| Increased attention is being paid to the internal controls used to
insure scientific integrity, safety and protection of subjects in clinical
trials. This is in part a result of revelations within the past few years
of falsified data in a breast cancer trial sponsored by the National Cancer
Institute. Although those incidents do not seem to have invalidated the
results of the study, the General Accounting Office, at the request of
Congress, conducted
an inquiry into how quality is maintained in NIH-sponsored trials.
In its investigation the GAO found that while there are federal and
grantee-level mechanisms for dealing with misconduct when it is discovered,
there is no established NIH-wide methodology built into trial design to
insure scientific quality and integrity. Currently some Institutes do employ
such strategies as funding data verification activities that are independent
of research investigators. However, the feeling within NIH is that mandating
such methods agency-wide would not be appropriate for the variety of trials
sponsored by the different Institutes.
Nevertheless specific Institutes are formulating their own policies.
On October 4, the National Institute on Aging issued a
detailed set of guidelines for conducting Phase III clinical trials
under its sponsorship. These guidelines will impose additional requirements
on research investigators in proposing and carrying out their studies.
For instance, investigators will have to provide explicit protocols for
quality control and records audits, and set up data analysis functions
that are independent of the principal investigator.
The following is an excerpt from that announcement:
Requirements for Investigators
The proposed trial must include:
- A data processing and analysis unit or function administered by a
designated individual other than the Principal Investigator(s) of the trial.
(In the case of multiple-award studies, the designated individual may be
the Principal Investigator of an administratively separate award). This
individual may report to the Principal Investigator, or, in multiple-award
studies, may report to the steering committee, of which the Program Administrator
is a member.
- An independent Data and Safety Monitoring Board (DSMB). The application
should indicate the proposed frequency of meetings for the DSMB, and include
a proposed list of data items to be provided to the DSMB and estimates
for DSMB-related expenses in the proposed project budget.
- Protocols for quality assurance/quality control, data management,
and analysis.
- A structured adverse event determination, monitoring and reporting
system, including standardized forms and protocols for referring and/or
treating subjects experiencing adverse events. The proposed schedule for
reporting adverse events to the Data and Safety Monitoring Board and Program
Administrator should be described.
- Specific criteria and procedures for unmasking (if the study is masked).
- Plans for independent auditing of primary subject records over the
course of the trial to verify conformance with eligibility criteria, recruitment
and outcome data, adverse events, etc., and reporting discrepancies to
the Principal Investigator, the DSMB, and the Program Administrator. These
plans should provide for reporting serious discrepancies within two weeks
of their discovery.
- A detailed Manual of Procedures. This is not required in the application
submitted to peer review. If an award is made, the applicant should
prepare the Manual of Procedures for review by the Data and Safety Monitoring
Board and Program Administrator, prior to implementation of the trial.
- Plans for notifying subjects of trial results after the conclusion
of the trial.
While these new standards currently apply only to NIA-sponsored
trials, other Institutes may well follow suit. This means that investigators
interested in proposing Phase III clinical trials would be well advised
to become proficient in designing and carrying out these types of internal
controls.
|
