| Comment Record|
Dr. Michael Gorin ||
2003-04-29 12:44:18 |
University of Pittsburgh |
| Comments for FDA General |
1. General Comments
There is always the potential for an investigator to have a financial interest in the outcome of a research project. Such interest may be direct such as payments from a company, but could also be as subtle as ongoing NIH funding for renewal of the research or funding of new grant applications. Identifying potential financial interests is basically a waste of time and essentially implies that some financial interests are more threatening than others when in fact, it is the actions of the investigator (not the motivating situations) that determine the potential conflicts. It is actually less problematic to approach the problem differently.
Rather than trying to define the nature of the financial interest, the emphasis should be on:
1) how the potential outcome of the research might be affected by any potential or preexisting financial interest
2) If such a potential influence may be present, what steps should be taken to ensure that the results and interpretation of the results are as objective and credible as possible.
The first objective must be focused on identifying potential mechanisms by which the research (and more specifically, patient welfare) may be affected by a financial interest.
a) Recruitment of subjects must be considered because of the impact of meeting recruitment goals on the well-being of the investigator?s current and future activities. Even without a classical financial interest, investigators? have a strong motivation to maximize subject recruitment.
? Clear delineation of inclusion and exclusion requirements ? this must be a requirement of every clinical research proposal.
? We have to carefully consider how to avoid the use of intentionally fraudulent data (unintentional fraudulent data can be minimized with computerized data entry).
The best approach is to not make a single person responsible for gathering and storing the data for inclusion and exclusion. If the PI is responsible for determining if the subject is eligible, then the data for eligibility should be collected and verified by another person (who clearly identifies themselves in the records). If the eligibility criteria are determined by a central coordinating center, the clinical data must indicate a primary source for external validation. For example there should be a mechanism for the coordinating center to contact the medical records of a subject directly for radiologic data or primary laboratory data. This can be done through an honest broker system within the clinical system to preserve confidentiality. If it is impractical to do this for every case, then a randomized audit system should be instituted and administered by the research organization.
b) Subject monitoring (concerns regarding exclusions and inclusions)
The problem most commonly is the failure to report relevant new exclusion data. This has been seen in several prominent investigations of clinical trials. There are two potential sources of data: 1) that gathered as part of the study (see above), and 2) data gathered outside of the study (such as by unrelated medical evaluations or from the patient). For the later, one should consider patient-oriented audits and again, honest broker?driven audits. Data safety monitoring committees are essential for this subject protection but there needs to be clear guidelines as to how these committees obtain critical information and carryout their mandates.
After addressing patient safety issues, it is essential to also consider the matter of research integrity. The public, the research community and the funding agencies have an enormous investment in clinical studies and the integrity of the data and analyses are crucial. Thus safeguards are necessary to avoid degrading research integrity because of potential (not necessarily real) financial conflicts of interest.
Research Integrity encompasses the following components:
a) Analysis of raw data and data integrity ? compliance, avoidance of tampering, completeness, confidentiality
b) Statistical analyses
c) Internal consistency checks
d) Analysis of results and conclusions
Some of these issues are actually covered by the peer-review process for publication and by the review processes used by the FDA. However rather than creating more onerous barriers, we should be facilitating the means by which the reviewers have access to primary data. In many cases, this is challenging because the data sets are huge and complex and reviewers have neither the time (nor the resources and expertise) to carryout audits and primary analyses. Traditionally, good publications and reports specifically address the limitations of a study and potentially confounding factors that might influence the outcomes and/or interpretations. Such ?self-critical? assessments are key and an important part of the investigators ensuring that their conclusions are well-considered and valid. I would recommend that this be somewhat more standardized as part of the reporting process. Investigators should have be strongly encouraged to include as part of their research reports, a section devoted to their own critical analyses of the study and how they would propose to address potential weaknesses and limitations. We do this routinely in NIH grant applications. If these sections are inadequately addressed, the investigators may be unable to publish or have approval by the FDA. The investigators should be able to produce the raw data and analyses for up to 7 years after completion of the study.
Finally, it is important to recognize that we are rapidly approaching the point it which the burdens on research compliance, rules for research studies, documentation and oversight requirements are making it very difficult to conduct clinical research in this country. The result will not be better or more ethical clinical research, but less studies will be done with existing funding and more research will be done in other countries which will then have the benefits (and profits) from these efforts. It is already much easier to do genetic research outside of the United States because of the restrictions and we are essentially giving away the fruits of the Human Genome Project to other countries and non-American industries who can do the clinical studies to make the genetic code meaningful to human health. We need to strengthen the existing mechanisms for human protections and research oversight rather than create new bureaucratic hurdles. We need to foster practices that promote good and ethical research without assuming that investigators are instinctively attempting to commit fraud. Rather than assuming that I am attempting to deceive and mistreat my research subjects and trying to dupe the public with my results and conclusions, give me clearer guidelines as to how to do my research properly with appropriate safeguards (which must be funded and available by my institution) and make sure that my presentations of results have been critically assessed. Don?t block the public?s right to hear about new research results, but do not allow investigators to release results and conclusions to the public without prior peer review. Investigators should be aware that public dissemination of research findings without such a peer review will result in loss of current and future research funding. If the finding is so critical that it cannot await the standard pathway of publication, then the journals should provide an expedited path of review.