| October, 2000
The task of staying in compliance with federal regulations when you
do research can be daunting. Among the welter of acronyms the investigator
must recognize are "MPA" and SPA". These refer to assurances made between
institutions and the federal government, represented, for human research,
by the Office for Human Research Protection (OHRP, the regulators formerly
known as OPRR).
Readers of this Newsletter know that the Bureau maintains a Multiple
Project Assurance or MPA, which covers all the research carried out at
Bureau sites, or by Bureau employees when they carry out research as employees
or agents of the Bureau. This assurance is like a contract that spells
out how we will oversee our research to protect research participants.
With this assurance, affiliates of the Bureau can carry out federally funded
research with oversight from our local Institutional Review Boards (IRBs).
But what happens when a Bureau investigator collaborates with another
institution? Suppose you're carrying out a study which involves schools,
or clinics operated by another health system. Does the Bureau MPA apply?
If you're acting in your capacity as an employee or agent of the Bureau,
then yes, your activities are covered by our MPA (and must be approved
by one of our IRBs), even if they take place off-site. But the story
doesn't end there. If the research is federally funded, the other institutions
may also need to have or obtain assurances before they can participate
in the research.
With any federal research grant, if the research is not eligible for
an exemption from review, all the institutions judged to be "engaged" in
the research must have an assurance with OHRP . The responsibility
for making sure all collaborators have assurances rests in large part on
the institution that receives the grant. This institution bears ultimate
responsibility for protecting human subjects under the award, at its own
and at its collaborators' sites. Indeed, this institution must document
that all collaborating sites have assurances before the grant money is
released. If the collaborating sites don't have their own MPA, then a Single
Project Assurance (SPA) must be applied for. An SPA is a special
assurance, made for this study only, that details how local oversight of
human subject protection will take place.
So far, so good. But how far does the need for an assurance reach? If
a clinic site will only be involved to the extent of telling potential
participants about the existence of the study, and giving them a phone
number to call, does that clinic need an assurance? If a community
organization provides a space for holding focus groups, is it "engaged"
in the research? What about the hospital that releases a subject's
records after the subject has signed a release for this information?
The questions about who is or is not "engaged", and therefore who needs
an assurance, are so convoluted that a guidance letter was issued by the
former OPRR in 1999 (see http://ohrp.osophs.dhhs.gov/humansubjects/assurance/engage.htm)
to clarify the distinction between "engaged" and not. What follows is a
summary of the main points of this letter, to help you determine if your
federally funded research needs additional assurances:
An Institution is "Engaged In Research"...
Any institution which receives a direct
award from the Department of Health and Human Services to carry out non-exempt
human research is automatically considered to be "engaged". This applies
even where all activities involving human subjects are carried out by a
subcontractor or collaborator.
Additionally, an institutions is engaged
in research if, for research purposes, the institution's employees or agents1:
• Perform invasive or noninvasive
procedures with living individuals. Examples include drawing blood;
collecting other biological samples; dispensing drugs; or administering
other treatments.
• Intervene with by manipulate the environment
of living persons. Examples include controlling environmental
light, sound, or temperature; presenting sensory stimuli; orchestrating
environmental events or social interactions; making voice, digital, or
image recordings.
• Interact with living individuals.
Examples include engaging in protocol-dictated communication or interpersonal
contact; conducting research interviews; and obtaining informed consent.
(As detailed below, there are some other personal interactions -- involving
informing prospective subjects about research -- which don't constitute
"engagement" in research.)
• Release individually identifiable
private information, or permit investigators to obtain individually
identifiable private information, without subjects' explicit written permission.
Examples include releasing patient names to investigators to recruit them
for a study; or permitting investigators to record private
information from medical records in individually identifiable form. Exceptions
would be when the subject has given prior written permission for the release,
or some cases in which information is released to a State Health Department
(see below)
• Obtain, receive, or possess private
information that is individually identifiable (either directly or indirectly
through coding systems) for research purposes. An example of this type
of information would be information from medical records in an individually
identifiable form. Exceptions may be made when information is in
a form that doesn't permit identification of individuals. (see below)
• Maintain a "statistical center" for
multi-site collaborative research that obtains, receives, or possesses
private information that is individually identifiable (either directly
or indirectly through coding systems). Where institutional activities involve
no interaction or intervention with subjects, and the principal risk associated
with institutional activities is that of breach of confidentiality, the
institution's IRB need not review each collaborative protocol. However,
the IRB must review the center itself. That is, the IRB determines and
documents that the statistical center can ensure that:
(i) the privacy of subjects and
the confidentiality of data are adequately maintained, given the sensitivity
of the data involved;
(ii) each collaborating institution holds
an applicable OPRR-approved Assurance;
(iii) each protocol is reviewed and approved
by the IRB at the collaborating institution prior to the enrollment of
subjects; and
(iv) informed consent is obtained from
each subject in compliance with HHS regulations.
• Maintain "operations centers" or "coordinating
centers" for multi-site collaborative research. Where institutional
activities involve no interaction or intervention with subjects, the IRB
need not review each collaborative protocol. However, the IRB should determine
and document that the operations or coordinating center can to ensure
that:
(i) management, data analysis,
and Data Safety and Monitoring (DSM) systems are adequate, given the nature
of the research involved;
(ii) sample protocols and informed consent
documents are developed and distributed to each collaborating institution;
(iii) each collaborating institution holds
an applicable OPRR-approved Assurance;
(iv) each protocol is reviewed and approved
by the IRB at the collaborating institution prior to the enrollment
of subjects;
(v) any substantive modification by the
collaborating institution of sample consent information related to risks
or alternative procedures is appropriately justified; and
(vi) informed consent is obtained
1Agents include all individuals
performing institutionally designated activities or exercising institutionally
delegated authority or responsibility.
|
| An Institution is NOT "Engaged In Research"...
Institutions would not be considered "engaged"
in human subjects research (and would not need an Assurance) if their involvement
is limited to the following:
If the institution's employees or agents:
• Act as consultants on research
but at no time obtain, receive, or possess identifiable private information.
Should a consultant access or utilize individually identifiable private
information while visiting the research team's institution, the consultant's
activities become subject to the oversight of the research team's Institutional
Review Board (IRB). However, the consultant's institution is not considered
to be "engaged" in the research and would not need an Assurance.
However, should the consultant obtain "coded" (i.e.,with the potential
to link back to identifiers) data for analysis at the consultant's home
institution, that institution becomes "engaged" in human subjects research,
and would need an Assurance, unless a written agreement unequivocally prohibits
release of identifying codes to the consultant.
• Perform commercial services
for the investigators and they adhere to commonly recognized
professional standards for maintaining privacy and confidentiality. An
example would be when a laboratory performs analyses of blood samples for
investigators solely on a commercial basis.
• Provide information to prospective
subjects, including:
(i) informing prospective
subjects about the availability of research;
(ii) providing prospective subjects with
written
information about research (which may include a copy of the relevant informed
consent document and other IRB-approved materials) but not obtaining subjects'
consent or acting as authoritative representatives of the investigators;
(iii) providing prospective subjects
with information about contacting investigators for information or
enrollment; or
(iv) obtaining and appropriately
documenting prospective subjects' permission for investigators to contact
them. As an example, a clinician provides patients with literature about
a research study, including a copy of the informed consent document, and
tells them how to contact the investigator if they want to enroll.
• Upon a subject's request,
release identifiable private information to investigators with the
prior written permission of the subject, An example would be when, with
written permission of the subject, a clinician releases the subject's medical
record to investigators.
• For public health purposes,
release identifiable private information or specimens to a State or Local
Health Department or its agent within the recognized authority of that
Department. However, if the Department's investigators use the information
or specimens for research purposes, then Department is engaged in research,
and would need an Assurance.
• Release existing information
and/or specimens to investigators in non-identifiable form, where such
information/specimens have been obtained by the institution for purposes
other than the investigators' research. As an example, a hospital
pathology department releases excess tissue specimens and relevant medical
record information to investigators, but these materials include no direct
or indirect identifiers through which the identity of individual subjects
could be ascertained, either by investigators or by hospital personnel,
including the pathology department.
• Receive information or specimens
for research from established repositories operating in accordance
with (i) an approved Assurance; (ii) federal guidance (iii) and written
agreements unequivocally prohibiting of release of identifying information
to recipient investigators.
Additionally, an institution is not
"engaged" if:
• It permits use of its facilities
for intervention or interaction with subjects by research investigators.
As an example, school permits investigators to test students whose parents
have provided written permission for their participation.
• The institution, its clinical
staff (or a private practitioner) provides protocol-related care
and/or follow-up to subjects enrolled at distant sites by clinical trial
investigators in officially recognized Cooperative Protocol Research Programs
(CPRPs).
|
If you would like to know more about the process of helping your collaborators
obtain an SPA, contact the Office of Research Development at 312-633-4940.
Back to top
|
