations formally requiring them to report their findings will contribute to the integrity of clinical data.
However, these regulations continue to deal with the issue after falsification has occurred. This in the opinion of these authors does not directly promote the quality of data generated. It does seek to lessen the motivation to falsify data.
Holding everyone involved responsible for any action that has the potential to harm the public is clearly fair and just. However, penalty options are not generally effective deterrents, if the financial incentive to falsify data remains greater than the potential financial penalty if discovered.
Evidence supporting this opinion is generated with almost every routine inspection of modern manufacturing plants. The financial incentive while a drug or device is in the clinical phase is as great, if not greater than a drug in the post marketing phase.
Financial motivation is clear.
The flexibility that can be built into a clinical study also provides opportunities for types of data falsification that is difficult to detect, and non-routine methods of reporting the data allow for further opportunity to mis-state or misuse data. The non-routine nature of the collected data, and the reporting techniques, also considerably amplify the review challenges presented to FDA.
Opportunity and means are also clear.
Consider the following as an example of the dangers that can be posed in light of the flexibility of clinical study data collection and interpretation mechanisms:
Company A manufacturers a drug that has been proven to increase suicidal thoughts in children. A fraud lawsuit against the company in 2004 claimed that three of five clinical trials verified that suicide-related behaviors were observed twice as often among young subjects given the active drug as those that were given the placebo. While an application for use for children was never filed; doctors have professional discretion to prescribe the drug for any group that is not contraindicated.
The suit claimed that the company delayed submitting the test results to government regulators for years and chose not to include the information contraindicating pediatric use in the “Medical Information Letters” sent to doctors. Without this warning, doctors prescribed the drug as they felt warranted. The company did not admit to guilt but later settled and agreed to pay more than $70 million to 6 states and to Medicaid.
This company is currently conducting a pediatric trial in Japan The nature of the trial design does not include a focus on links to suicidal behavior which will allow any observed link to suicide behaviors to also go unreported or under-emphasized.
Given this example, it is clear, that the clinical investigators are not the only players in the game that may be incentivized or have an opportunity to falsify or misuse clinical data. In this case, the investigator may have failed to highlight the findings, but the sponsor must have realized the significance of the data, as the resultant application did not include pediatric usage. The sponsor also chose to completely omit this information from subsequent communications with the medical community.
This example is not unique.
The greatest impact these proposed regulations could have is if they encourage sponsors to develop and implement robust data monitoring systems. If these systems are not developed and implemented, then the effect of this regulation will be limited to expansion of the scope of individuals and/or organizations that can be legally penalized once fraudulent activity has been discovered.
While we sincerely hope that the new reporting regulations encourage sponsors to establish and implement more robust collection and monitoring systems, we also believe that requiring the implementation of these enhanced systems is a step that should be strongly considered.
It is clear that the nature of clinical investigation will always require the ability to design unique protocols, and to interpretively report data, it also seems clear that building systems that can provide effective oversight in real time is also possible. If it is not practical for industry to provide this oversight in real time, perhaps the nature of the FDA’s involvement during clinical trials should be modified in order to allow more effective federal oversight.
We will continue to monitor progress in this area and report it to our readers.
Stay tuned…
© Coda Corp USA 2009. All rights reserved.
__________________________
Authors:
Corrine R. Knight and Gina Guido-Redden
Coda Corp USA
References
Financial Disclosures by Clinical Investigators: http://www.fda.gov/ScienceResearch/SpecialTopics/RunningClinicalTrials/ucm119145.htmReporting Information Regarding Falsification of Data: http://edocket.access.gpo.gov/2010/pdf/2010-3123.pdfhttp://industry.bnet.com/pharma/10008290/glaxo-is-testing-paxil-on-7-year-olds-despite-well-known-suicide-risks/http://www.fda.gov/RegulatoryInformation/Guidances/ucm126832.htm
http://www.consumeraffairs.com/news04/paxil_ny.html
