NAACCR Virtual Pooled Registry Cancer Linkage System

The VPR-CLS is designed to facilitate minimal risk linkage studies with multiple cancer registries. In order to utilize the VPR-CLS, the following requirements must be met:

• Study has (or will have) an existing cohort
• Study has a current IRB-approval, IRB exempt determination, or documentation of Not Human Subjects Research
• The study protocol includes linkage with cancer registries
• Study consent form includes linkage with cancer registries OR study has a specific waiver of informed consent to link with registries

The VPR-CLS consists of, and has been developed in, two distinct phases:

• Phase I I supports a standardized linkage with participating registries and release of provision of aggregate match counts (by registry and diagnosis year), allowing the researcher to prioritize registries from which to request additional data. Phase I includes a web-based application, secure data transfer protocols between researchers, IMS and registries, and use of a single record linkage software (Match*Pro) optimized for linkages between cancer registries and research cohorts. Phase I functionality has been completed.
• Phase II supports the researcher's process of applying to registries and their IRBs for release of individual-level cancer data on matched cases identified in Phase I. It includes use of a Templated IRB/Registry Application, compilation of any required registry applications or agreements, a Templated Data Use Agreement, and will soon include a dedicated Central IRB for the reviewing VPR linkage requests. Phase II development and initial testing has been completed. Additional enhancements to Phase II are currently underway with planned completion in 2022.

Phase II of the VPR-CLS will streamline the process of applying for release of individual-level data on matched cases by offering registries and state/local IRBs optional use of the following resources:

1. Templated Forms: NAACCR has led efforts to create the following templated forms that can be used in lieu of the state-specific forms:
   1. Templated IRB/Registry Application (TIRA): The TIRA is a single application that includes common questions from across 50+ individual registry and IRB applications. The majority of VPR-participating registries have adopted use of the TIRA in lieu of their state-specific applications.
   2. VPR Templated Data Use Agreement (VPR DUA): The VPR DUA is a single DUA, with standard terms and conditions, developed by the NAACCR Templated DUA Task Force and vetted with registries, researchers, and legal teams. Over half of the VPR-participating registries have adopted the VPR DUA has been adopted in lieu of their state-specific DUAs.
2. Central IRB: The Central IRB will serve as a single reviewing body, in lieu state/local IRBs, for multi-site minimal risk linkage studies, such as those coming through the VPR-CLS. Funded by NCI, the Biomedical Research Alliance of New York (BRANY) is currently talking with registries and their IRBs about entering into a reliance agreement that would allow BRANY to perform IRB review for all VPR linkage studies.

All VPR-CLS linkages are performed using the record linkage software, Match*Pro, developed by IMS. Match*Pro conducts probabilistic linkage based on the Fellegi and Sunter model. The following variables are used, as available, to link the study file with the registry file: First name, middle name, last name, maiden name, date of birth, social security number, telephone number, gender, and street address. After probabilistically identifying potential matches, deterministic filters classify each linked pair as a match, non-match, or uncertain.

Phase I: Application to use VPR-CLS to link with registries behind their firewall and receive aggregate match counts only (no IRB review needed).

Anticipated Timeline: 2-4 weeks for application review; 2-4 weeks after the researcher provides validated, edited cohort file, the VPR-CLS will return match counts by registry and diagnosis year.

1. Researcher submits online VPR-CLS application and supporting documents. Supporting documents include the current institutional IRB approval, approved study protocol, consent form or waiver of consent, investigator's curriculum vitae, and signed data use agreement with IMS.
2. NAACCR reviews application for completeness and resolves any issues with researcher.
3. Research Review Committee (RRC) reviews application and researcher is notified of the decision.
4. Researcher creates, edits, and uploads an encrypted study linkage file to the VPR-CLS in accordance with established file specifications and editing software.
5. Encrypted study file is validated and posted for registries to securely download.
6. Registries perform linkage behind their firewalls using Match*Pro and a pre-defined, standard linkage configuration file.
7. Registries generate aggregate match counts (no patient records) of the number of high quality and potential matches by diagnosis year and upload to the VPR-CLS.
8. Match counts are provided to the research through the VPR-CLS.

Phase II: Application for release of individual-level data on matched cases identified during Phase I.

Anticipated Timeline: IRB/Registry approval and release of data will vary considerably based on the review process of the Central IRB and processes in each state (if necessary).

9. Researcher reviews match counts and selects which registries to approach for release of individual-level data on matched cases.
10. The VPR-CLS assists the researcher in completing a streamlined application process, which includes filling out a Templated IRB/Registry Application, uploading supporting documents, and providing links to state-specific applications, if needed.
11. Central IRB, registries, and local/state IRB, as appropriate, review application and enters the review determination into the VPR-CLS tracking system. Registries may require a data use agreement, confidentiality form, and payment to cover costs.
12. Upon approval and fully executed DUA (if applicable), registry creates a file of individual-level data on the matched cases.
13. Registry provides data directly to the researcher through a secure site, independent of the VPR-CLS, as specified by either the researcher or the registry.

The VPR-CLS provides a web-based portal through which researchers submit an application to use the system to link with registries. A Data Use Agreement is signed between the researcher and IMS prior to uploading any study data files containing patient identifiers to the secure data transmission service. The website uses Transport Layer Security, ensuring that communication and files transferred between a client and the IMS server are securely encrypted. All study data files are encrypted by the researcher (using Match*Pro) prior to uploading them to the VPR-CLS. All files uploaded to the VPR-CLS are first scanned for viruses and then stored on a secure server behind the IMS firewall. Only authorized full-time IMS staff can access the study files provided by researchers and all IMS staff have been trained in the handling of files that contain personal identifying information.

Once an uploaded study file has been reviewed and approved by IMS, it will be accessible for secure download by an authorized liaison from each of the participating registries. All registry liaisons are authenticated and verified by IMS prior to receiving access to the VPR-CLS. In addition, each registry has confirmed compliance with a list of common security protections described in the VPR-CLS Cancer Registry Security Protections document. Individually signed checklists are located on the About/Participating Registries tab and a summary of these protections across registries is available here. Study files are used solely for data linkage by the participating registries. Once all registries have downloaded the study file, access to the file on the VPR-CLS will be removed. Each registry will perform the linkage behind their firewall. Both the registry and IMS will destroy the study file at the conclusion of the match.

Study files are used solely for data linkage by the participating registries. Each registry will perform the linkage behind their firewall. After the match count report is uploaded, the study file will no longer be accessible for download from the VPR and the registry will be prompted to delete the study file and provide confirmation of destruction. Later in the request workflow, after all the participating registries have sent their case data matches to the study requestor, the requestor is provided an option to request file deletion after the data have been reviewed. If the requestor selects to have data deleted, all registries are required to delete the linkage results file produced by Match*Pro and provide confirmation of destruction. In addition, the study file is deleted from the IMS server. All backups at IMS are removed after 3 months.

Registry participation it the VPR-CLS is voluntary. As of 2022, 46 registries are participating representing 90% of the U.S. population (see map below). Details about the participating registries, including years of available data, population size, annual cancer counts, contact information, and security protections can be found here.

U.S. cancer registries have varying years of complete data available for linkage as noted on the NAACCR website. The majority of registries will include the most recent data in their VPR-CLS linkage file (e.g. In 2019, cancer cases diagnosed through the end of 2017 would be included). Cancer registries provide standardized, curated, population-based data on patient demographics, incidence, stage, treatment, and follow up.

The VPR-CLS development team would like to acknowledge the role of the National Heart, Lung, and Blood Institute and its BioLINCC Program in the approach used in the development of this website. The VPR-CLS code was based on the success of the BioLINCC framework, which shares many of the same advanced features and a well-defined yet flexible workflow for request submission and fulfillment.

BioShare is developed and maintained by Information Management Services, Inc.