30 Day Comments and responses

1
Response to Comments Received During the 30-day Public Comment Period
The Office of Management and Budget received four comments during the 30-day public comment period
from November 23, 2011, to December 23, 2011. NIH’s response to those comments follows. Each of
the four comments is provided in its entirety following NIH’s response.
Coalition for 21st Century Medicine
Several of the Coalition’s suggestions (e.g., adding common test name, clarifying the definition of “test
performed in-house”) have been addressed in the Genetic Testing Registry (GTR). Suggestions for
additional data fields and categories for GTR menu options (e.g., gene expression testing and protein
expression testing to the data field for laboratory types of services) will be considered in consultation with
GTR clinical advisory groups. The National Institutes of Health (NIH) notes that gene and protein
expression tests are generally somatic tests, which are outside the scope of the initial phase of the GTR,
but will be added in subsequent phases. Regarding the comment to revise "internal test validation method
description" to include a field for reference value/intervals, this field is in the GTR. NIH will clarify the
instructions and note that reference values can be included. The Coalition expressed concern that
requesting information about arrays and instruments could reveal proprietary information. At the
November 2011 public stakeholder meeting and in subsequent presentations, NIH explicitly stated that it
will not request proprietary information. This point also will be made clear in instructions for submitters.
The Coalition pointed out that definitions for “precision” and “accuracy” were reversed; the definitions
have been corrected. Regarding the option to describe the test methodology, submitters can choose
"other" from the menu list and provide the test methodology or can provide this information in the text
field for "Description of test procedure/protocol." NIH will consider the Coalition’s suggestion to
incorporate additional data fields for validation studies at a later date. Regarding the estimated burden,
after the 60-day comment period, NIH recalculated the cost burden using a mean hourly wage more
appropriate for experience laboratory personnel ($30.63 instead of $22.85). NIH re-evaluated the time
burden and decided that it remained valid.
Robert Cook-Deegan, Duke Institute for Genome Science & Policy
NIH shares Dr. Cook-Deegan’s concern about avoiding duplication with other government databases and
has engaged in interagency discussions to ensure that the GTR does not duplicate other information
collections. Before NIH’s March 2010 publicly announcement of its plan to develop the GTR, it met with
the Food and Drug Administration (FDA) and the Centers for Medicare & Medicaid Services (CMS) to
discuss whether such a registry would be useful to these agencies. Avoiding duplication and minimizing
reporting burden were important elements of this discussion. Once GTR development began, NIH held
two broader trans-agency meetings that included FDA, CMS, the Centers for Disease Control and
Prevention, Health Resources and Services Administration, and Federal Trade Commission. The
objectives of these meeting were to demonstrate GTR prototypes and gather feedback, discuss approaches
that would reduce reporting burden for laboratories, and learn of federal activities relevant to the GTR.
In developing the GTR, NIH was sensitive to Dr. Cook-Deegan’s concern about the burden to provide
certain data repeatedly (e.g., quality control programs, contact person, laboratory certification). GTR
submitters will be able to use time-saving features such as a copy function and bulk submission of data.
His suggestion that GTR integrate with resources/tools such as OMIM and MESH categories has already
been implemented. The suggestion to link to mutation databases such as Human Variome and
MutaDATABASE will be done through ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/), a companion
database to the GTR that will be available later this year.

2
With regard to Dr. Cook-Deegan’s recommendation to include patent and licensing information in the
GTR, the primary purpose of the GTR is to enhance the transparency of information about the scientific
and clinical basis of the test, and NIH does not plan to request data elements that do not relate directly to
this goal. At the same time, the database will continue to evolve after its launch and as it matures and
gains greater stakeholder support and wider utility, NIH will consider adding data elements that have
other public policy purposes.
Human Genetics Society of Australasia (HGSA)
Several of HGSA’s comments have been addressed in the design of the GTR (e.g., presenting information
in tiers, providing pull-down menus to complete data fields, collecting accreditation information from
international laboratories). HGSA expressed concern about the detail of some data fields (e.g., platforms,
instrumentation, gene variants). These data fields are optional; submitters can choose not to complete
them. HGSA suggested that test price and turnaround time be added to the information collection. Based
on the collective feedback from extensive stakeholder engagement prior to the public comment periods
required by the Paperwork Reduction Act, NIH decided not to include test price or turnaround time in the
initial phase of the GTR. It will reconsider these data elements in later phases of the GTR. HGSA also
suggested that NIH facilitate the importation of data from databases in addition to GeneTests. Using a
bulk upload option, laboratories can place data held in personal or other databases onto a standard form
for import into the GTR.
National Society of Genetic Counselors (NSGC)
NSGC requested that NIH implement a structured peer-review process to safeguard against false and
inaccurate data. Absent this process, NSGC would like a notice that NIH, or other government entity,
does not endorse the information in the GTR. NIH has taken a number of steps to address this concern,
including an explicit disclaimer that states NIH or the U.S. government makes no endorsements of tests or
laboratories listed in the GTR. Data fields suggested by NSGC (e.g., CLIA certification, common or
commercial name of each test, and target population) have already been included in the GTR. NSGC also
recommended that GTR focus on a primary audience. At the public stakeholder meeting in November
2011, NIH clarified that it will focus initially on health care providers as the primary GTR audience and
will expand to other audiences in subsequent phases of the Registry. This point has been made in
subsequent presentations and will continue to be made in the future.

December 23, 2011
Via Electronic Mail Submission to: OIRA_submission@omb.eop.gov; gtr@od.nih.gov
Amy P. Patterson, M.D.
Associate Director for Science Policy
Office of Biotechnology Activities
National Institutes of Health
6705 Rockledge Drive, Suite 750
Bethesda, MD 20892
RE: Submission for OMB Review; Comment Request Information Program on the Genetic
Testing Registry, 76 Fed. Reg. 72,424 (Nov. 23, 2011)
Dear Dr. Patterson:
On behalf of the Coalition for Twenty-First Century Medicine (the “Coalition”), we are pleased to submit
comments in response to the above-captioned request for comments from the National Institutes of Health
(the “NIH”) on the proposed Genetic Testing Registry (“GTR”). As we emphasized in our July 31, 2010
letter to Dr. Francis Collins responding to the NIH’s request for information on the development of a
genetic testing registry (the “July 31, 2010 Letter”), our Coalition strongly supports the NIH’s efforts, and
we appreciate the opportunity provided through this request for comment to discuss further the
implementation of the GTR.
The Coalition represents diagnostic technology companies, clinical laboratories, researchers, and other
relevant industry stakeholders who work together to improve the quality of healthcare by encouraging
research, development, and commercialization of diagnostic technologies in order to improve patient
outcomes and reduce healthcare costs. Through its diverse membership, the Coalition seeks to ensure that
innovative diagnostic and therapy management tests are available to patients and their physicians, and
consistent with this mission, the Coalition applauds the NIH on its development of the GTR, which is
designed to serve as a public resource for information about the availability and scientific basis of
genetic-based tests. As a representative of a wide range of industry stakeholders, the Coalition is wellpositioned to provide comments on the development of the reporting fields in the GTR.
The Coalition generally supports the majority of the proposed fields developed to collect the information
for the GTR and believes these fields will effectively assist the NIH in achieving its goal of increasing
access for the public to information about genetic-based tests. However, we have included below
comments and concerns for specific proposed fields where certain definitions of the terms of the field will
likely cause significant confusion to the laboratories reporting these data. Within these comments, we
also provide considerations for additional data fields that NIH may want to consider including in the GTR
at a future date. Finally, we address concerns with regard to the estimated burdens on laboratories to
compile and complete submissions to the GTR.
**********

DM_US 30976721-2.068154.0023

Dr. Amy Patterson
December 23, 2011
Page 2 of 6
I.

Comments on Specific Genetic Testing Registry Fields
A.

Laboratory Definitions

The Coalition generally supports the reporting fields under the “Laboratory Information” category. With
regard to the field for “Laboratory Types of Service,” however, the Coalition is concerned that the types
of testing performed by its members are not included in the list of possible types of services. The
majority of the laboratories that are members of the Coalition perform gene expression or protein
expression testing used to (i) assess the risk of progression or recurrence of an underlying disease, or
(ii) evaluate the potential benefits of therapy for a specific disease. In their article Developing the
Blueprint for a Genetic Testing Registry, Javitt, et al. specifically define a genetic test to include an
analysis of DNA, RNA, chromosomes, proteins, or metabolites to detect levels of gene expression in a
human sample.1 The decision tree developed in this article for inclusion of tests in the GTR would also
support the conclusion that the types of tests offered by the Coalition’s members would be included in the
GTR. Moreover, the Coalition’s laboratory members are single-source providers of the tests performed,
and therefore it is important that their types of services be represented in the options for this field. Insofar
as the NIH uses the Javitt, et al. article as a reference for creating the “Laboratory Types of Service” field,
we recommend that NIH add two categories entitled “Gene Expression Testing” and “Protein Expression
Testing” to the proposed list under “Laboratory Types of Services,” and we also request that the NIH
make the corresponding additions to the “Test-Specific Laboratory Services” field.
B.

Test Information

As the Coalition requested in its July 31, 2010 Letter, the NIH has provided a field for laboratories to list
the “Laboratory Test Name.” The Coalition appreciates the NIH’s inclusion of this field, but requests that
the NIH also provide a field for the laboratory to list the common name for the test (which may be a
proprietary name or a non-proprietary name), which will assist individuals – particularly researchers and
other industry stakeholders – with another way to identify the test.
With regard to the field entitled “Test Performed In-House,” the Coalition has two primary concerns.
First, the Clinical Laboratory Improvement Amendments of 1988 (“CLIA”) uses the term “in-house
testing” to refer to the development and performance of laboratory-developed tests.2 In contrast, we
understand the purpose of the GTR data field “Test Performed In-House” to be limited to data regarding
the performance of different portions of the reported test. Accordingly, we would strongly recommend
that the NIH provide a clarifying definition of “in-house” so as to avoid confusion for laboratories as they
submit information for this field.
In addition, if the Coalition is correct to assume that “in-house testing” refers exclusively to information
regarding the performance of a test (and not to its development), the Coalition is concerned that the
options in this field may not accurately capture how a test is often performed by the laboratories. By way
of example, for many diagnostic tests, specimen preparation may take place both at the performing
laboratory and the referring laboratory. This may occur because certain steps always occur at both
locations (e.g., retrieval of blocks always performed at the referring laboratory and review of specimen
slides to confirm the adequacy of the sample performed at the performing laboratory) and/or because
certain steps may be performed at the referring laboratory on some occasions and at the performing
laboratory on others (e.g., cutting of blocks to prepare slides for analysis). As drafted, however, these
1
2

Javitt, et al. Developing the Blueprint for a Genetic Testing Registry. Public Health Genomics 2009; 13:95-105.
42 C.F.R. § 493.1253(b)(2).

-2DM_US 30976721-2.068154.0023

Dr. Amy Patterson
December 23, 2011
Page 3 of 6
fields do not provide an opportunity for the laboratory to explain these multiple preparation steps when
completing this field.
Given the complexity of advanced diagnostic testing and the variability as to the location of where steps
of a test are performed, the Coalition requests that the NIH revise the category “Test Performed InHouse.” We recommend that the NIH keep the fields “Entire test performed in-house” and “Entire test
performed externally,” but if a test does not fall into either of these categories, then the fields for
specimen preparation, wet lab work, interpretation, and report generated would contain an open text field
to allow the laboratory to explain where each of these steps occur. This will minimize confusion as the
laboratory completes this portion of its submission, while also ensuring that the GTR contains the most
accurate description of the performance of the test.
C.

Indications for Use

Consistent with the Coalition’s recommendations in the July 31, 2010 Letter, the NIH proposes to include
a field in the GTR for the laboratory to identify the indication(s) for use of the test. While the Coalition
appreciates the NIH’s efforts to develop this field, many of the Coalition’s members perform tests that are
designed to (i) assess the risk of recurrence or progression of a disease, or (ii) predict the benefit of a
particular treatment. However, these types of tests do not fit into any of the options listed in this proposed
data field. The Coalition recommends that the NIH add the categories “Disease Progression” and
“Treatment Benefit Assessment” to this field in order to allow laboratories to identify appropriately the
type of test being reported.
On a similar note, the NIH has proposed to include under the “Indications for Use” category several fields
related to the disease for which the test has been developed. The GTR would require the laboratory to list
the “Condition for which test is offered,” defined as the name of the disease/syndrome/drug response for
which the test can be ordered, and the “Description of the Target Population,” defined as the segment of
the population that should be tested for this disease. The Coalition is concerned that it will not be clear to
laboratories completing the fields how to differentiate between the condition for which a test is offered
and the population that should be tested, when both describe the category of individuals who should be
offered the test.
The MMWR article included as a reference for the proposed GTR definition of “Description of the Target
Population” uses patient population as the basis for reporting the intended use of the test.3 In contrast, the
Logical Observation Identifiers Names and Code (“LOINC®”) database identifier referenced for purposes
of the field “Condition for which test is offered” focuses on the “Genetic Disease Assessed.” The
inclusion of both categories may demonstrate an attempt by the NIH to capture the fields recommended
by each source, but these fields would appear to measure the same population and are therefore somewhat
duplicative in nature. As such, the Coalition urges the NIH to remove one of these reporting fields from
the “Indications for Use” category in order to streamline the submission process, or, in the alternative,
provide clarification and examples to demonstrate the data requested for these two categories.

3

Chen et al., Morbidity and Mortality Weekly Report (MMWR): Good Practices for Molecular Genetic Testing for
Heritable Diseases and Conditions (2009), available at
http://www.cdc.gov/mmWR/preview/mmwrhtml/rr5806a1.htm.

-3DM_US 30976721-2.068154.0023

Dr. Amy Patterson
December 23, 2011
Page 4 of 6
D. �

Test Methodology

Although the Coalition supports the NIH’s efforts to include information in the GTR regarding test
methodology, we are concerned generally with the level of detail that would be required from laboratories
under this field. Moreover, the Coalition urges the NIH to consider that these fields may provide only
limited benefit to users of the GTR, most importantly the general public, while creating significant
confusion for laboratories attempting to submit these data.
1. �

Test Methodology: Pull-Down List

With regard to the field for the test method used in the assay (“Test Methodology”), we would note that
developers of the LOINC database have specifically stated that LOINC distinguishes tests by the type of
methodology only “if a given type of method has an important effect on the interpretation of the result.”4
Moreover, the LOINC database does not include broad categories of test methodologies which potentially
encompass a wide range of laboratory tests and for which multiple categories might be appropriate for a
specific test. Instead, the LOINC database permits laboratories to spell out the test method fully in the
LOINC identifier.5 Creating distinct requirements under the GTR as compared to the LOINC database
may create substantial confusion for laboratories when determining under which, if any, category the test
methodology should fall.
By way of example, certain diagnostic tests developed by the Coalition’s members target specific genes in
order to assess the risk of disease recurrence or progression; other tests analyze specific genes in order to
assess the likely benefit of certain therapies; other tests are designed for both purposes. The Coalition is
concerned that its members would be unable to determine whether their tests would fall under the “gene
expression profiling” category, the “GeneID” category, or would be appropriately reported under both
categories in the proposed GTR. In order to avoid confusion and to provide consistency with the
processes used in the LOINC database, the Coalition requests that the NIH instead allow laboratories the
option – but not the requirement – to describe the methodology for a test in the GTR. As an alternative,
the NIH should provide examples of the relevant sub-categories it would include under each method
category.
2. �

Platforms: Laboratory-Specific Pull-Down List; Instrument(s) Used During
Testing: Pull-Down List

With regard to the proposed fields for Platforms and Instrument(s) Used During Testing, we understand
that the NIH anticipates that laboratories would submit the names of specific manufacturers whose assays
and/or instruments are utilized by the laboratory to perform the test. The Coalition urges the NIH to
consider that requiring laboratories to list arrays and instruments with manufacturer-specific information
may raise significant proprietary concerns. By providing this information, laboratories would run the risk
of disclosing confidential data with respect to the procedures used to perform their highly unique testing.
This is especially important for the Coalition’s members insofar as they are single-source laboratories,
and disclosure of their assays and instruments would reveal proprietary information as to how the specific
test conducted at their laboratory is performed. Moreover, the Coalition seeks clarification as to how to
differentiate between the general field “FDA-Approved tests” and the fields listing specific
manufacturer’s arrays, and whether the laboratory would list one or both in describing a certain array used
to perform the test.
4
5

McDonald et al., Logical Observation Identifiers Names and Codes (LOINC®) Users’ Guide 17 (June 2011).
Id. at 39.

-4DM_US 30976721-2.068154.0023

Dr. Amy Patterson
December 23, 2011
Page 5 of 6
Overall, the Coalition is also confused as to the utility of requiring laboratories to report this information
in the GTR. Specifically, if the goal of the GTR is to provide access to the public for information
regarding tests offered by laboratories, then requiring laboratories to complete fields that will not be
available to the public provides little, if any, benefit to the public. In addition, the risk of disclosing
proprietary information – especially in the context of single-source laboratories – outweighs any minimal
benefit to submitting this information. As such, although we agree with the NIH’s decision to prevent the
public from viewing this field, the Coalition strongly urges the NIH to remove this field entirely in order
to minimize proprietary concerns and further streamline reporting procedures under the GTR.
E.

Quality Control and Quality Assurance
1.

Observations on Performance Specification Fields

The Coalition supports the NIH’s inclusion of fields for laboratories to submit data regarding precision
and accuracy consistent with the recommendations in our July 31, 2010 Letter. We would note, however,
that the definitions of “accuracy” and “precision” appear to be reversed. Specifically, “precision” is
defined as reproducibility or repeatability, meaning the degree to which the results are the same when
reproduced under the same conditions, but this definition is included under the heading for “accuracy” in
the proposed GTR definitions. In contrast, “accuracy” is customarily defined as how close the
quantitative results of a test are to the actual (true) quantitative value, but this is the definition for
“precision” under the proposed GTR definitions. We would therefore request that the NIH revise the
definitions for these fields accordingly.
In addition, the Coalition supports the NIH’s decision to include “Analytical Sensitivity” and “Analytical
Specificity” fields to describe a test’s performance specifications, and believes that these are important
data for purposes of demonstrating how laboratories validate the results of the tests which they perform.
With regard to validation methods, however, the Coalition is concerned that the NIH has not provided a
field for laboratories to report the reference intervals, or normal values, for the test. NIH has relied on
CAP.MOL.31245 “Reference/Reportable Range” for purposes of creating the field to list the reportable
range in the GTR. It is important to note that CAP.MOL.31245 also suggests that the laboratory be
required to report the reference value (normal versus abnormal result). The Coalition believes this
information is especially pertinent for healthcare professionals so as to understand better the results
received from a particular test, and therefore we urge the NIH to revise the Internal Test Validation
Method Description to include a field for laboratories to submit the reference value for the test.
2.

Additional Clinical Validity Fields

Finally, the Coalition encourages the NIH to consider expanding the GTR at a later date to incorporate
additional fields regarding validation studies for the test reported by the laboratory. Because many tests
reported under the GTR have an intended use which includes a claim of clinical meaning or usefulness of
the analytical result, the Coalition urges the NIH to add a field to the GTR that permits laboratories to
submit primary and secondary endpoint measures for the test, including the value for each measure, and
the results of any tests measuring the statistical significance of these endpoint measures. These data are
significant for purposes of providing healthcare professionals with important clinical details about a test,
and publishing these data in the GTR provides an opportunity for healthcare professionals to have quick
and easy access to these measures after they receive results from the laboratory.
The Coalition also believes that researchers, health care professionals, and the general public would
benefit from more information regarding clinical studies conducted to establish the clinical validity of a
test. The NIH should consider creating a field for laboratories to submit a synopsis of the study protocol

-5DM_US 30976721-2.068154.0023

Dr. Amy Patterson
December 23, 2011
Page 6 of 6
so that health care professionals and researchers will be able to interpret and evaluate the results of the
studies accurately. This information would be reported by creating additional optional fields under the
“Quality Control and Quality Assurance” category, so long as the NIH determines that inclusion of these
studies would not be misleading or promotional in nature. Although these fields would require additional
reporting by laboratories, the Coalition believes this provides the public with access to important
information regarding the test. Moreover, postponing implementation of these fields until a future date
will permit laboratories to adjust to the fields currently proposed for the GTR before requesting that
laboratories complete additional fields.
II.

Observations Regarding the Burden on Laboratories to Report Data

In the original Request for Comments published in the July 27, 2011 Federal Register,6 the NIH
estimated that it would take laboratory personnel – at a mean hourly wage of $22.85 – an average of three
hours to complete each submission to the GTR. Although the Coalition appreciates that the NIH has
made efforts to streamline and simplify the reporting procedures for the GTR, the complexity of the data
to be reported necessitates that a Laboratory Director review, verify, and complete many of the fields
proposed for the GTR. The participation of Laboratory Directors in the GTR submission process will
therefore significantly affect the NIH’s previous estimates by substantially increasing the three-hour
completion time to review the data, as well as increasing the mean hourly wage estimated by the NIH.
Accordingly, the Coalition would urge the NIH to reconsider its proposed estimates with regard to the
financial burden on laboratories to participate in the GTR.
**********
We appreciate the opportunity to submit comments on the Genetic Testing Registry and we hope that you
have found the information and suggestions in this letter to be helpful. If you have any questions about
our comments, please contact Paul Radensky at 202-756-8794 or pradensky@mwe.com.

6

National Institutes for Health, Request for Comments Under the Paperwork Reduction Act, Section 3506, 76 Fed.
Reg. 44,937 (July 27, 2011).

-6DM_US 30976721-2.068154.0023

22 December 2011
Amy P. Patterson
Associate Director for Science Policy
National Institutes of Health
Via gtr@od.nih.gov and OIRA_submission@omb.eop.gov
Dear Dr. Patterson,
Thank you for the opportunity to comment on the proposed form that will be used

to generate the Genetic Testing Registry (GTR), per Federal Register notice of 23

November 2011 (Vol. 76, No. 226, pp. 72424-5).

The test registry could replace genetests.org as the first place to go to determine

who is offering genetic testing for a particular condition or locus. The OMB

notice appropriately focuses on the potential burden on those submitting data.

Only data essential to make sense of the list and not available more readily

elsewhere should be included.

I have several concerns based on reviewing the proposed fields and mock-up entry. One is
overinclusion of information for which the GTR is not the appropriate archival resource, and therefore
not a good use of time for those submitting entries; another concern centers on patent and licensing
elements that bear directly on the availability of tests but would not be included. My final comments
will focus on overlap and coordination with the regulatory data required of testing services, so that the
NIH database is integrated with rather than separate from data required for those tests (now and in the
future) that will be regulated by the Food and Drug Administration (FDA; both Laboratory Developed
Tests and In-Vitro Diagnostic Multi-Index Assays) and for regulation, coverage and reimbursement by
the Centers for Medicare and Medicaid Services (CMS) under federally funded health programs and the
Clinical Laboratories Improvement Amendments. There is a danger that three agencies will set up
somewhat different databases for their different purposes and with extensively overlapping data
requirements that are nonetheless not coordinated, so that those submitting data have to comply with
these and any state and private health plan data needs separately. Explicit explanation of how data
submitted to GTR will be used by FDA and CMS would be most welcome, so the system is not an
added data burden but instead a single database serving multiple purposes and so streamlining federal
data submission on genetic tests.
Overinclusion and redundancy
While it makes sense to list participation in proficiency testing and quality control national programs
and to designate a contact person, the level of detail seems unduly high. The nature of quality control
programs is constantly changing; moreover, many programs relate to multiple tests or by laboratory
rather than by individual tests. Asking for exactly which QC program is used for each test will require a
substantial effort centrally to generate an up-to-date set of pull-down items to select, and redundancy
test-by-test for those submitting data. This test-by-test data entry format will be particularly
burdensome for laboratories that do many tests; for them it would seem to make better sense to instead
link to a page they make public that shows how they are complying with proficiency testing and CLIA
certification requirements, which are done laboratory-by-laboratory rather then test-by-test. A single

entry per laboratory noting which QC programs it subscribes to might be more logical, obviating repeat
entries for the laboratories that offer multiple tests. This suggests the option of organizing the data
according to test and according to laboratory, with ability to connect between them, rather than only testby-test.
Regarding the contact person, the level of specificity about certification, licensing, and specialty seems
beyond the need of a list focused on what tests are available, and items associated with personnel are
sure to require regular updating and thus often be out of date and inaccurate. I don’t think I would use
GTR to find an individual’s medical certification, but I would use it to find out who is the laboratory
director. Beyond an initial contact, however, I do not think GTR is the right place to list genetic
counselors, physicians, or others except to the degree they are the primary contact for information about
a particular test.
Regarding the details about which tissue samples, how samples are submitted, how linkage to
counseling services is maintained, and exemplars of negative, VUS and positive reports, a central
database seems a poor way to try to capture the information both because it will be constantly changing,
and because in many cases, the same procedures will be used for multiple tests. Another option would
be to require each laboratory offering a test to have a site that explains these details, and perhaps specify
the data that must be available, but not to require data entry test-by-test. That is, require a link to a
document that supplies stipulated information, but not directly incorporate the information centrally in
the database. This would leave the control of updates in the hands of the laboratories, and require
central management only of links rather than the constantly changing data.
The section on the condition tested (currently called “disease,” but that is probably going to encounter
some confusion for some testing, for example pharmacogenomic tests or broad screening tests). The
option for detail here is welcome, but GTR seems unlikely to be the place that information about
acronyms, mode of inheritance, disease mechanism, or prevalence are documented. A link out to
OMIM, where possible, would be much more logical, but it does not appear that OMIM categories are
cross-referenced here, so the data are apt to be redundant with OMIM, which is an actively archived and
curated dataset and more likely to be current with the literature than GTR will be (as a database about
what tests are available not about what is known about the diseases being tested). Explicit attention to
MESH categories and OMIM categories might reduce the amount of data entered and also connect users
to the relevant database. Finally, the information on individual mutations will be accumulating at
databases such as MutaDATABASE, Breast Cancer Information Core, Human Gene Mutation Database,
and Human Variome database. Linking to these rather than duplicating the information in them seems a
wiser strategy, and these databases are much more appropriate archives for the basic data on which
interpretation of tests will be based, whereas GTR will be the database of clinical tests and testing
services.
In the section on “method,” this seems likely to be especially fast moving and liable to change, and the
categories will be shifting to a degree that may make a central database hard to maintain. And yet this
information is exactly what someone contemplating ordering a test will want to know. My comments
here are not apt to be as useful as those from active laboratory directors, but this does seem both very
difficult and yet essential to have in a useful database. I will say that some of the level of detail implied
by the mock-up (down to the level of reference sequences and which variants) seems impractical. In
thinking of CF or BRCA testing, for example, there would be thousands of variants listed and it seems

unreasonable to expect that list to be maintained and up to date in a central database unless the database
maps directly to one that is operational at the testing service, and automatically updated as test methods
change. For sequence-based tests, the “all variants” item would work, but for multi-allele methods,
there are many tests now that have dozens or even hundreds of variants. Is GTR the right place to try to
keep track of this? The answer may be yes, but it may also be no. One solution might be to require a
link to a page that would describe the test method that would be under control of the laboratory doing
the testing, with specifications of the level of detail that those offering tests are expected to provide, but
not trying to keep all this information in the central database.
The information about how VUS and other categories of interpretation would be most welcome to know,
but seems more likely to arise more from other databases like the variome database, Mutadatabase, and
human gene mutation database than from GTR. The part that would be essential to capture is how the
lab bases its interpretation, and how to find the data and algorithms on which the interpretation is based.
If there is an interpretive algorithm, then link to where that can be found. The questions about whether
additional family members are invited for analysis (although right now, the only option is “without
charge”) seems excellent, as well as information about recontact if interpretation changes. These
answers of course have legal ramifications, but this seems like a reason to include, rather than exclude
them.
There will surely be a debate, however, about whether GTR should be the place where such information
is archived and documented. These are generally not laboratory-specific features, but about the nature
of the test in general, and of course answers to such questions entail liability and spill over to regulatory
compliance (under FDA, CLIA, New York State, etc.). The information about performance
characteristics and clinical validity and utility are welcome, in part for these very reasons. Asking each
laboratory to state utility and validity and cite the basis for it would go a long way towards
accountability. These items would also mean the federal government would be able to readily note the
conditions under which tests are being offered, including medical claims.
Missing elements about patents and licensing
Huys, et al., in their analysis of genetic testing for 22 commonly tested conditions, found at least one
blocking claim in patents associated with genes for 16 of the conditions (Nature Biotechnology 29: 903909, October 2009). Only a small fraction of claims were blocking, and most that were hard to work
around were method claims, many of which may be invalid under shifting jurisprudence in the United
States. But nonetheless, one fair interpretation of their findings is that patenting and licensing could
affect availability of 73 percent of Mendelian conditions commonly tested.
Cho, et al., surveyed laboratories offering genetic tests a decade ago, and found that 65 percent had been
contacted about patent enforcement (J Molec Diagnostics 5: 3-8, February 2003). Indeed, their
sampling strategy started from genetests.org, indicating that GTR is the logical locus for identifying
when intellectual property considerations might be relevant. In eight case studies of genetic testing for
ten clinical conditions prepared for the Secretary’s Advisory Committee on Genetics, Health and
Society, the evidence about effects of patenting and licensing on clinical access to tests was complex
and often equivocal; but the evidence that patenting and licensing matter to which laboratories offer
which tests was overwhelming and unequivocal. This work was summarized (Nature 458: 405-406, 26
March 2009; and Nature Biotechnology 28: 784-791, August 2010) and the case studies published
(Genetics in Medicine 12 (Suppl): S1-S211, April 2010).

It is thus quite clear that one reason a test may or may not be offered is patent rights. Moreover the
locus of responsibility for managing it will be at the level of laboratories. Just as other items on the list
seem intended to assist in establishing accountability in genetic testing, it seems odd to leave licensing
status entirely off the list of GTR data items.
The GTR is not the logical place to gather the detail information about patent status of relevant genes, or
terms of licensing. It does seem appropriate to have an acknowledgement of and link to further
information (where relevant) and patents and licensing. A simple check-box about whether patent rights
are licensed, with a link to a laboratory-maintained page that lists patenting and licensing status when
the box is checked (e.g., the list of patents, or public statement about licensing status) would both be
simple to submit, would require only disclosure of information that a service offering a test should know
about without undue research, and would greatly increase transparency and inconsistency in reporting
that directly affects which tests are available from whom.
Since the function of intellectual property, like property, depends on defining metes and bounds, it
seems remiss to leave out any mention of patents and licensing rights in a database of genetic tests given
the evidence that such rights directly affect which tests are available from which laboratories. Indeed
leaving such information out seems to signal acquiescence to patent-holders not disclosing their patents
and laboratories continuing to indulge in wink-and-nod avoidance of intellectual property conditions for
some genetic tests. At the least, if the decision is not to include such a check-box and link to further
information, then an explicit reason for not doing so seems warranted, and this letter is an invitation to
provide that justification.
This has obvious implications for infringement liability, but many other items being included in the
database also have liability implications. The patenting and licensing situation for some tests can be
murky, but this is in no small part because of failure in transparency for which GTR could be a partial
remedy. To the degree that intellectual property considerations operate in the shadows, it breeds
uncertainty and inefficiency. Patenting and licensing are at least as relevant to genetic testing as many
other items on the list, and demonstrably relevant to whether genetic tests are available from a given
laboratory. Moreover, if the legal regime under which genetic testing is to operate is to be both
transparent and also respect intellectual property rights where they exist, the logical place to effect the
transparency is where the test is publicly offered, i.e., the GTR. If a decision is made not to include this
information, the reasons for the decision should be explicit and public.
Coordination with other federal (and state) agencies that require data submission
CLIA certification seems essential. But presumably such CLIA certification comes with certain
information submitted to CMS; will GTR be linked to such data, reducing the need to submit redundant
information and keep it updated in two places? Regarding state licensing, it is not clear whether the
information requested is for the lab or the lab director or both. And of course, the special requirements
of New York State introduce yet another potential data redundancy. GTR could become the repository
of relevant information for the federal government, serving this function for CMS as well as NIH, but it
could also simply link to a separate CMS database (or the laboratories’ own records of CLIA
certification), but it would be good to avoid outright duplication.

The information about FDA status also seems essential, analogous to CLIA certification. The same
issues of coordination arise. Asking if the test is FDA-approved as a kit, an IVDMIA, or (in the future)
LDT or is under investigational use or is research use only is a core element, and should be required.
For tests that a laboratory administers but was not responsible for FDA review, however, there needs to
be an option (e.g., if they offer a kit test manufactured elsewhere).
Again, thank you for the opportunity to comment. I can be reached at gelp@duke.edu; 919 668-0790 if

I have been unclear or if you have questions or need more information.

Sincerely,

Robert Cook-Deegan, MD
Director, Center for Genome Ethics, Law & Policy,
Institute of Genome Sciences & Policy
Research Professor of Public Policy Studies and of Biology
Research Professor of Medicine
Duke University

HUMAN GENETICS SOCIETY OF AUSTRALASIA
ARBN. 076 130 937 (Incorporated Under the Associations Incorporation Act)
The liability of members is limited

Telephone: 02 9669 6602 Fax: 02 9669 6607
Email: secretariat@hgsa.org.au
PO Box 6012, Alexandria, NSW 2015
ABN No. 17 076 130 937

NIH GTR RFI Comments,
National Institutes of Health,
Office of Management and Budget,
Office of Regulatory Affairs
6705 Rockledge Drive,
Bethesda, MD 20892.
OIRA_submission@omb.eop.gov
December 23, 2011
Dear Sir,
Re: Comments on the practical utility of proposed data collection for the Genetic Testing Registry
The Human Genetics Society of Australasia (HGSA) is grateful for the opportunity to further comment on the
development of the Genetic Testing Registry (GTR). The HGSA is the peak Australasian professional body to
provide a forum and, in some instances, certification for the various disciplines related to human genetics in
Australia and New Zealand. As with our previous comments regarding the GTR, we are submitting this
response with another Australasian organization with a professional interest in genetic testing, the Royal
College of Pathologists of Australasia (RCPA). The RCPA is the peak body representing pathologists and
scientists in the delivery of medical testing in Australasia, has a major role in developing accreditation standards
for laboratories, and is a partner in the assessment of medical testing laboratories in Australasia.
We applaud the concept of the GTR as a resource for patients, referrers, laboratories, and (potentially) funders
to guide the utilization of useful, cost-effective, high-quality genetic tests in healthcare. The Australian
Government has commenced a discussion on similar issues, and we are following the development of the GTR
story with keen interest.
The broad range of stakeholders in genetic testing, and the variety of genetic competencies within each
category of stakeholder, makes it challenging to provide a resource that will be useful and contemporary for
everyone. This challenge is compounded by the GTR being a repository of data that is provided on a voluntary
basis by laboratories.
The PDF screenshots of the proposed interface indicate that a lot of detailed information will be provided by
each laboratory for each test. We are concerned that many laboratories will not volunteer to provide this level of
detail, and that the detail will render the information inaccessible to many potential users. The use of free-text to
describe a test and its utility provides freedom of expression and allows a laboratory to provide detail or
nuances, but free text can obscure the very facts that the user is seeking. And the facts that one user seeks
may be different to those sought by another.
We submit that GTR would be enhanced by the following:
o The target audiences need to be clearly defined. Rather than having one interface meeting all needs, it
may be appropriate to have interfaces for both the professional and the consumer
o Information needs to be presented in different tiers, with the principal information being on the surface
and more detailed information below. The screenshots already have this structure, with the overview
providing a “one screen” summary. But the language in this overview is too dense for most non-experts
(consumer and professional alike).
o Where possible, the key message for a field should be selected from a list, thereby ensuring ease of
reading (consistent layout and language) and searching. This summary statement could then be linked
to more detailed free-text statement that captures the detail.

o

There should be a central resource of statements regarding the validity and utility of certain tests that
any laboratory can utilize. If a gene test for a certain mutation is offered by a number of laboratories, the
validity and utility of the test should be independent of the laboratory providing the test (assuming that
each laboratory has appropriate Clinical Laboratory Improvement Amendments (CLIA) certification). But
the proposed screenshots indicate that each laboratory needs to draft its own statement regarding
these parameters. It seems inefficient to have each laboratory replicate this task.

The considerations above are “high level”. We have the following more specific comments.
o

We appreciate that not all options on drop down boxes or lists are visible, making it difficult to determine
the utility of the information collected and the burden of data entry on the respondents. We have
assumed the full list of options is as described in GTR Proposed Field Definitions V 0.25.

o

It appears the Personnel, and Licensure and Accreditation fields are built around North American
qualifications and certifications. It would be a pity not to collect accreditation information from
International laboratories, especially since ISO15189 accredited labs are required to ensure referral
laboratories and consultants are competent to perform requested examinations. We note over 1/3rd of
the laboratories in Gene Tests are outside the USA and an International resource such as the GTR
should capture accreditation information from non-CLIA labs.

o

The Ordering information fields are difficult to follow and not intuitive, they refer to URLs for detailed
ordering instructions in several places but in reality there is probably only one URL for the laboratory
which is required.

o

Under the Methodology section, collection of detail on platforms and instruments would appear to be
unnecessary for users of the GTR. Multiple instruments may be used for some testing. The section on
listing all relevant gene variants and clinical significance of each would be impossibly arduous unless it
was automatically filled from locus specific databases.

o

The interpretation fields include policy on reporting Variants of Unknown Significance. This could be
extremely complex and take a great deal of time for high level personnel to complete. It is unclear what
is meant by the final field in this section. We are also uncertain whether this will be useful. The experts
who would understand this section would only rely on it if they could be assured that it was current and
comprehensive. Given the voluntary nature of submissions, this is unlikely and the experts will contact
the laboratory directly. Non-experts could simply be confused by this section.

o

The whole section on Performance Characteristics is very complex and often likely to be contextspecific. Clinical validity and utility seems to be targeted at Direct-to-Consumer Testing. Clinicians
using the GTR should already have their reasons for ordering the test and what they want to do with the
results.

o

Potentially useful information not in the proposed collection includes price and turnaround times.

o

Import of at least minimal data sets from existing databases should help to reduce the burden of data
entry, however, we suggest there should be ability to import from other national or regional databases of
genetic tests.

Overall the HGSA and RCPA continue to support the concept of the GTR as a resource for clinicians, laboratory
staff, researchers and the public. However, the burden of completing many of the optional fields appears to
outweigh the utility of this information for users. We submit that it would be preferable to have simpler resource
of consistent high quality than something more ambitious of variable quality.
Thank you for considering our views.
Yours Sincerely
Prof Yee Khong
President RCPA

A/Prof Kevin Carpenter
President HGSA

December 22, 2011
NIH GTR Comments
National Institutes of Health
Office of Science Policy
6705 Rockledge Dr., Room 750
Bethesda, MD 20892

Dear GTR Staff,
The National Society of Genetic Counselors (NSGC) is responding to the National
Institutes of Health’s (NIH) November 23, 2011 request for comments on the practical
utility of the proposed collection of information for the Genetic Testing Registry (GTR).
NSGC supports efforts to enhance access to information regarding the availability,
validity, and utility of genetic tests. Additionally, NSGC appreciates the previous work
that NIH has conducted on accessibility and we encourage NIH to continue to seek out
genetic counselors and NSGC for expertise in fine-tuning the GTR. Further, we
respectfully request that NIH consider the following comments and recommendations of
NSGC.
General recommendations:
 NSGC requests that NIH safeguard against false or inaccurate GTR data by
implementing a structured peer-review process.
While the GTR seeks to increase transparency by establishing a general clearinghouse
for genetic data, more information is not always better. The volunteer-based
submission criteria make the GTR vulnerable to erroneous and false data that can be
misconstrued as valid and accurate information. If the GTR links to marketing
materials on a company or laboratory’s website, there should be disclosure that the
viewer is being directed to an external website.
Further, NSGC is concerned that tests that have little or no clinical validity will gain
credibility simply by being listed on the GTR. The increased transparency and
information available should be weighed against the quality of the data submitted
through a peer-review process. If formal peer-review is not feasible, then NSGC
recommends that NIH enable providers and other laboratories to submit external
comments to all publicized data within the GTR. In the absence of oversight of the
information presented, there should be notice that NIH, or any other government
entity, does not endorse the information presented on the GTR.

 NSGC recommends that the GTR provide adequate information on all tests.
The GTR should also include Clinical Laboratory Improvement Advisory Committee
certifications, as well as the common and commercial name for each test. Issues such as
unique product identifiers would help to identify and compare tests, though no current
regulations exist that mandate such standards. An example of useful information that should
be included is population or demographic data to ensure that tests are appropriately applied to
those populations.
 NSGC recommends that the GTR focus on the audience that will benefit most from the GTR.
It is not possible to create a resource that is equally valuable to all audiences. While
consumers, researchers, providers, payers, and policy makers may have some overlapping
needs, many are distinct. For example, clinicians may value and correctly interpret accuracy
and analytical sensitivity/specificity, but consumers may misinterpret these data elements as
clinical validity. Even genetics professionals and non-genetics professionals have
substantially different needs. Genetics providers will likely use this service as they used
GeneTests – a resource for genetic test availability and application. Non-genetics providers
and the public may use the GTR to infer validity.
Payers could find the GTR helpful if it addresses clinical validity in addition to analytical
validity, as they seek guidance on covering certain genetic tests. However, the GTR, in its
current form, will not be sufficient for payers’ purposes because it cannot assess the
circumstances under which testing should be offered and covered on a case-by-case basis.
NSGC appreciates the opportunity to provide comments. We look forward to collaborating with
NIH to ensure that the GTR is a valuable tool for genetics professionals.
Sincerely,

Karin M. Dent, MS, LCGC
President