Mr. Chairman and members of the Subcommittee, I am Dr. Edward Sondik,
Director of the National Center for Health Statistics (NCHS), of the Centers for
Disease Control and Prevention (CDC). I also serve as the Senior Advisor to the
Secretary on Health Statistics, and in that capacity I provide technical and
policy advice on statistical and health information issues that affect the
Department of Health and Human Services (HHS).
I am pleased to be with you today, and am particularly pleased that the
Subcommittee has chosen to address two important issues within the Federal
Statistical System - confidentiality and coordination among statistical
agencies.
Confidentiality and Use of Health Statistics
It is impossible to overstate the importance of confidentiality in the
Federal statistical community. Protecting the confidentiality of information
from our respondents is a fundamental value at NCHS and in all statistical
agencies. Not only do we have an ethical responsibility to our respondents, but
we recognize that our ability to obtain voluntary cooperation - and valid data
- is directly related to our ability to protect confidentiality. We must be able
to assure respondents that we can -- and will -- fully protect the
confidentiality of the information they provide.
The NCHS confidentiality protection, like that of other Federal statistics
agencies, has evolved over several decades. NCHS' authorizing legislation
includes carefully crafted legislative prohibitions against use for
non-statistical purposes, and requires us to obtain informed consent for the
uses we do make of data. NCHS has developed rigorous confidentiality practices
and safeguards to assure that our promises to respondents are kept. We are proud
of our strong record in maintaining confidentiality, as are our partners in
other statistical agencies, in public health, and in health research. We are
also gratified that these principles have been reinforced by the Federal
Statistical Confidentiality Order recently issued by the Office of Management
and Budget (OMB).
We also recognize that we have a responsibility to taxpayers to make maximum
use of limited funds, and a responsibility to our respondents to make sure that
their time and attention is used to full advantage. For this reason, we seek to
maximize the use of data that we do collect. We continue to develop new
approaches to making data available in the detail needed by researchers and
analysts, without compromising our responsibility to avoid disclosure of
information about individuals. Meeting ever greater data needs - for example, at
smaller geographic levels, and for more detailed subgroups of the population -
is a continuing challenge. In a moment, I will comment on the concept of sharing
information between protected statistical data centers as one solution.
Coordination of Statistical Programs
First, I want to briefly address two important and related topics -
coordination of efforts within our respective Departments and subject matter
areas - health, in the case of NCHS - and coordination across elements of the
Federal Statistical system. Each is important to an efficient statistical system
that produces analytically useful products.
Federal statistics serve multiple purposes. At the "macro" level, they are
important components of national economic analysis, providing information for
business and economic forecasting, and profiling the population and social
trends. They are also critical at the "micro" or program level, where they are
integral to the operation, evaluation, management, and accountability of
innumerable Federal programs. Within HHS, for example, statistics are used for
guiding medical research, tracking public health objectives, measuring progress
under civil rights laws, allocating grant funds, and safeguarding the Medicare
trust fund.
It is a particular responsibility of Federal statistical agencies (such as
NCHS, the Federal government's principal health statistics organization) to work
closely with programs, subject matter specialists, and policy-related offices
from within their respective Departments to assure that data needs are met. As
Director of NCHS and as Senior Advisor to the Secretary, I work closely with the
HHS Data Council in an ongoing effort to integrate statistical efforts within
HHS, and to bring a more strategic focus to meeting the information needs of our
programs and initiatives.
Each of the statistical agencies also has a distinct role in coordinating
statistical efforts across Departments as well, and I have been impressed as a
relative newcomer by the enthusiasm of my statistical agency counterparts for
strengthening ties among our agencies. Despite the different subject matter
addressed by these agencies, the "culture" and interests of our parent
Departments, and different authorizing statutes that have made each of the
Statistical agencies somewhat unique, we are anxious to coordinate efforts
across Departments, promote system-wide efficiency, and minimize duplication of
effort. I feel that the strength of a decentralized system in which data is
collected close to its use, assuring relevance, is a solid foundation to build
on in tapping into each other's unique expertise, resources, and technologies.
Data Sharing Among Statistical Agencies
Many of the most important interagency initiatives can only be accomplished
with legislative changes proposed by the Administration (the "Statistical
Confidentiality Act"), and introduced by Chairman Horn and cosponsored by Rep.
Maloney in the 104th Congress (H.R. 3924).
Many of the confidentiality statutes currently in place were written
narrowly to address statistical agencies one by one, rather than to apply to the
system as a whole. As the Federal statistical system has evolved, the increasing
complexity of our efforts has made it clear that there are efficiencies and
analytic benefits that could result from greater flexibility.
As an example, NCHS' legislative mandate was drafted several decades ago to
cover virtually all of the important issues in health. This legislation was
accompanied by strong confidentiality protections. While both the broad mandate
for NCHS and those protections remain valid today, the analytic focus has
changed considerably. Thirty years ago, health statistics focused primarily on
vital statistics, health status, and medical treatment. Today, with health
constituting 13.6 percent of Gross Domestic Product, issues of health care
financing and economics, the health care delivery system, productivity and
performance have become relatively more important. While the preponderance of
our statistical efforts are still unique to the health field, the confluence of
interest between NCHS and its counterparts in economic statistics has grown
considerably. The Bureau of the Census (Census) and the Bureau of Labor
Statistics (BLS) conduct surveys of employers and businesses, and from the
health perspective we have a need to look at employer-provided health insurance.
We need new authorities to allow us to work together effectively, and the
Statistical Confidentiality Act would provide us with such tools.
Examples of how we could effectively use this new authority include:
- Improving the design and sampling of our surveys of populations - and, in
particular, hard to locate subpopulations, such as race/ethnic groups, using
more detailed information and samples from the Bureau of the Census.
- Improving our health-related data on businesses, starting with the types
of health insurance benefits provided by employers. We are beginning to work
with BLS, Census, and other agencies on an overall review of data needs and
survey mechanisms in this area, and the ability to share sampling frames and
other resources will provide new opportunities for interagency collaboration.
- Improving our ability to assess the supply-side of the health care
system, including the types and distribution of health care providers and their
capacity. Since these providers are also businesses, legislation providing for
sharing among data centers would allow us to better use information already
collected by BLS and Census.
- Developing new longitudinal studies of children, beginning at birth, that
can address the interaction of issues such as child health, development, and
education. With legislation permitting data sharing, NCHS and the National
Center for Education Statistics (NCES) can more readily ccal
Confidentiality Act and the recently issued Federal Statistical Confidentiality
Order establish uniform standards for maintaining the confidentiality of
statistical information. We are confident that data could be shared, under these
provisions, in controlled, limited ways that will provide the fullest
protections to our respondents. We will carefully assess the reasons for, and
risks from, sharing information with other statistical agencies, and proceed
only with those that we would be comfortable explaining in full to our
respondents.
Related Confidentiality Issues
Before closing, I want to comment briefly on an aspect of confidentiality
that is of great importance to those of us in the health statistics, public
health, and research communities. From its longstanding work in this area, the
Subcommittee is well aware of the need to address fundamental issues in
protecting the privacy of medical records, and I expect that this issue will be
before the Subcommittee again in the nould allow important collaboration between the smaller agencies, such
as the NCHS/NCES example described previously.
I also want to emphasize the importance of enacting not only broad data
sharing principles, but also specific conforming amendments included in the
Administration's proposed Statistical Confidentiality Act- Without these
detailed changes in existing law, the promise of data sharing will not be
realized in practice.
Finally, it is important to note that the proposed Statistical
Confidentiality Act and the recently issued Federal Statistical Confidentiality
Order establish uniform standards for maintaining the confidentiality of
statistical information. We are confident that data could be shared, under these
provisions, in controlled, limited ways that will provide the fullest
protections to our respondents. We will carefully assess the reasons for, and
risks from, sharing information with other statistical agencies, and proceed
only with those that we would be comfortable explaining in full to our
respondents.
Related Confidentiality Issues
Before closing, I want to comment briefly on an aspect of confidentiality
that is of great importance to those of us in the health statistics, public
health, and research communities. From its longstanding work in this area, the
Subcommittee is well aware of the need to address fundamental issues in
protecting the privacy of medical records, and I expect that this issue will be
before the Subcommittee again in the near future.
There is a growing consensus that the lack of any uniform nationwide privacy
protections for records in the health care system is increasingly problematic,
and that a new era of electronic medical records systems poses potential risks
to individuals. New impetus for addressing this issue has been provided from
enactment of the Health insurance Portability and Accountability Act of 1996
(P.L. 104-191), which mandated the development of new Federal privacy
protections in this area.
As mandated by P. L. 104-191, HHS is nearing the end of a careful review of
this issue, based in part on input from public hearings held by the National
Committee on Vital and Health Statistics. Recommendations for privacy
legislation will be made by the Secretary at the end of August.
The implications of legislation in this area are immense, both in providing
for the protection of individuals and in providing for important and appropriate
statistical, research, and public health uses of medical information. As you are
aware, many of the advances in the science of biomedical research, the detection
and control of diseases, and advances in our health care system have come from
the aggregation of individual medical records. I want to emphasize the
importance of achieving dual objectives with this legislation: assuring that
privacy is protected, and assuring that we maintain our ability to provide
answers to important health questions through carefully controlled access to
records. I am looking forward to working with the Subcommittee as it considers
the Secretary's recommendations.
Again, I thank the Subcommittee for the opportunity to discuss these issues.