Attachment_D.pdf

National Program of Cancer Registries

Cancer Surveillance System (NPCR-CSS)

Data Release Policy

December 2006

Cancer Surveillance Branch Division of Cancer Prevention and Control

NCCDPHP, CDC, Mailstop K-53 Atlanta, Georgia 30341-3717

Phone: 770-488-4783 E-mail: cancerinfo@cdc.gov (specify “NPCR-CSS” in subject line) mailto:cancerinfo@cdc.gov

Table of Contents

Page

Acknowledgement 3

I. Introduction 4

II. Assurance of Confidentiality 4

III. Overview of the Data 5

IV. Requests for State or Local Data and Notification of States 6

V. Public-Use Data Sets 7

A. United States Cancer Statistics (USCS) Data Set 8

B. USCS Expanded Data Set 8

C. USCS County Cancer Incidence Data Set 8

D. Public Release Disclosure Statement 8

VI. Restricted-Access Data Set 9

A. State Level Restricted-Access Data Set 9

B. Regional Level Restricted-Access Data Set 9

VII. Emergency and Provisional Data Releases 10

VIII. Data Release Under Controlled Conditions 10

IX. References 11

Table 1 Comparison of NPCR-CSS Data 12

Appendices

Appendix A NPCR-CSS Scientific Working Group Members 16 NPCR-CSS Small Data Release Working Group Appendix B NPCR-CSS 308(d) Assurance of Confidentiality 18 Appendix C NPCR-CSS 308(d) Assurance of Confidentiality Frequently Asked Questions 19 Appendix D NPCR-CSS Overview of Data Security 23 Appendix E State and Metropolitan Area Cancer Registries by Federal Funding Source and First Diagnosis Year for Which Cancer Cases Were Reportable to CDC’s NPCR or NCI’s SEER Program 25 Appendix F NPCR-CSS Data Sharing Agreement 27 Appendix G Data Items for Restricted Access Data Sets 29

NPCR-CSS Data Release Policy

Acknowledgement

The Cancer Surveillance Branch, Division of Cancer Prevention and Control, NCCDPHP, CDC, wishes to acknowledge and thank the members of the state-based NPCR-CSS Scientific Working Group for their contributions to the concepts included in this document. In 2005 and 2006, the members of the Small Data Release Working Group played an important role in the review and release of the USCS expanded data set on CDC WONDER and the restricted-access data sets.

Appendix A lists the members of both working groups.

NPCR-CSS Data Policy

Cancer Surveillance System

Data Release Policy

I. Introduction This document describes the format and content of data that the National Program of Cancer Registries’ Cancer Surveillance System (NPCR-CSS) will release or share. This multi-year policy updates the October 2005 NPCR-CSS Data Release Policy. This policy applies to data submitted to Centers for Disease Control and Prevention (CDC) for the January 2007 data submission and for all future data submissions until a new policy is provided.

The NPCR-CSS Privacy Steward (Mr. Joseph Rogers) will clear all releases of state data, ensuring that the data are released according to the terms of the NPCR-CSS Data Release Policy.

It is possible that, in future years, data release practices or the content and format of released data may vary from those described in these guidelines. Such changes may occur as a result of improvements in the quality of the data, changes in information technology, and evolving data needs. However, if such variations occur, they will provide comparable protection (or more protection) for patient confidentiality than is described in this policy. If it is anticipated that any data will be released with less protection (as determined by the NPCR-CSS Privacy Steward) for patient confidentiality than is described in this policy, NPCR states will be notified and will have ample time to respond before the data are released. This policy will be reviewed annually by the NPCR-CSS Privacy Steward and other appropriate CDC staff members to determine whether revisions are needed. If revisions are needed, NPCR states will be notified and allowed to review and comment on the revisions before they become final.

II. Assurance of Confidentiality All data collected and maintained by NPCR-CSS must be managed, presented, published, and released in accordance with strict attention to confidentiality and security, consistent with the general principles and guidelines established by CDC for confidential case data1-3 and specific restrictions imposed on the NPCR-CSS data (Appendices B, C, and D).4 Special care is needed even for cancer incidence data that are not directly identifiable because geographic and small cell data may be indirectly identifying when combined with detailed information in case reports or from laboratory, medical records, or linkage with other data files.5-10

NPCR-CSS has approval for protection under section 308(d) of the Public Health Services (PHS) Act (42 USC 242m(d)) (Appendices B and C). The 308(d) confidentiality assurance protects identifiable and potentially identifiable information from being used for any purpose other than the purpose for which it was collected unless the person or establishment from which it was obtained has consented to such use. This assurance protects against disclosures under a court order and provides protections that the Privacy Act of 1974 (5 USC 552a) does not. For example, the Privacy Act of 1974 protects individual participants, but the 308(d) confidentiality assurance also protects institutions. Confidentiality protection granted by CDC promises participants and institutions that their data will be shared only with those individuals and institutions listed in the project’s consent form or in its specified policies.

III. Overview of the Data In 1992, Congress established the National Program of Cancer Registries (NPCR) by enacting the Cancer Registries Amendment Act, Public Law 102-515.4 The law authorized CDC to provide funds and technical assistance to states and territories to improve or enhance existing cancer registries and to plan for and implement population-based central cancer registries where they did not exist. NPCR’s purpose is to assure the availability of more complete local, state, regional, and national cancer incidence data for planning and evaluation of cancer control interventions and for research. NPCR adopted reporting requirements and definitions consistent with the National Cancer Institute’s (NCI's) Surveillance, Epidemiology, and End Results Program (SEER);11,12 required the use of uniform data items and codes and record layouts as defined by consensus of members of the North American Association of Central Cancer Registries (NAACCR);13 and established standards for data management and data completeness, timeliness, and quality similar to those recommended by NAACCR.13,14 In 1994, the first 37 states received funding from CDC. Currently, 45 states, three territories, and the District of Columbia are funded by NPCR (Appendix E).15,16 NPCR-funded state registries collect data on patient demographics, primary tumor site, morphology, stage of disease at diagnosis, and first course of treatment. In addition, NPCR registries conduct follow-up for vital status by linking with state and national death files.

Invasive and in situ cancer case reports are submitted to CDC by population-based statewide central cancer registries in all 45 participating states, the District of Columbia, and three territories. In each state or territory, state laws and regulations mandate the reporting of cancer cases by facilities and practitioners who diagnose or treat cancer to the state health department or its designee.4 The state cancer registry receives case reports from facilities and practitioners throughout the state and processes them according to standard data management procedures.14 Personal identifiers including the patient’s name, social security number, and street address are removed prior to the encryption and electronic transmission of these case reports to a contractor acting on behalf of CDC. CDC and the contractor adhere to strict data security procedures when receiving, processing, and managing the data (Appendix D). For more information on the NPCR- CSS data, see the Technical Notes as posted on the United States Cancer Statistics Web site (http://www.cdc.gov/uscs) which is updated annually. NPCR-CSS received formal approval (protocol #2594) from CDC’s Institutional Review Board (IRB) in October 1999. The approval is updated annually. CDC has an Office for Human Research Protections (OHRP)-approved federal-wide assurance of compliance with rules for the protection of human subjects in research (45 CFR 46) (http://ohrp.osophs.dhhs.gov/nsearch_t.htm).

State cancer registries and federal agencies routinely publish cancer incidence data between 22 and 25 months after the close of each diagnosis year based on data that meet standards for completeness and quality.16,17 However, other versions of the same data, based on the data file as it exists at different time periods, are usually available. For example, some states have preliminary data available as soon as 12 months after the close of each diagnosis year. Even after the publication of official statistics, state cancer registries (and CDC and NCI) continue to update and re-publish data with new information incorporated. Thus, when cancer incidence data are published it is common practice to document either the data submission date (i.e., when the data were submitted to CDC or NCI) or the date the file was prepared. Changes in state cancer registry incidence data that occur more than 22 to 25 months after the close of a diagnosis year are likely to be small; however, delays in reporting are more likely to impact certain cancer sites and may be important for some research studies.18 http://www.cdc.gov/uscs http://ohrp.osophs.dhhs.gov/humansubjects/guidance/45cfr46.htm http://ohrp.osophs.dhhs.gov/nsearch_t.htm

IV. Requests for State or Local Data and Notification of States The Freedom of Information Act (FOIA) (http://www.cdc.gov/od/foia/foi/htm) generally provides that, upon written request from any person, a Federal agency (i.e., CDC) must release any agency record unless that record falls (in whole or part) within one of nine exemptions.

FOIA applies to only Federal agencies, and covers only records in the possession and control of those agencies at the time of the FOIA request (except in certain narrow instances involving grantee-held data). Because State based data becomes a Federal record in CDC’s possession, such records are subject to disclosure in response to a FOIA request. The FOIA exemptions that may be available to protect some aspects of State data from public disclosures in response to a FOIA request are:

1) Exemption 3, which specifically exempts information from disclosure by statute (in this instance pursuant to an Assurance of Confidentiality under Section 308(d) of the Public Health Service Act), and

2) Exemption 6, which exempts from disclosure personnel and medical files and similar files the disclosure of which would constitute a clearly unwarranted invasion of personal privacy.

In general, non-FOIA requests to CDC from the public, the media, and other government agencies for local cancer incidence data are referred to the state health department for a reply.

There are three reasons for this: (1) the state health departments can release cancer incidence data in accordance with locally established policies and procedures and consistent with provisions of the 1992 Cancer Registries Act, Public Law 102-515; 4 (2) the relative infrequency of data submission to federal agencies assures that the state health department or its designated cancer registry will have the most complete, accurate, and up-to-date information; and (3) the state registry may be able to provide more detailed data that can better meet the needs of the requestor. When the request is for data regarding cancer incidence involving more than one state, CDC will refer the requestor to published reports or to NPCR-CSS data sets that are released in accordance with practices described in this document. At this time, it is anticipated that two kinds of data sets will be released: public-use data sets and restricted-access data sets (see definitions below).

CDC staff members or contractors perform analyses of NPCR program data as needed, including assessment of the completeness, timeliness, and quality of cancer incidence data; and analyses of the cancer burden as needed for meeting national cancer control objectives. Such analyses of state and national data are conducted routinely by federal agencies, including CDC, for programmatic or statistical purposes as needed to achieve the agency's mandate. Publications or presentations describing the quality of the data or the burden of cancer may be one outcome of such analyses. Examples of topics for such analyses are descriptive analyses by racial and ethnic populations for specific cancers and descriptions of cancer incidence trends. In compliance with the 308(d) Assurance of Confidentiality, CDC employees and contractors are required to handle the information in accordance with principles outlined in the CDC Staff Manual on Confidentiality and to follow the specific procedures documented in the NPCR-CSS Confidentiality/Security Statement (Appendices B, C, and D).

In addition to adhering to strict requirements for protecting confidentiality, CDC staff members will notify the state cancer registry in advance whenever they plan to present, publish or release state-specific information on cancer incidence that have not been previously presented, published, or released. This notification will include, when possible, sending a pre-publication

NPCR-CSS Data Policy December 2006 http://www.cdc.gov/od/foia/foi/htm copy of the entire publication or other information to the specific states. When that is not possible (for example if the information is embargoed), the specific state cancer registries will receive a summary of the information before it is published or released. In addition, CDC staff members are required to acknowledge state cancer registries whenever NPCR-CSS data are presented, released, or published by CDC by making available the following (or similar) statement:

These data were provided by cancer registries participating in the National Program of Cancer Registries (NPCR) and submitted to CDC in the (insert submission date) NPCR-Cancer Surveillance System data submission.

V. Public-Use Data Sets For purposes of this policy, public-use data sets (PUDS) are defined as data sets that are comprised of aggregated data (i.e., not individual case-specific data or microdata) that have been modified as needed, according to accepted procedures, to block breaches of confidentiality and prevent disclosure of the patient’s identity or the patient’s confidential information. 2, 5-10 A combination of confidentiality protection measures is employed for each PUDS (Table 1). PUDS will not contain information that is identifiable or potentially identifiable according to currently accepted procedures for reducing disclosure risk. 2, 5-10 One of the PUDS is available as a Web-based query system and has a database behind a CDC firewall that is case-specific microdata;

however, even for this database users will be able to access only aggregate counts and rates with all confidentiality protections built in. Before each PUDS is finalized, the aggregate values will be analyzed to determine whether there is a need for complementary cell suppression.2, 5-10 If appropriate, the analysis will include consultation with a statistician with specific expertise in statistical disclosure limitation techniques. Following the analysis, complementary cell suppression will be applied as needed.

There will be no restrictions on access to PUDS. A public release disclosure statement (see page

9) will caution users against inappropriate use of the data or inappropriate disclosure of information. PUDS will be released as delimited ASCII files and/or on a Web-based query system, or possibly through other vehicles (Table 1). States will have an opportunity to review their state's data before each PUDS is released and have adequate time to notify CDC if they see a problem with the data. As a convenience to NPCR state registries, states may request from CDC a copy of their complete state-specific analytic database that is used to create each PUDS.

The following three PUDS are currently being released:

• United States Cancer Statistics (USCS) data set,

• USCS expanded data set, and

• USCS county cancer incidence data set.

All NPCR-CSS PUDS will consist of cancer incidence data selected from the NPCR-CSS analytic database. This is the same database that provides cancer incidence data for the annual publication of USCS.16 Data sources, case definitions, basic registry eligibility criteria in terms of required data completeness and quality, population denominator sources, methods for calculating incidence rates, and the rationale for specific cell suppression thresholds are as described in the Technical Notes for USCS unless noted in separate documentation that accompanies the PUDS.

Data from NCI’s SEER program are included in some, but not all, data releases described in this document. However, based on the underlying principles for coordinated nationwide cancer control as stated in the CDC and NCI Surveillance Memorandum of Understanding, the goal is to have all data sets include data from both NPCR and SEER.

Separate documentation may accompany each PUDS which will describe its unique features; for example, the data submission date, percentage of the U.S. population covered, diagnosis years and cancer sites included, variables included, any special data quality criteria required for inclusion, and any unique statistical methods employed.

A. United States Cancer Statistics (USCS) Data Set The USCS data set contains the same aggregate counts and rates for incidence and mortality published annually (Table 1). The PUDS is an HTML edition of USCS. Tables of male and female combined counts and rates are in the HTML edition in addition to the published sex-specific tables. Users can download the data in ASCII and MS Excel formats for use in other applications.

B. USCS Expanded Data Set The USCS expanded data set displays the aggregate counts, rates, and 95% confidence intervals published yearly in the USCS, plus additional aggregate values created from the same analytic file containing finer breakdowns of counts and rates based on selected variables (Table 1). This PUDS is available on WONDER (http://wonder.cdc.gov), a Web-based query system, that has a database behind a CDC firewall with case-specific microdata;

however, users will be able to access only aggregate counts and rates with all confidentiality protections built in. Because this PUDS will present data in more detail than is presented in USCS, states will have the option to notify NPCR if they prefer not to have their state data included. Planned enhancements include the addition of mortality data to the current system.

C. USCS County Cancer Incidence Data Set The USCS county cancer incidence data set consists of aggregate cancer incidence counts and rates for selected counties in the United States (Table 1). This PUDS is available as an ASCII file. Because this PUDS presents data at a sub-state geographic level, states have the option to notify NPCR if they prefer not to have their county data included. A limited version of this PUDS was released in 2004 to a small number of users, including the American Cancer Society for its Community Assessment "E-Tool" CD-ROM, NCI for the State Cancer Profiles project (www.statecancerprofiles.cancer.gov), and the U.S. Department of Health and Human Services' Women's Health Initiative Project. Future versions may contain more detail about cancer at the county level. Unless states are notified well in advance, this PUDS will always contain a cell suppression threshold that is at least as high as that used in 2004 with counts and rates averaged over at least three years. When states have accumulated more consecutive years of high quality data, we will routinely publish county data averaged over five years. It is anticipated that diagnosis years before 1997 will be added when resources permit and when data quality for those years has been tested and found to meet the specified criteria.

D. Public Release Disclosure Statement The following (or similar) public release disclosure statement will be prominently displayed for users of all NPCR-CSS PUDS:

Data Use Restrictions: Read Carefully before Using By using these data, you signify your agreement to comply with the following statutorily based requirements. The National Program of Cancer Registries (NPCR), Centers for Disease Control and Prevention (CDC), has obtained an assurance of confidentiality pursuant to Section 308(d) of the Public Health Service Act, 42 U.S.C. 242m(d). This assurance provides that

NPCR-CSS Data Policy December 2006 http://wonder.cdc.gov/ identifiable or potentially identifiable data collected by the NPCR may be used only for the purpose for which they were obtained unless the person or establishment from which they were obtained has consented to such use. Any effort to determine the identity of any reported cases, or to use the information for any purpose other than statistical reporting and analysis, is a violation of the assurance. Therefore users will:

• Use the data for statistical reporting and analysis only.

• Make no attempt to learn the identity of any person or establishment included in these data.

• Make no disclosure or other use of the identity of any person or establishment discovered inadvertently, and advise the Associate Director for Science, Office of Science Policy and Technology Transfer, CDC, Mailstop D-50, 1600 Clifton Road, N.E., Atlanta, Georgia, 30333, Phone: 404-639-7240) (or NCI’s SEER Program if SEER data) and the relevant state or metropolitan area cancer registry, of any such discovery.

VI. Restricted-Access Data Sets For purposes of this policy, restricted-access data sets (RADS) are defined as versions of the full NPCR-CSS analytic data set, either aggregated data or microdata (i.e., individual case-specific data), that have been modified as needed to minimize (but may not remove entirely) the potential for disclosure of confidential information. RADS will not contain personal identifiers such as a patient’s name, street address, or social security number as this information is not transmitted by state cancer registries to CDC as part of their annual data submission. However, they may contain information that is potentially identifiable especially when linked with other data sets, such as the occurrence of a rare cancer in a person of a certain age or racial or ethnic group. The following two RADS are described:

• Regional level restricted-access data set (rRADS)

• State level restricted-access data set (sRADS)

The list of the variables included in these RADS is in Appendix G.

Because restricted-access data sets may potentially contain identifiable information, states will have the option to notify NPCR if they prefer not to have their data included in either or both of these restricted access data sets.

A. Regional Level Restricted-Access Data Set (rRADS) The rRADS will be released after the requestor signs a detailed data sharing agreement and provides a description of the proposed project. The intent of the regional-level dataset is to provide access to individual case-specific data with reduced potential for disclosure of confidential information and correspondingly less oversight from CDC. The rRADS is similar to the sRADS, but with reduced geographic specificity and fewer data items.

Reducing geographic specificity is an effective way to reduce the potential for disclosure of confidential information.

B. State Level Restricted-Access Data Set (sRADS) The sRADS will be released only after CDC authenticates the requestor's identity and the requestor either signs a detailed data sharing agreement or signs a less detailed data sharing agreement and submits confirmation of IRB approval by an institution with appropriate assurance from the U.S. Department of Health and Human Services’ Office for Human Research Protections (http://ohrp.osophs.dhhs.gov/nsearch_t.htm).

To define and develop the RADS, CDC staff members are working collaboratively with a subset (Small Data Release Working Group) of the NPCR-CSS Scientific Working Group. The

NPCR-CSS Data Policy December 2006 http://ohrp.osophs.dhhs.gov/nsearch_t.htm resources needed for preparation, data quality testing, releasing, logging 2 and monitoring the release of restricted data sets have been and will continue to be substantial. Thus, release of these data sets will be phased in over several years. The sRADS was first made available to ACS, NCI and members of the Small Data Release Working Group starting in November 2005. In November 2006, the sRADS was made available to ACS, NCI, and all NPCR programs. In future years, if all goes as planned, the sRADS will be available to other researchers who meet the criteria jointly established by CDC and NPCR programs. These RADS are currently available as an ASCII file but eventually will be made available as a SEER*STAT data base.

User documentation is being developed and is facilitated by the availability of a NAACCR data dictionary for every diagnosis year since 1994.13 Detailed data sharing agreements, and procedures for user authentication and for logging and monitoring of data releases, are being developed. The recently completed CDC/ATSDR/CSTE Data Release Guidelines for Re-Release of State Data described in detail the recommended contents of such a data sharing agreement and provided guidance for user authentication and logging and monitoring of data releases.2 CDC staff members and partners who are working together to plan for the release of these data sets have reviewed these CDC recommendations and other relevant materials and developed data sharing agreements that meet the requirements of state and federal agencies. A similar approach will be used for developing methods for user authentication, logging of data releases, and compliance monitoring.

VII. Emergency and Provisional Data Releases It is not anticipated that CDC will ever need to release NPCR-CSS data before the files have been modified as needed to protect confidentiality as described in this policy. This is prohibited by the 308(d) Assurance of Confidentiality (Appendices B and C).

Provisional data and draft data tables will be shared with CDC employees and contractors, NPCR states, and other partners as needed in order to facilitate quality reviews of the data.

When appropriate, individuals who participate in such reviews will sign a data use agreement before accessing the data or tables.

VIII. Data Release under Controlled Conditions CDC-wide policy stipulates that a CDC program may consider release of data that cannot be released as either a PUDS or a RADS under certain controlled conditions.1 These controlled conditions may include a CDC-controlled data center such as the data center established at NCHS (http://www.cdc.gov/nchs/r&d/rdc.htm) or through special licensing. NPCR-CSS data will not be released under these controlled conditions while this policy is in place. Release of data under controlled conditions will be considered as part of discussions with partners toward development of RADS, and a determination will be made as to whether such releases of data will be considered in the future for NPCR-CSS data.

http://www.cdc.gov/nchs/r&d/rdc.htm

IX. References

1. Centers for Disease Control and Prevention. CDC/ATSDR Policy on Releasing and Sharing Data.

Atlanta: Centers for Disease Control and Prevention; 2003. Available at http://www.cdc.gov/od/foia/policies/sharing.htm.

2. Centers for Disease Control and Prevention. CDC/ATSDR/CSTE Data Release Guidelines for Re- Release of State Data. Atlanta: Centers for Disease Control and Prevention; 2003 (available upon request).

3. Centers for Disease Control and Prevention. CDC Staff Manual on Confidentiality. Atlanta: Centers for Disease Control and Prevention; 1984 and National Center for Health Statistics. NCHS Staff Manual on Confidentiality. Hyattsville MD: National Center for Health Statisitics; 1999.

4. Cancer Registries Amendment Act, Public Law 102-515, Stat. 3312 (October 22, 1992). Available at http://www.cdc.gov/cancer/npcr/npcrpdfs/publaw.pdf.

5. American Statistical Association. Privacy, Confidentiality, and Data Security Web Site. Alexandria, VA: American Statisitcal Association; 2005. Available at http://www.amstat.org/comm/cmtepc/index.cfm.

6. Doyle P, Lane JI, Theeuwes JM, Zayatz LM (eds). Confidentiality, Disclosure, and Data Access:

Theory and Practical Application for Statistical Agencies. Amsterdam: Elsevier Science BV; 2001.

7. Federal Committee on Statistical Methodology. Checklist on Disclosure Potential of Proposed Data Releases. Available at http://www.fcsm.gov/committees/cdac/cdac.html.

8. Federal Committee on Statistical Methodology. Report on Statistical Disclosure Limitation Methodology. (Statistical Working Paper 22). Washington (DC): Office of Management and Budget;

1994. Available at http://www.fcsm.gov/working-papers/spwp22.html.

9. McLaughlin C. Confidentiality protection in publicly released central cancer registry data. Journal of Registry Management 2002; 29(3):84-88.

10. Stoto M. Statistical Issues in Interactive Web-Based Public Health Data Dissemination Systems.

Draft report prepared for the National Association of Public Health Statistics and Information Systems, Rand Corporation; September 2002.

11. Surveillance, Epidemiology, and End Results Program. The SEER Program Code Manual. 3rd ed.

Bethesda (MD): National Cancer Institute,1998.

12. Percy C, Van Holten V, Muir C (eds). International Classification of Diseases for Oncology, 2nd edition. Geneva, Switzerland: World Health Organization; 1990.

13. Hultstrom D, editor. Standards for Cancer Registries, vol. II: Data Standards and Data Dictionary, version 9.1, 6th ed. Springfield (IL): North American Association of Central Cancer Registries; 2001.

14. North American Association of Central Cancer Registries. Standards for Cancer Registries, vol. III:

Standards for Completeness, Quality, Analysis, and Management of Data. Springfield (IL): North American Association of Central Cancer Registries; 2000.

15. Hutton MD, Simpson, LD, Miller DS, Weir HK, McDavid K, Hall HI, Progress toward nationwide cancer surveillance: an evaluation of the National Program of Cancer Registries, 1994-1999. Journal of Registry Management 2001;28(3):113-120.

16. U.S. Cancer Statistics Working Group. United States Cancer Statistics: 2003 Incidence and Mortality, Technical Notes. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute; 2006 (in press).

17. Method for estimating case completeness is available in: Wu XC, Hotes JL, Fulton PJ, Cormier M, Correa CN, McLaughlin CC, Kosary C, Howe HL, Chen VW, editors. Cancer in North America, 1995-1999, vol. I: Incidence. Springfield, IL: North American Association of Central Cancer Registries; 2002. Available at http://www.naaccr.org/Stats/CINAPubs.html.

18. Clegg LX, Fueur EJ, Midthune DN, Fay MP, Hankey BF. Impact of reporting delay and reporting error on cancer incidence rates and trends. Journal of the National Cancer Institute 2002;94(20):1537-45.

NPCR-CSS Data Policy December 2006 http://www.cdc.gov/od/foia/policies/sharing.htm http://www.cdc.gov/cancer/npcr/npcrpdfs/publaw.pdf http://www.amstat.org/comm/cmtepc/index.cfm http://www.fcsm.gov/committees/cdac/cdac.html http://www.fcsm.gov/working-papers/spwp22.html http://www.naaccr.org/Stats/CINAPubs.html

Table 1. Comparison of the National Program of Cancer Registries –Cancer Surveillance System Data Sets Overview

Public-Use Data Sets (PUDS) Restricted-Access Data Sets (RADS) USCS USCS Expanded USCS County Regional Level State Level

Format Database of aggregate counts and rates, with text documentation

Database of aggregate counts and rates, with text documentation. The database behind the CDC firewall is case-specific microdata.

Database of aggregate counts and rates, with text documentation

Database of individual record level data

Database of individual record level data

Mode of Access Web-based query system and downloadable ASCII file tables

Web-based query system Flat ASCII file and Web-based query system and separate brief text documentation;

www.statecancerprofiles.

cancer.gov

Flat ASCII file Flat ASCII file

Web Address or Contact Information

USCS website (www.cdc.gov/uscs)

CDC Wonder (http://wonder.cdc.gov)

Request from cancerinfo@cdc.gov

(specify “USCS County” in subject line)

Request from cancerinfo@cdc.gov (specify “Regional

RADS” in subject line)

Request from cancerinfo@cdc.gov

(specify “State RADS” in subject line)

Contains Potentially Identifiable Information

No

No Yes Yes

Registry Eligibility Criteria for Data Completeness and

Quality

USCS

publication criteria

USCS publication criteria;

data meet criteria for unknown county

USCS

publication criteria

USCS

publication criteria

When Available Updated 2007

Updated 2007 2007 Updated 2007

NPCR-CSS Data Release Policy December 2006 http://www.statecancerprofiles.cancer.gov/ http://www.cdc.gov/uscs http://wonder.cdc.gov/ mailto:cancerinfo@cdc.gov

Table 1. Comparison of the National Program of Cancer Registries –Cancer Surveillance System Data Sets, continued Cases Included

Public-Use Data Sets (PUDS) Restricted-Access Data Sets (RADS) USCS USCS Expanded USCS County Regional Level State Level

States/Territories

NPCR and SEER states that meet eligibility criteria

NPCR states that meet eligibility criteria*

NPCR states that meet

NPCR states that meet

Diagnosis Years

Single year of data from 1999-2004; combined years of data from 2001-2004 (prior diagnosis years may be added in future years)

Single year of data from 1999-2004 (prior diagnosis years may be added in future years)

Aggregated years of data from 2001-2004 (prior diagnosis years may be added in future years)

Single year of data from 1999-2004 (prior diagnosis years may be added in future years)

Single year of data from 1999-2004 (prior diagnosis years may be added in future years)

Cancer Sites All reportable invasive cancers; in situ female breast, and benign and borderline primary intracranial and central nervous system tumors (diagnosis year 2004)

All reportable cancer sites combined; female breast; in situ female breast; cervix uteri; colon and rectum; lung and bronchus; melanoma;

bladder; prostate; oral cavity and pharynx; brain and other nervous system; thyroid;

kidney; stomach; ovary;

corpus and uterus, NOS;

leukemias; non-Hodgkin lymphoma, liver and intrahepatic bile duct, pancreas, esophagus, and childhood cancers**

All reportable invasive and in situ cancers and benign and borderline primary intracranial and central nervous system tumors (diagnosis year 2004)

All reportable invasive and in situ cancers and benign and borderline primary intracranial and central nervous system tumors (diagnosis year 2004)

* Future plans may include the addition of SEER data similar to the USCS data set.

** Pending further data quality investigation and discussion with the NPCR-CSS Scientific Working Group.

Table 1. Comparison of the National Program of Cancer Registries –Cancer Surveillance System Data Sets, continued Variables Included

Public-Use Data Sets (PUDS) Restricted-Access Data Sets (RADS) USCS USCS Expanded USCS County Regional Level State Level

Geographic Levels All areas combined; U.S.

census region and division;

NPCR and SEER state or territory; SEER metropolitan area

All areas combined; NPCR and SEER state or territory;

MSA for cities of 500,000 or more* (additional levels may be added in 2007)**

NPCR state;

county* (additional levels may be added in 2007; i.e., Health Service Regions)**

United States Census Region (and Division**)

NPCR state or territory*

Race All races combined; white; black; Asian/Pacific Islander;

American Indian/Alaska Native (subpopulations may be added in 2007)**

All races combined; white;

black (additional groups or subpopulations may be added in 2007)**

All races reported

All races reported

Ethnicity (Hispanic) Yes May be added in 2007** Yes Yes Age Groups All ages combined and standard 5 year age groups for adults and <15 and <20 for childhood cancers

All ages combined and standard 5-year age groups that can be combined by the user.

May be added in 2007** Standard 19 age groups Individual ages and standard 5-year age groups

Summary Stage No No No Yes Yes

Histology International Classification of Childhood Cancers, Third Revision (all geographic areas combined) Mesothelioma (national and state level) Kaposi Sarcoma (national and state level) Consensus Conf on Cancer Registration of Brain and CNS Tumors (all areas only)

No Yes Yes

* Future plans may include the addition of SEER data similar to the USCS data set.

** Pending further data quality investigation and discussion with the NPCR-CSS Scientific Working Group.

NPCR-CSS Data Policy

Table 1. Comparison of the National Program of Cancer Registries –Cancer Surveillance System Data Sets, continued Confidentiality Protection/Disclosure Limitation Measures Employed

Public-Use Data Sets (PUDS) Restricted-Access Data Sets (RADS) USCS USCS Expanded USCS County Regional Level State Level

No Direct Identifiers

Yes

Yes Yes Yes

Aggregation

Yes Yes No No

Limited Number of Variables

Yes Yes Yes Yes

Grouping/Collapsing of Variables or Response Codes

(1) Average Annual Counts Rounded to the Nearest Whole Number

(2) Average Annual Rates

(3) Annual averages are based on at least 3 years of data

Yes No No

Cell Suppression Counts and rates: count of less than 16

Counts and rates: 3-5 year total count of less than 16**

No No

Complementary Cell Suppression

As needed

As needed No No

Public Release Disclosure Statement

Yes Yes Yes

Data Sharing Agreement and/or IRB Approval

User Authentication No

Logging and Monitoring Limited

Limited Yes Yes

** Pending further data quality investigation and discussion with the NPCR-CSS Scientific Working Group

NPCR-CSS Data Policy

Appendix A

NPCR-CSS Scientific Working Group Members

Faruque Ahmed, MD, PhD, MPH, Centers for Disease Control and Prevention Mark E. Allen, MS, California Cancer Registry Virginia C. Andrews, MSPH, South Carolina Central Cancer Registry Janet Bates, MD, MPH, California Cancer Registry Sally Bushhouse, DVM, PhD, Minnesota Cancer Surveillance System Cheryll Cardinez, MSPH, Centers for Disease Control and Prevention Vivien W. Chen, PhD, Louisiana Tumor Registry Susan T. Gershman, MS, MPH, PhD, CTR, Massachusetts Cancer Registry Georgette G. Haydu, MS, Ohio Cancer Incidence Surveillance System Holly L. Howe, PhD, North American Association of Central Cancer Registries Jeannette Jackson-Thompson, PhD, MSPH, Missouri Cancer Registry Missy Jamison, MPH, Centers for Disease Control and Prevention Alison T. Johnson, CTR, Vermont Cancer Registry Amy Kahn, MS, CTR, New York State Cancer Registry Karen L. Knight, MS, North Carolina Central Cancer Registry Betsy A. Kohler, MPH, CTR, New Jersey State Cancer Registry Sue Min Lai, PhD, MS, MBA, Kansas Cancer Registry Jill A. MacKinnon, CTR, Florida Cancer Data System Howard J. Martin, PhD, Virginia Cancer Registry Xiaoling Niu, MS, New Jersey State Cancer Registry Emily Reed, CTR, Kentucky Cancer Registry Patricia Reese, CTR, Wyoming Central Tumor Registry Maria J. Schymura, PhD, New York State Cancer Registry Tiefu Shen, MD, PhD, Illinois State Cancer Registry Laura Stephenson, BA, Wisconsin Cancer Reporting System Sherri Stewart, PhD, Centers for Disease Control and Prevention Thomas C. Tucker, PhD, MPH, Kentucky Cancer Registry Lydia Voti, MS, Florida Cancer Data System Hannah Weir, PhD, Centers for Disease Control and Prevention Nancy S. Weiss, PhD, Texas Cancer Registry Melanie Williams, PhD, Texas Cancer Registry Reda J. Wilson, MPH, CTR, Centers for Disease Control and Prevention Brian D. Wright, BS, Pennsylvania Cancer Registry Xiao Cheng Wu, MD, MPH, CTR, Louisiana Tumor Registry

Appendix A

NPCR-CSS Small Data Release Working Group

Mark E. Allen, MS, California Cancer Registry Pamela Agovino, MPH, New Jersey State Cancer Registry Patricia Andrews, MPH, CTR, Louisiana Tumor Registry Vivian W. Chen, PhD, Louisiana Tumor Registry Georgette G. Haydu, MS, Ohio Cancer Incidence Surveillance System Amy Kahn, MS, CTR, New York State Cancer Registry Karen L. Knight, MS, North Carolina Central Cancer Registry Sue Min Lai, PhD, MS, MBA, Kansas Cancer Registry Jill A. MacKinnon, CTR, Florida Cancer Data System Xiaoling Niu, MS, New Jersey State Cancer Registry Maria J. Schymura, PhD, New York State Cancer Registry Jeannette Jackson-Thompson, PhD, MSPH, Missouri Cancer Registry

NPCR-CSS Data Policy

Appendix B

National Program of Cancer Registries Cancer Surveillance System

308(d) Assurance of Confidentiality Statement

A surveillance system of population-based cancer incidence data received from cooperative agreement holders for the National Program of Cancer Registries is being conducted by the National Center for Chronic Disease Prevention and Health Promotion (NCCDPHP) of the Centers for Disease Prevention and Control (CDC), an agency of the United States Department of Health and Human Services, and Macro International Inc., a contractor of the CDC. The information to be received by the CDC is a subset of a standard set of data items that the State central cancer registry routinely receives from hospitals, pathology labs, clinics and private physicians on all cancer patients diagnosed in the state. This information includes patient demographics and cancer diagnosis and treatment data. Once a year in January, CDC will request cumulative data from central cancer registries. The variables reported to CDC may vary from year to year. The cancer registries maintain these data permanently in longitudinal databases that are used for public health surveillance, program planning and evaluation, and research. CDC will update its longitudinal database each year with data received from the states.

These data are used by CDC scientists for routine cancer surveillance, program planning and evaluation, and to provide data for research. NCCDPHP, recognizing the sensitivity of the data being furnished by the States, has applied for and obtained an Assurance of Confidentiality to provide a greater level of protection for the data while at CDC and at the contractor site.

Information received by CDC or its contractors as part of this surveillance system that could lead to direct or indirect identification of cancer patients is collected and maintained at CDC under Section 306 of the Public Health Service (PHS) Act (42 USC 242k) with an assurance that it will be held in strict confidence in accordance with Section 308(d) of the PHS Act (42 USC 242m).

It will be used only for purposes stated in this Assurance and will not otherwise be disclosed or released, even following the death of cancer patients in this surveillance system.

Information collected by the CDC will be used without personal identifiers for publication in statistical and analytic summaries and for release in restricted release datasets for research.

Information that could lead to direct or indirect identification of cancer patients will not be made available to any group or individual. In particular, such information will not be disclosed to:

insurance companies; any party involved in civil, criminal, or administrative litigation; agencies of Federal, State or local government; or any other member of the public.

Collected information that could lead to direct or indirect identification of cancer patients will be kept confidential and, aside from NCCDPHP employees, their contractors, and qualified researchers, no one will be allowed to see or have access to the information. CDC employees and contractors will be required to handle the information in accordance with principles outlined in the CDC Staff Manual on Confidentiality and to follow the specific procedures documented in the Confidentiality Security Statement for this project. Qualified researchers and organizations (e.g., the North American Association of Central Cancer Registries, the American Cancer Society, the National Cancer Institute) will be required to sign a detailed data release agreement to have access to restricted release data.

NPCR-CSS Data Policy

Appendix C

308(d) Assurance of Confidentiality Frequently Asked Questions

Background The Centers for Disease Control and Prevention (CDC) is responsible for public health surveillance in the United States. CDC collects, compiles, and publishes a large volume of personal, medical, epidemiologic, and statistical data. The success of CDC’s operations depends, in part, on the agency’s ability to protect the confidentiality of these data. While it is a matter of principle for CDC to guard sensitive information, and federal statutes such as the Privacy Act provide a degree of protection for personally identifiable data, the Public Health Service Act, Section 308(d) enables CDC to provide the highest level of confidentiality protection for sensitive and mission-significant research and surveillance data.

CDC received a formal delegation of authority from the National Center for Health Statistics (then a separate agency) to grant 308(d) confidentiality protection in 1983. Section 308(d) of the Public Health Service Act (42 U.S.C. 242 m(d)) ensures the confidentiality of data collected under Sections 304 and 306 of the Public Health Service Act. These special legislative authorities were the provisions under which NCHS collects and safeguards most of its survey data, along with the mortality data within the National Death Index. CDC was required to establish a stringent application process, and continues to use the authority sparingly. The agency has granted confidentiality assurances to projects deemed significant to CDC’s mission, such as surveillance of hospital infections, AIDS and HIV infections, pregnancy related mortality, and congenital defects. Fewer than fifty projects have received 308(d) protection since CDC received this authority, and currently there are only approximately 25 active projects with 308(d) confidentiality assurances. As a testament of the importance of this project to the mission of CDC, the National Program of Cancer Registries (NPCR) has been afforded this special data protection.

1. What is stated in Public Health Service Act, Section 308(d)?

The first clause of Section 308(d) states that CDC must explain the purpose for collecting data to persons or agencies supplying information, and it guarantees that CDC will be limited to those specified uses unless an additional consent is obtained. Moreover, the information obtained may be used only by staff of CDC, or its contractors, in the pursuit of such stated purposes. The second clause states that CDC may never release identifiable information without the advance, explicit approval of the person or establishment supplying the information or by the person or establishment described in the information.

2. What process did NPCR undertake to obtain 308(d) confidentiality protection?

NPCR staff worked with the CDC Office of General Counsel and the CDC Confidentiality and Privacy Officer to prepare the application for the NPCR Cancer Surveillance System (CSS)

December 2006 project. The application contained the following four components:

a. A Justification Statement summarizing the NPCR-CSS project’s programmatic purpose, the type of data to be collected, and the uses to be made of the information.

This statement also included an assurance that a) the requested data would not be furnished without the guarantee of a confidentiality assurance, b) confidentiality assurance is important to protect the individuals described in the data and to reassure the institutions submitting data, c) the information cannot reliably be obtained from other sources, d) the information is essential to the project’s success, e) granting the confidentiality assurance would not prohibit CDC from fulfilling its responsibilities, and f) the advantages of assuring confidentiality outweigh the disadvantages.

b. An Assurance of Confidentiality Statement delineating anticipated data uses and those with whom identifiable data would be shared, along with general advisements regarding the confidentiality protection.

c. A Confidentiality Security Statement detailing the stringent safeguarding measures in place to ensure that the promise of confidentiality would not be jeopardized by practices of staff handling the data.

d. An IRB Review Status Statement verifying NPCR-CSS’s exemption from CDC Institutional Review Board (IRB) approval. (The Human Subjects Administrator at the National Center for Chronic Disease Prevention and Health Promotion determined that NPCR-CSS activities are routine surveillance and not research on human subjects. Therefore, protocol review by CDC IRB was deemed not necessary.)

The application was submitted to the CDC Confidentiality Officer for review and modification, prepared for presentation to the CDC Confidentiality Review Group (CRG), and in May 2000 NPCR received 308(d) confidentiality protection approval for NPCR-CSS data, including authorization for retroactive confidentiality protection beginning with diagnosis year 1995.

NPCR must file for continuation every five years to maintain the assurance. In 2006, NPCR filed and received approval for continuation.

3. What makes 308(d) confidentiality assurance the best protection for NPCR-CSS data?

The 308(d) confidentiality assurance is the only confidentiality protection that covers routine surveillance activities, such as those conducted by NPCR-CSS. The assurance specifies that data protected by 308(d) may be used only for statistical or epidemiological purposes and not released further in identifiable form without consent. Another exclusive advantage of 308(d) is that it also protects indirectly identifiable data.