PLCO Data Access Policies and Procedures by hijuney4

VIEWS: 0 PAGES: 7

									PLCO Cancer Screening Trial

PLCO Data Access Policies and Procedures
Version 8/27/08 INTRODUCTION The Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) has established policies to promote access to PLCO data by qualified investigators and to make transparent the review process and ongoing research in all areas. Detailed policies are provided below in Section 1.0 PLCO Data Access Policies. Investigators may request access to PLCO data to support proposal development or to conduct a proposed study. Data may be provided by the NCI or NCI contractors, Information Management Systems (IMS), and Westat as simple frequencies and cross tabs in response to a request for exploration (i.e. proposal development), or as more complicated analyses to complete a study. In addition, data sets may be provided for use by an investigator to conduct his or her own analyses. All requests for exploration of PLCO data must be submitted on a PLCO Data Exploration Form (DEF) and approved according to procedures given below in Section 2.0. All requests for data sets or analysis of PLCO data by IMS, Westat, or NCI for the purposes of preparing a manuscript, abstract, poster, or other public access document, must be submitted on a PLCO Proposal Application Form (PAF) and approved according to the procedures described below in Section 3.0. Any requests to access PLCO data that also include a request for use of PLCO biospecimens must be submitted to the PLCO Etiologic and Early Markers Studies (EEMS) Review Panel. There are two EEMS application forms: 1) EEMS Concept Form – complete to obtain feed back on an idea or concept prior to developing a full proposal, and 2) EEMS Proposal Form – complete to obtain approval for release of PLCO biospecimens and data. (Note: EEMS applications are only accepted two times per year, June and December.) These forms are available on the PLCO STARS website (www.plcostars.com) or the investigator may contact the EEMS Coordinator for more information (e-mail: plco-eems@westat.com).

PLCO_Data_AccessPolicy_Proc

1

11/14/2009

PLCO Cancer Screening Trial 1.0 PLCO DATA ACCESS POLICIES The following are policies governing access to PLCO data: 1. 2. 3. The PLCO Chief Statistician or designee may have access to any and all PLCO data for the purpose of monitoring the Trial and conducting analyses. All proposals that request the use of PLCO data, without requesting biospecimens; will be assigned to a PLCO subcommittee. Proposed work shall be collaborative and may involve one or more investigators from NCI (DCP, DCEG, or other divisions) or it may involve a Principal Investigator from a PLCO Screening Center. If the NCI PLCO collaborator is from a division other then DCP the PLCO Project Officer and Chief Statistician reserves the right to assign a DCP collaborator to any paper or study that is either central to the PLCO Trial or controversial. All studies must comply with the PLCO Analysis Plan and policies regarding the use of PLCO data and/or biospecimens. Investigators must sign a Confidentiality Disclosure Agreement prohibiting the use of PLCO data for other than the pre-specified uses. Release of PLCO data is time-limited and for approved purpose(s) only. Data cannot be used for additional analyses and/or publications without going through the study approval process for the additional work. Studies that are not completed within reasonable timeframes, as agreed during the approval process, may be superseded by other investigators. During development of a proposal, data access will be controlled and performed through NCI until final study approval. Only the data needed for the final approved proposal will be released outside of NCI control if such a release was negotiated during the review process. Data will be released only as "restricted-use" datasets. No identifying information and no PLCO Participant Identifier numbers (PID) are to be provided in the released data sets. A PLCO PID is a unique number assigned to a PLCO participant at enrollment. Only the SC has access to the identifying information linked to the PID. Only Westat and IMS have access to the study data linked to the PID. This document only applies to restricted use datasets that are not subject to anonymization or statistical disclosure controls. Other datasets must go through a different process and are subjected to an expanded impact analysis and varying levels of disclosure control depending on the specific data and usage. Discrete data elements that uniquely identify an individual will not appear in any datasets, including restricted use datasets, (e.g. PID, name, address, social security number, medical record number, 9-digit zip code, etc.). A final impact analysis will be performed prior to release of data to ensure that inference channels are addressed. Data are non-transferable unless prior authorization by NCI has been granted. Data cannot be used for any purpose other than originally approved and cannot be released to anyone but the agreed collaborators. Information about datasets released for approved studies will be available to PLCO Trial collaborators. This includes data dictionaries corresponding to the datasets (including calculated fields), as well as release dates and other tracking information.

4. 5.

6.

7.

8.

9.

PLCO_Data_AccessPolicy_Proc

2

11/14/2009

PLCO Cancer Screening Trial 10. Analyses may be repeated by NCI prior to publication, at NCI’s discretion, ensuring that investigators have correctly interpreted PLCO data. This decision will be made towards the end of the study and cannot be included in the impact analysis. Laboratory data and/or study results from non-anonymized studies that complete the review process can be released to the PLCO Trial by other investigators for use with other Trial data, unless otherwise negotiated during the usual review process. Procedures for the release will include a check for methodological studies to be reviewed on a case-by-case basis regarding the appropriateness of the methodology chosen by the investigator. PLCO Trial outcomes will not be compromised. To ensure outcomes are not compromised, a “living list” of variables that will require special approval prior to release of data will be maintained and reviewed at least semi-annually. The PLCO Project Officer and Chief Statistician will establish this evolving policy in conjunction with the DSMB. See Guidelines for PLCO Release of Data in Section 2.1 below. Datasets will be released to investigators for use locally (by investigator programmers) or on the NIH computer mainframe Center for Information Technology (CIT) as determined during the study approval process. Generally, it is preferred that datasets be released as SAS or ASCII files. Additional costs may be incurred for other formats. NCI prefers that all data analysis be kept central, unless otherwise negotiated during the review process. PLCO policy accounts for release of data for use outside of direct NCI control. This will be necessary for anonymized data that would be analyzed by the principal investigator or under grants, as well as analyses supported and funded by NCI.

11.

12.

13.

1.1

Guidelines for PLCO Release of Data The following guidelines governing release of data to investigators were developed by the NCI PLCO Chief Statistician and the NCI PLCO Project Officer in conjunction with the PLCO Data and Safety Monitoring Board. The content of this document will be reviewed regularly by the Chief Statistician and will be revised as appropriate. The guiding principles for release of data are: 1) 2) Analyses of released data should not compromise trial endpoints. All proposals for obtaining access to PLCO data should contain a discussion of why the trial will not be harmed. All analyses are subject to approval by the PLCO Publications Subcommittee, the PLCO Project Officer, and the PLCO Data and Safety Monitoring Board.

3)

PLCO_Data_AccessPolicy_Proc

3

11/14/2009

PLCO Cancer Screening Trial The guidelines for release of data fall into three categories: 1. Category I – Acceptable for Release – all data items in this category will be eligible for expedited approval by the PLCO Chief Statistician or his designee: 1. 2. 3. 4. 5. 6. 7. 2. All baseline questionnaire data All dietary data – DQX and DHQ T0 screening test results T0 secondary screening results, for example, cyst size (TVU), granuloma (XRY) – see PLCO screening examination forms for details T0 cancer characteristics already in T0 papers T0 compliance Non-PLCO cancer data

Category II – May be Acceptable for Release: All requests for analysis that fall into this category require review by and approval of the Chief Statistician (or designee). 1. 2. 3. 4. 5. 6. 7. 8. 9. T1+ screening results, such as compliance, findings, PPV and yield, and secondary results T1+ cancer characteristics such as stage and grade (no comparisons between trial arms) T0+ diagnostic evaluation procedures T0+ treatment information Comparison of screen vs. interval diagnosed cancers and characteristics Comparisons among Screening Centers Cancer survival data (excluding Prostate, Lung, Colon, Ovary cancer specific data) Mortality data (excluding Prostate, Lung, Colon, Ovary cancer specific data) PLCO cancer incidence rates (no comparisons between trial arms)

3.

Category III – Not Acceptable for Release 1. 2. 3. 4. Prostate, Lung, Colon, Ovary cancer specific mortality data Prostate, Lung, Colon, Ovary cancer specific survival data Prostate, Lung, Colon, Ovary cancer stage distributions by arm and stage specific incidence rates Contamination (HSQ) data

All data questions not appearing on this list will fall into Category II (May be Acceptable for Release).

PLCO_Data_AccessPolicy_Proc

4

11/14/2009

PLCO Cancer Screening Trial 2.0 EXPLORING PLCO DATA (PLCO DATA EXPLORATION FORM – DEF) NCI, Westat, and IMS have worked together to develop procedures for exploring PLCO data. Usually most data exploration will be done to develop a proposal; it is expected only basic frequencies and cross tabulations will be required. The procedures for requesting approval to explore data are provided below:

1.

All investigators must complete a PLCO Data Exploration Form (DEF) and a PLCO Confidentiality Disclosure Agreement (CDA) to obtain approval for exploration of PLCO data.* These forms are available on the PLCO Study Tracking and Review System (STARS) Web site (www.plcostars.com) or the investigator may contact the PLCO Data Access Coordinator (DAC) (e-mail:PLCO-Requestdata@westat.com) for more information. * A PLCO investigator does not have to submit a PLCO Data Exploration form to explore data limited to his/her own PLCO Screening Center. However, if the PLCO investigator plans to conduct a study with data from his/her own PLCO Screening Center, then the PLCO investigator must submit a PLCO Proposal Application Form (PAF) and comply with the proposal application approval process (see Section 3.0 below).

2.

Investigators will submit, via PLCO STARS, a completed PLCO Data Exploration Form (DEF) to the Data Access Coordinator (DAC), who will log the form, assign a PLCO Study ID number, review it for completeness and accuracy, and forward it to the PLCO Chief Statistician. The PLCO Chief Statistician will review the PLCO DEF to approve or reject requests for data exploration. The DAC will document the results of the review, including as appropriate, reasons for rejection or conditions for approval. If the request is approved, the DAC will notify the investigator to contact IMS, Westat, or NCI and begin work. If the request is denied, the form will be returned to the investigator. The DAC will track all submitted and reviewed DEFs and report approved requests to the appropriate PLCO Subcommittee(s) and NCI. Data Exploration Forms may be submitted at any time. The PI should allow two weeks for review and approval of a DEF. All investigators will have six months to obtain data for an approved DEF request. It is anticipated that an investigator should have thoroughly explored an idea in this time and either will have completed a Proposal Application Form (PAF) to conduct a study or determined the idea to be infeasible. An investigator will receive notice at four and five months post approval of impending DEF expiration. This will allow time for the investigator to confer with a PLCO subcommittee as needed, prior to the expiration of the DEF. At six months post-approval the investigator’s DEF will be considered “Expired” and no further data may be released. Until a PLCO Proposal Form (PAF) or EEMS form is submitted and approved, an idea for a study/manuscript may be superseded by another investigator.

3.

4. 5. 6.

7.

PLCO_Data_AccessPolicy_Proc

5

11/14/2009

PLCO Cancer Screening Trial

2.0

REQUESTING USE OF PLCO DATA (PLCO PROPOSAL APPLICATION FORM – PAF) After completing an approved exploration of PLCO data, an investigator may be ready to conduct a detailed analysis of selected PLCO data to prepare a manuscript, abstract, or poster. Investigators must submit a PLCO Proposal Application Form (PAF) to obtain support for detailed analysis or to obtain data files. The PLCO PAF will be reviewed by a PLCO Subcommittee and by the NCI Project Officer and Chief Statistician. The procedures for submission and review of a proposal to obtain support for analysis or to obtain data files are given below. 1. All investigators must complete a PLCO Proposal Application Form (PAF). The form and specifications for completion are available on the PLCO STARS Web site (www.plcostars.com) or the investigator may contact the Westat PLCO DAC (e-mail: PLCO-Requestdata@westat.com) for more information. The investigator should complete and submit a PAF on the PLCO STARS Web site. The DAC will conduct a review of the PAF for accuracy, completeness, and an impact analysis (an evaluation of compliance to the Guidelines for PLCO Release of Data). The DAC will forward the completed application to the PLCO Project Officer and Chief Statistician who will perform a suitability review to determine that the proposal does not conflict with the PLCO Data Analysis Plan and that it does not compromise the study endpoints. They will document the outcome of the suitability review in PLCO STARS and assign the proposal to one or more suitable subcommittees for review. The DAC will notify the appropriate Subcommittee Chair(s) that a new proposal is ready for review and schedule a discussion of the proposal with the subcommittee(s) and the PI for the proposal. Following the discussion of the proposal, the Subcommittee Chair will request that voting members make recommendations for approval or rejection of the proposal. The Subcommittee Chair will decide the method of voting such as e-mail after a call or hand vote during a meeting (where the PI steps out of the room), and the number of voters required for a quorum. Voting members will provide reasons for approval and/or rejection. (See companion document “PLCO Subcommittees: Responsibilities, Operations and Administration” for an explanation of voting membership.) The PLCO Subcommittee will document and submit their recommendation using the PLCO STARS Web site. a. b. The Subcommittee Chair may bring in outside experts if he/she believes they are needed to contribute to the discussion and evaluation of a particular proposal. Subcommittees may recommend approval of proposals that fall under their subject areas to the PLCO Project Officer and the PLCO Chief Statistician and a DCEG PLCO Senior Scientist, as appropriate. Only the NCI Project Officer and Chief Statistician have the final authority to approve or reject a proposal.

2.

3.

4.

5.

6.

The PLCO Project Officer and Chief Statistician will review Subcommittee recommendations and comments to determine if the proposals involve epidemiologic issues for which consultation and approval by a DCEG PLCO Senior Scientist should be sought. After consulting with a DCEG Senior Scientist as appropriate, the PLCO Project Officer and the PLCO Chief Statistician will document their decisions using the PLCO STARS Web site to indicate approval or rejection (with reasons) of a proposal.

PLCO_Data_AccessPolicy_Proc

6

11/14/2009

PLCO Cancer Screening Trial 7. The DAC will notify an investigator when a proposal is approved and the investigator may contact IMS or Westat to begin work. Please note that data will not be released to an investigator until he or she has submitted a PLCO Confidentiality Disclosure Agreement (CDA). NCI prefers that all data analysis be kept central, meaning that data will be released to the investigator through an analytic group (NCI, IMS, or Westat), and the investigator may request analyses and review outcomes. However, an investigator may apply for special permission from the PLCO Chief Statistician to work independently on his or her own computer: a. b. Prior to the release of the data set, the PLCO Chief Statistician will review the request, the final data set, and the data dictionary; A process must be developed to ensure that the investigator understands the nuances of the data and data collection issues (this will require input from Westat and the analytic group and can be coordinated through Subcommittee activities), and The investigator must be prepared to send a copy of the analytic file which corresponds to the accepted manuscript to NCI. NCI’s analytic group may rerun the investigator’s analysis to verify results prior to publication, at NCI discretion.

c.

8. 9. 10.

If a PAF is not approved, the investigator may appeal the decision with the PLCO Project Officer and Chief Statistician and/or revise the proposal and resubmit. The DAC will post an abstract for all approved proposals on the STARS Web site. A PLCO PAF may be submitted at any time. The PLCO Subcommittees meet six times a year in January, March, May, July, September, and November. Investigators should submit applications to the DAC two months prior to the next subcommittee call or meeting to facilitate a timely review. Investigators should expect a six to eight week turn around for proposals submitted according to the recommended schedule. The DAC will request that the investigator complete a PLCO Semi-Annual Progress Report for an approved proposal. If after twelve months the investigator reports no progress, the monitoring PLCO Subcommittee will determine if the proposal should be withdrawn. The Analysis Cost Effectiveness and Modeling (ACEM) Subcommittee Chair will monitor all new research in each Subcommittee to determine if any proposals should be submitted to the ACEM Subcommittee. If an investigator wishes to obtain additional data or a new analysis for an approved PAF, he or she must submit a PLCO Addendum Data Request Form, and complete the PLCO Addendum Review Process. Information on this process can be found on the PLCO STARS Web site. All manuscripts, abstracts, posters, or other proposed publications using PLCO data must have an approved PLCO PAF on file before the PLCO Publication Subcommittee will accept them for review.

11.

12.

13.

14.

PLCO_Data_AccessPolicy_Proc

7

11/14/2009


								
To top