Frequently Asked Questions
Questions
I. Grants and Cooperative
Agreements
II. Contracts
III. Ancillary Studies
IV. Genetic and Pedigree data
V. Other Questions
Answers
I. Grants and Cooperative Agreements
Which grants and cooperative
agreements are subject to the limited access data policy?
For grants, studies requesting $500K or more in direct
costs in one or more years and identified as being of high programmatic
interest to NHLBI are subject to the limited access policy. For cooperative
agreements, clinical trials or epidemiology studies that are supported by the
U01 mechanism and have 500 or more participants are subject to the
limited access policy.
Back to
Table of Contents
Is there standard language that will be added to
the grant award?
The Grants Operations Branch, working with program
staff, will ensure that the following language is included in the Notice of
Grant Award:
"Limited access data will be released under this
study. Limited access data refers to study data, with certain deletions and
recoding, that are released to requesting institutions and investigators for
specific purposes and with certain restrictions and conditions. Limited access
data will be made available to the public in accordance with the NHLBI Policy
for Distribution of Data (http://www.nhlbi.nih.gov/resources/deca/policy_new.htm)
as revised on June 27, 2005."
Back to
Table of Contents
A study funded by a grant (e.g., R01 or cooperative
agreement) that was not required under the previous policy to provide a limited
access data set is undergoing a competitive renewal. Does the new policy now
apply to previously collected data?
Studies will not be required to include previously
collected data in the limited access data set; however, since the intent of the
policy is to make data available to the research community, every effort should
be made to release previously collected data. The added burden of documenting
and cleaning previously collected data should be considered when developing
budgets for the renewal period.
Back to
Table of Contents
A study funded by a grant is using population
samples from existing contract studies that are already required to provide
limited access data sets. What data would be released by each funding mechanism
through the limited access process?
In this specific situation, the grant is also an
ancillary study to more than one contract study. In order to avoid releasing an
implied identifier for the contract studies, the data that belong to the grant
(e.g., special labs, imaging studies or procedures) would only be released from
the grant. For completeness, the grant would also need to release some parent
study data regarded as an integral part of the grant analyses; however, the
limited access data from the grant should not contain any contract study
identifiers. The contract studies that provided the population samples would
not release data collected by the grant even if it is ancillary to the contract
study (see Ancillary Study question #7).
Back to
Table of Contents
A grant includes cohorts from populations
explicitly excluded from the policy (i.e., American Indian or Alaskan Native).
Should a limited access data set still be provided to NHLBI?
Participants from populations excluded from the policy
should be removed from the data set. If the contribution of the excluded
participants is so large as to seriously compromise the value of the data set
for scientific purposes, then program officials, with concurrence of program or
division directors, will decide if the grant should be excluded from the
policy
Back to
Table of Contents
What is the role of the NHLBI program official in
ensuring the limited access data set from a grant is properly prepared and
documented?
The program official will ensure that a study prepares
the limited access data set and appropriate documentation. The limited access
project official (currently Sean Coady) is
available to staff and investigators for consultations on strategies for the
review and preparation of the limited access data set and appropriate
documentation. General guidelines for the preparation of the limited access
data set can be found on the NHLBI public website
(http://www.nhlbi.nih.gov/resources/deca/policy_new.htm#3b).
Once the data have been reviewed (and, if necessary, modified) and
documentation assembled, the material should be sent to the program official.
The program official will then inform the limited access project official that
the data and documentation from the study have been prepared. The data and
documentation are then reviewed for acceptability for release through the
limited access program. If specific revisions are needed to the data or
documentation, the program official serves as the point person for all
communications with the study.
Back to
Table of Contents
Limited access data from a grant are scheduled for
release after funding expires. Should the limited access data be prepared
potentially years in advance of the scheduled release? Will funds in the
estimated amount of the limited access data preparation budget be restricted so
that the limited access data can be prepared closer in time to the actual
release date?
Limited access data could be prepared prior to the end
of funding even though this may be several years prior to the scheduled release
date. The data and documentation should be delivered to NHLBI for review with
sufficient lead time to ensure that any revisions can be completed while
funding is still in place. Once a final data set has been prepared, either the
study or NHLBI can hold the data until the appropriate release date or any time
prior if the principal investigator approves an earlier release. The advantage
of preparing the data early is that study investigators are familiar with data
and documentation issues and may therefore be able to prepare the limited
access data and documentation more efficiently. Another option would be to
restrict funding in the amount of the expected limited access preparation
budget and the study could complete the data and documentation preparation at
any time prior to the scheduled release date.
Back to
Table of Contents
How will NHLBI track which grants are required to
prepare limited access data sets and when the data should be delivered to
NHLBI?
Working with appropriate staff, NHLBI will develop
checkboxes to the competing (new) awards checklist indicating whether a limited
access data set is required for the grant. The limited access project official
(currently Sean Coady) will periodically
generate reports to develop a list of grants required to provide limited access
data sets. Program officials will periodically be asked to update certain
relevant dates regarding the progress of the study such as when recruitment is
expected to be completed, anticipated dates of final data cleaning for
analysis, and for clinical trials, the anticipated date of publication of the
primary results.
Back to
Table of Contents
What is the expected budget for preparation of a
limited access data set, and what components should be included in it?
The budget for preparation of a limited access data
set is dependent on the size and complexity of the study data. In general, the
staff involved would include a statistical programmer to review the data and
perform any actions needed to conform to the NHLBI limited access policy,
support or other staff to review documentation and create electronic versions
suitable for new investigators to use to conduct analyses, and a
biostatistician or other investigator to manage the overall effort. In brief,
informed consents would need to be reviewed to determine patients or subjects
eligible for inclusion in a limited access data set and whether or not
commercial/noncommercial use versions of the data are required (see Other Questions #2). The review of the data would
include looking at frequency distributions of data fields that involve visually
identifiable traits (such as height and weight) or traits that are reasonably
discernible through a casual knowledge of a study subject (such as education,
marital status, income), traits of a sensitive or potential legal nature (such
as illicit drug use, sexual behaviors), long comment fields with potentially
identifiable information, other potentially identifying traits with low
frequencies (such as specific causes of death), and data fields that may have
validity issues (such as a test or procedure found to not provide reliable
data). Modifications to the data would include replacing the original IDs with
new random IDs, recoding all dates to a time interval such as time since
randomization or enrollment, top/bottom coding of visual traits such as height
and weight, removing any geographic identifiers, grouping variables such as
race, marital, status, education, and removing invalid fields or long comment
fields. The documentation to prepare would include items such as forms, data
dictionaries, descriptions of variables that were derived, study protocols and
manuals, and publication lists. For a study with approximately 2,000
participants, 30-40 baseline and follow-up forms, and outcome data a rough
estimate of time would be 0.25-0.40 annual FTE for staff and 0.05-0.15 FTE for
project management.
Back to
Table of Contents
What points should be discussed in the =$500K
request letter and the grant application about limited access data
sets?
The general points to discuss include: a general
statement addressing expected compliance with the limited access policy, brief
statement on plans for preparation of the limited access data set, and a
statement indicating that the budget reflects inclusion of resources for the
limited access data set. In the $500K letter, this could be relevant to the
acceptance criterion about a dissemination plan. However, both a limited access
data plan and a dissemination plan for results should be discussed. Below is an
example of addressing the limited access policy in a recent $500K request
letter:
- In compliance with the NHLBI current policy
for observational studies (October 1, 2005), two years after the data are
finalized and cleaned, we will send to the NHLBI Limited Access Data
Sets (LADS), consisting of cleaned data on individual genotypes and
phenotypes, (excluding frank PHI identifiers such as names, addresses, etc.).
Of course, such limited access data sets will not include those subjects whose
informed consents will not allow inclusion. As we will be completing the
genotypes in phases, we anticipate that there will actually be 3 LADS
deliveries corresponding to data for 1) GWAS in the Caucasian cases-controls
(genotypes completed in YR01, cleaned for publication by mid YR02, LADS
delivery mid YR04) 2) GWAS in African-Americans (genotypes completed in YR02,
cleaned for publication by mid YR03, LADS delivery mid YR05), 3) Fine mapping
of regions in Caucasians and African Americans (genotype compete mid YR05 and
cleaned for publication by end of YR05, LADS delivery mid YR07). Our budget
includes one third FTE of a Statistical Data Analyst programmer at the CC
(approximately $25K direct) to support the limited use data preparation,
documentation, and distribution in these years.
Back to
Table of Contents
II. Contracts
Is there standard language that will be
incorporated in the RFP and subsequent contract for delivery of a limited
access data set?
The Contracts Operation Branch will ensure that the
following language is included in each RFP and contract:
- The National Heart, Lung, and Blood Institute
(NHLBI) has supported collection of data from participants in numerous clinical
trials and epidemiologic studies. These well-characterized population samples
are rare and valuable scientific resources. In order to take full advantage of
them and maximize their research value, it is important that data collected
with public funds be made available, under appropriate terms and conditions, to
the largest possible number of qualified investigators in a timely manner.
Limited access data will be released under this study. Limited access
data refers to study data, with certain deletions and recoding, that are
released to requesting institutions and investigators for specific purposes and
with certain restrictions and conditions. Limited access data will be made
available to the public in accordance with the NHLBI Policy for Distribution of
Data
http://www.nhlbi.nih.gov/resources/deca/policy_new.htm)
as revised on June 27, 2005. All changes to the policy are hereby incorporated
by reference without further amendment to the contract. Limited access data is
a deliverable under this contract for this study, as described in Section C.
Description/Specification/Work Statement and/or Section F. Deliveries or
Performance of the contract."
Back to
Table of Contents
Are existing contracts expected to comply with the
new policy in terms of the timing for release of limited access data, or
continue to follow the previous policy until the study is renewed?
Existing studies are encouraged but not required to
comply with the new policy. There may be potential benefits to migrating to the
new policy, such as simplifying timelines for the study. For example, it is
possible that a new ancillary study would have a release date far earlier than
the contract data.
Back to
Table of Contents
For an existing study to comply with the new
policy, the study will need to provide one or more years worth of catch-up data
at the time of their next scheduled release, creating a burden on study staff.
Should the additional burden and, therefore costs, be considered by program
staff?
Providing the catch-up data does create at least some
additional burden. Program or contract staff should be consulted on possible
cost reimbursement options.
Back to
Table of Contents
III. Ancillary Studies
Are all ancillary studies required to submit a
limited access data set?
An ancillary study is expected to provide a limited
access data set only if the parent study is subject to the limited access
policy.
Back to
Table of Contents
Currently, there is no provision in the [parent
studys] ancillary study policy for an ancillary investigator to provide
limited access data to the coordinating center. Is there standard language that
should be added to the ancillary study policy?
Ancillary study policies should make clear that
ancillary study data are to be included in the limited access data set
submitted to NHLBI. Although this may require tailoring to the policies of
individual studies, the following language should be added in the section
related to return of ancillary study data to the main study:
- Limited access data will be released under
the [PARENT STUDY]. Limited access data refers to study data, with certain
deletions and recoding, that are released to requesting institutions and
investigators for specific purposes and with certain restrictions and
conditions. Limited access data will be made available to the public in
accordance with the NHLBI Policy for Distribution of Data
(http://www.nhlbi.nih.gov/resources/deca/policy_new.htm)
as revised on June 27, 2005. This policy requires that data from studies
ancillary to [PARENT STUDY] be included in the limited access data set
submitted to NHLBI.
It may be worthwhile to add a statement to the
ancillary study proposal form or agreement, if the parent study has one,
stating:
...the ancillary study investigator understands
that data from this ancillary study will be made available to the public in
accordance with the NHLBI Policy for Distribution of Data
(http://www.nhlbi.nih.gov/resources/deca/policy_new.htm)
as revised on June 27, 2005, and that adequate funds will be designated for
this purpose.
Back to
Table of Contents
Are the coordinating centers expected to produce
and appropriately document ancillary study data or should this be the
responsibility of the ancillary study investigator?
The parent study will need to consider how it should
implement the policy. Since the coordinating centers are more familiar with the
limited access guidelines, it may be advantageous for the coordinating center
to review and, if necessary, revise the ancillary data (based on NHLBI
guidelines) and add the ancillary study data to the parent study database. The
coordinating center would need to request that the ancillary study investigator
provide appropriate documentation. The coordinating center should require
appropriate compensation either through the ancillary study award or directly
as part of the studys contract or grant award.
Back to
Table of Contents
If the ancillary study timeline differs from the
parent study timeline, will the limited access data set release date relate to
the timeline of the ancillary study or the parent study exam cycle? For parent
studies with many ancillary studies, there may be lots of timelines to track.
Are the coordinating centers responsible for this as well? Note this further
increases the burden on coordinating centers.
To the extent possible, ongoing observational studies
will be updated only on an annual basis and ancillary study data should be
delivered to NHLBI during the annual update that most closely approximates the
timeline as described in the policy. For some ancillary studies, this may mean
a slight delay in when the data are actually released through the limited
access program. One possible approach would be to track ancillary studies on a
fiscal year basis, i.e. ancillary studies funded within a given
time frame will be released at a prescribed date. For clinical trials for which
the publication date of the main study results will differ from that of the
ancillary study, the ancillary study data should be delivered to NHLBI
according to the timeline as described in the policy. Regardless, the added
burden to coordinating centers should be taken into account when budgeting
costs for the ancillary study.
Back to
Table of Contents
Does the new limited access policy apply to
ancillary study applications reviewed or funded prior to October 1,
2005?
The new policy will not be applied retrospectively.
Unless NHLBI required otherwise, only ancillary study applications reviewed by
the parent study on or after October 1, 2005 are expected to provide limited
access data to the coordinating center.
Back to
Table of Contents
Are there situations in which an ancillary study is
exempt from the policy?
There may very well be situations where a limited
access data set from an ancillary study will not be subject to the policy. For
example, an ancillary study on a very small subgroup of participants may not
have sufficient value to the research community to justify a limited access
data set, or the data collected may have identifiable properties that cannot be
modified without compromising scientific value, or provisions in the informed
consent may prohibit release through the limited access program. There may also
be situations in which the ancillary study is a component of a larger grant or
cooperative agreement that includes participants/specimens/data from other
studies that are not required to produce a limited access data set. Program
officials, with concurrence from program or division directors, will make the
decision on when an ancillary study is exempt from the policy.
Back to
Table of Contents
An ancillary study is using participants from a
parent study along with participants from other contract studies as part of a
large grant or cooperative agreement. Will the parent study and the large
grant/cooperative agreement both be required to provide a limited access data
set?
In this situation, the ancillary data (for example,
special labs, imaging studies or procedures) collected as part of the
grant/cooperative agreement would only be released from the grant or
cooperative agreement and not from the parent study. Although the parent study
would not be releasing the ancillary data, the grant/cooperative agreement
would release the parent study data that is regarded as an integral part of the
grant/cooperative agreement. The grant/cooperative agreement would not release
any identifiers that associate a subject with a specific parent study (See Grants and Cooperative Agreements question #4).
Back to
Table of Contents
If an NHLBI funded parent study has an ancillary
study funded by another NIH institute or some other entity, is a limited access
data set from the ancillary study still expected?
Since provisions for an NHLBI limited access data set
should be incorporated into the ancillary study policy, there is an expectation
that regardless of funding source, NHLBI will receive an ancillary study data
set. If the ancillary study funding is from an NIH institute with similar data
sharing policies, then it is possible that the ancillary study data will be
released by more than one institute; however, parent study data can only be
released by NHLBI.
Back to
Table of Contents
Ancillary studies may be generational,
that is, the ancillary study itself may have ancillary studies. How far down
the line does the policy apply? How would offspring ancillary study
data be coordinated?
If the ancillary study uses parent study data or
specimens from the parent study, then the data should be subject to the
policy.
Back to
Table of Contents
IV. Gentetic and Pedigree data
Should studies provide to NHLBI both a
self-reported and a genetic pedigree?
Although there are scientific benefits to both types
of pedigree data, only genetic pedigrees should be sent to NHLBI for
distribution through the limited access program due to the sensitivity of
non-paternity.
Back to
Table of Contents
How should non-paternity be masked in genetic
pedigrees?
One approach would be to replace the self-reported
paternal ID with a phantom or dummy ID. The study should ensure
that the paternal dummy ID is unique, but indistinguishable from
other IDs (the ID cannot be linked with any phenotype data, but the format of
the ID should not reveal it as a dummy ID).
Back to
Table of Contents
Should large pedigrees be masked?
In a single center study, a large pedigree may be
considered an identifying characteristic and a study may wish to consider
removing such pedigrees from the limited access data set or truncating
pedigrees at a certain size (limit pedigrees to size 6 or less). What
constitutes a large pedigree is entirely subjective, and a study
should carefully consider the scientific value of large pedigrees and the
protections NHLBI has in place (investigators must sign a data distribution
agreement and obtain IRB approval) in deciding when large pedigrees should be
removed. In multicenter studies with no clinic or center identifier and no
special sample selection criteria by center (for example, race as a criteria
for a specific site), large pedigrees may be kept.
Back to
Table of Contents
Will genetic/pedigree data be provided to all
investigators requesting data from a particular study?
Genetic/pedigree data will only be provided to
investigators specifically requesting them. The genetic/pedigree data and
documentation should be sent to NHLBI as stand-alone data since not all
investigators will receive genetic/pedigree data. For example, if a large
population study also has ancillary data on specific SNPs, then an investigator
requesting the genetic data would be required to complete a data distribution
agreement for both the clinic/follow-up data and a separate agreement for the
genetic data. The IRB approval would need to indicate that the proposed project
included the use of genetic data.
Back to
Table of Contents
V. Other Questions
A studys informed consent was very detailed
and included several checkboxes for how the participants data could be
used. For example, there were checkboxes for: type of research (i.e., the
participant could consent to using their data for research on any disease, or
for cardiovascular disease and risk factors only), investigators that can do
research on their data (i.e., any qualified investigator or study investigators
only), and whether or not commercial interests can have access to the data. How
should the limited access data set be prepared in terms of which participants
are included?
The limited access data set must be prepared in a
manner consistent with each participants informed consent. Since NHLBI
cannot restrict the type of research conducted on a limited access data set,
participants requesting that their data be used only for a specific research
purpose (CVD only research, for example) should be removed from all limited
access data. The NHLBI policy also requires that for studies in which the
informed consent allows participants to request that their data not be used for
commercial purposes, the limited access data should include a commercial
purpose and a non-commercial purpose version. In preparation
for the limited access data set, the coordinating center will need to enter
data on the level of consent for each study participant.
Back to
Table of Contents
The terms "Commercial Purpose" and "Non-Commercial
Purpose" seem confusing. Are they named for an intended purpose or for the type
of investigators that are eligible to receive them?
The terminology is based on what the participant
indicated on the informed consent. A participant indicating on the informed
consent that his or her data may be shared with investigators developing
commercial products will be included in the commercial purpose data set.
Therefore, the commercial purpose data set may be used by either academic or
commercial entities developing commercially viable products. If a participant
indicated that his or her data may not be shared with investigators developing
commercial products, then that data will not appear in the commercial purpose
data set, but will be in the non-commercial purpose data set providing the
informed consent does not prohibit sharing of data beyond the study itself.
Investigators from both academic and commercial institutions may receive the
non-commercial data set; however, the investigator must indicate a
non-commercial use of the data in the data distribution agreement.
Back to
Table of Contents
Why should investigators provide a list of
participant IDs to the NHLBI?
Study IDs should be replaced with new randomized IDs.
Since NHLBI serves as the data repository, the link between the randomized IDs
and the original study IDs is requested as a backup for the study in case
delinking the limited access data is needed for an unforeseen purpose.
Participant IDs may be sent to NHLBI in several formats. One is to include all
participant IDs with indicators for permission to share data and
permission to share data for a commercial purpose. Another approach
is to supply separate files of IDs. One file would list only the IDs for which
the data can be shared for the purpose of non-commercial research
(non-commercial purpose) and one file for which the participants indicated that
their data could be shared for any reason (commercial purpose).
Back to
Table of Contents
Must all of the data be
reviewed for either truncation at the extremes (continuous traits) or grouping
so that all counts are 20 or greater (categorical traits)?
The guidelines for modifying data are general and
therefore not universally applicable. In general, only traits that have
features that may be identifiable visually (e.g., height, weight, waist
circumference) or could be ascertained through a casual knowledge of a subject
(e.g., race, Hispanic status, marital status, employment status, education,
income, significant outcomes of low frequencies such as cancers or specific
causes of death) should be reviewed for possible modification. In general,
multicenter studies have more leeway in performing fewer modifications since
the removal of study center (geography) will, by itself, help to resolve many
of the data modification issues.
Back to
Table of Contents
How should image files or imaging data be
transferred to NHLBI for distribution through the limited access
mechanism?
If possible, summary scores or other reduced data are
preferred from imaging studies such as MRIs, carotid ultrasounds, or CT scans.
If a study prefers to release raw image data, then the limited access project
official should be consulted for options in releasing image data
Back to
Table of Contents
Is the limited access process the only mechanism
for sharing data?
In addition to the limited access mechanism, study
investigators may want to establish a process for collaborations. Study
investigators have a particular expertise regarding the study and in some
situations may be willing to collaborate with other investigators. Coordinating
centers may wish to consider establishing a website to describe the study and
mechanisms for collaboration.
Back to
Table of Contents
Web sites
Available data sets, study descriptions, data
distribution agreements, etc.
http://www.nhlbi.nih.gov/resources/deca/default.htm
The current and previous limited access data policy
http://www.nhlbi.nih.gov/resources/deca/policy.htm
NHLBI guidelines for applications with direct costs of
$500K+ in any one year
http://www.nhlbi.nih.gov/funding/policies/500kweb.htm
|