The CIBMTR manages one of the world’s largest blood and marrow transplant research databases and is a unique information resource for clinicians, researchers and the general public interested in HCT. It is the CIBMTR’s policy to provide maximum access to and use of its data.
13.1 ACCESS TO CIBMTR DATA
Access to CIBMTR data falls into five major categories: Publicly available data;
Customized requests for data and analyses;
Requests to conduct a CIBMTR clinical outcomes study; Requests for datasets from previous CIBMTR research; Transplant center requests for data.
The processes the CIBMTR follows for handling these requests, and timelines for completion, vary depending on the type of request, complexity, need for statistical resources, and available external funding.
The CIBMTR provides summarized, general information through its websites, described below and in Chapter 15. In order to efficiently handle all requests, the CIBMTR has a few common pathways to receive and triage the requests.
An MS-level Statistician is the first to receive, triage, and respond to information requests that are received via the web-based Data Request System at http://www.cibmtr.org/Data/Request/ Pages/index.aspx. CIBMTR Scientific Directors work with the MS-level Statistician to provide scientific input, address questions and when appropriate review data prior to release.
13.1.1 Publicly Available Data
The CIBMTR regularly publishes data in a variety of report formats and makes these data available through the CIBMTR (http://www.cibmtr.org), NMDP/Be The Match
http://bethematch.org/ and HRSA (http://bloodcell.transplant.hrsa.gov/) websites or upon request. Examples of information readily available include, the annual CIBMTR Summary Slides and the SCTOD Reports (US Patient Survival Report, Transplant Data by Center Report, US Transplant Data by Disease Report) as well as synopses of studies published by the CIBMTR. These data are provided as reports and are created to support both research as well as general informational uses. More information about CIBMTR websites can be found in Chapter 15.
13.1.2 Customized requests for Data and Analyses
Requests that cannot be addressed using existing reports require a customized response. Reasons for these requests include self-education and decision making, patient counseling or clinical decision making, presentation support, and transplant center assessments. They range from simple queries of patient, disease, and transplant frequencies to those with greater complexity involving specific data combinations and / or statistical analysis of outcomes. The CIBMTR receives these requests by email or via the Custom Information Request Form on its website (http://www.cibmtr.org/Data/Request/Documents/Custom_Information_Request_ Form.pdf). Requests related to clinical decision making are responded to within three days and
results from other simple requests are generally returned within one week. Requests requiring a more complex custom analysis or dataset may take additional time, up to four weeks. If a request will take more than the estimated one to four weeks to fulfill, the requestor will be notified of the new time estimate. Most requests will require Scientific Director review prior to sending the response. Corporate requests follow the guidelines of the Corporate Program and may require a Statement of Work and budget (Chapter 17). In other instances, the requestor may be advised to submit a formal study proposal to a CIBMTR Working Committee for approval and prioritization. Transplant center requests may be referred to the Data Back to Centers (DBtC) web tool to obtain TED-level data.
13.1.3 Requests to Conduct a CIBMTR Clinical Outcomes Study
The most common request to access CIBMTR data and statistical resources is through
submission of a study proposal to a CIBMTR Working Committee (Chapter 6). Information on how to propose a study is on the CIBMTR website (http://www.cibmtr.org/Studies/
Observational/ProposeStudy). A list of previously published CIBMTR research, studies that are in progress, and copies of data collection forms are available at http://www.cibmtr.org/Data Management/DataCollectionForms/pages/index.aspx, and investigators are encouraged to review these before submitting a proposal.
Working Committee members evaluate new proposals based on their scientific merit and feasibility as well as the CIBMTR’s ability to complete the study in a timely fashion. Feedback from Working Committee members at the annual meeting is considered in deciding the relative merits of proposals. Final decisions regarding which studies to pursue are made by the Working Committee leadership and the CIBMTR Advisory Committee. Proposals may also be submitted for review by Committee leadership between meetings and may be assigned resources if deemed to be of exceptionally high impact and timeliness.
13.1.4 Requests for CIBMTR Datasets from Previous Research
The CIBMTR supports requests for datasets assembled for previously published studies and the annual Center Specific Outcome Analysis to conduct additional, independent analyses.
Requestors are required to submit a study proposal using the format noted in 13.1.3 to define their planned analysis and study design. These requests are for external use of the data and do not require CIBMTR statistical resources to conduct the proposed analyses. Requests for
existing datasets are reviewed and approved by Coordinating Center staff. These datasets often require some review and update before they can be provided to the requestor and are
accompanied by a data dictionary when distributed.
Responses to these requests are typically handled in three to four weeks but can take longer depending on complexity, verification, and approval.
13.1.5 Transplant Center Requests for Data
Many dataset requests received by the CIBMTR are from transplant centers that submit data to the CIBMTR and are for data specific to their own center. Requestors are instructed to use the DBtC application to retrieve TED level data submitted to the CIBMTR and stored in the Research Data Base (Chapter 14). More complex requests for existing data) are individually addressed by CIBMTR IT staff.
CIBMTR Manual of Operations Chapter 13: Data – Access and Release
Responses to requests for existing data for transplant centers are typically handled within one to two weeks.
13.2 RELEASE OF CIBMTR DATA
The CIBMTR releases only de-identified datasets that comply with all relevant federal regulations regarding privacy and confidentiality. It has standard policies and procedures in place for protecting CIBMTR data; these are addressed in further detail in Chapter 11 and
Chapter 12.
When releasing data, the CIBMTR is obligated to ensure that datasets do not contain Protected Health Information. CIBMTR staff follow a standard procedure for creation of de-identified datasets that specifies removal of all patient, donor, and center identifiers, which could lead to the identification of a patient or transplant center from data files. The CIBMTR does not release to the PI or any other member of his / her research team identifiable patient or center variables unless these data are critical to the approved study / project or will be used for linking to
another data file via an established honest broker relationship. In these cases, special procedures are outlined in the Linking CIBMTR Data to External Databases or Data Sources Standard Operating Procedure and documented with a Data Use Agreement (Appendices K and
D1) and established prior to final approval of the request.
Datasets are prepared with SAS statistical software and generally contain a standard set of essential data with pre-defined categories. A data dictionary defining each variable, valid values and labels accompanies the dataset. Other file formats are provided upon request.
In cases of an approved traditional clinical outcomes study (Section 13.1.3) or when datasets are requested from previous research that will not utilize CIBMTR statistical resources for analysis (Section 13.1.4), the PI must submit a Data Use Agreement (DUA) that specifies the requirements for using CIBMTR data (Appendix D1 or D2) before final approval of the project is offered. The DUA is provided to the PI, or requestor of data, when a study protocol or
statement of work has been submitted.
13.3 PUBLISHING CIBMTR DATA
Authors requesting to reproduce CIBMTR figures or unpublished data, data in manuscripts, or other printed or online media must first receive permission from the CIBMTR and must acknowledge use of the CIBMTR’s data. These requests require the use of disclaimer text and completion of the Publication Permission Request Form (located at
http://www.cibmtr.org/Copyright/PubPermission/Pages/index.aspx) prior to publication or presentation. A letter granting the requestor permission to publish CIBMTR data includes the disclaimer text.