AoU_CLAD_ROA-OTA-23-003 ROA.docx.pdf

National Institutes of Health (NIH) Office of the Director (OD) All of Us Research Program

Research Opportunity Announcement (ROA):

All of Us Center for Linkage and Acquisition of Data No. OTA-23-003

Issue Date: February 15, 2023

TABLE OF CONTENTS

SECTION 1: RESEARCH OPPORTUNITY ANNOUNCEMENT INFORMATION

1.1 Purpose

1.2 Overview

SECTION 2: ISSUING OFFICE AND POINTS OF CONTACT

2.1 Contact Information Agency

2.2 Questions

2.3 Amendments

SECTION 3: AWARD INFORMATION

3.1 Award Type and Authority

3.2 Funding and Period of Performance

3.2.1 Anticipated Funding

3.2.2 Period of Performance

SECTION 4: PROGRAMMATIC BACKGROUND

4.1 All of Us Research Program

4.2 Purpose of CLAD

4.3 Relationship Between CLAD and the Data and Research Center (DRC)

4.4 Awardee High-Level Tasks

4.5 Resources and Previous Announcements

SECTION 5: CLAD AWARD STRUCTURE, CORES AND FUNCTIONS, BASE PERIOD

DATA LINKAGES, BASE PERIOD MILESTONES, AND OPTIONAL PERIODS

5.1 Award Structure

5.2 Cores and Functions

5.2.1 Research Management Core (RMC)

5.2.2 Data Platform Core (DPC)

5.2.3 Data Linkage Core (DLC)

5.2.4 Program Compliance Core (PCC)

5.3 Base Period Data Linkages

5.3.1 Mortality Data

5.3.2 Residential History

5.3.3 Geocoding

5.3.4 Environmental Data

5.3.5 Health Care Claims Data

5.3.6 Pilot for Health Information Network (HIN) / Health Information Exchange (HIE)

5.4 Base Period Milestones

Milestone 1 - Award Onboarding and Access Milestone 2 - Regulatory and Policy Milestone 3 - DLP Security, Privacy, and Authorization To Operate (ATO) Milestone 4 - DLP Development and Maintenance Milestone 5 - Data Quality, Governance, and Sustainability Milestone 6 - Linkage for Mortality Data Milestone 7 - Linkage for Residential History Milestone 8 - Geocoding Milestone 9 - Linkage for Environmental Data Milestone 10 - Linkage for Health Care Claims Data Milestone 11 - Pilot for Health Information Network / Health Information Exchange Milestone 12 - Strategic and Innovation Planning

5.5 Optional Periods

SECTION 6: ELIGIBILITY INFORMATION

6.1 Higher Education

6.2 Nonprofits

6.3 For-Profits

6.4 Non-Eligible Entities

SECTION 7: PRE-SUBMISSION WEBINAR AND LETTER OF INTENT

7.1 Pre-Submission Webinar

7.2 Letter of Intent

SECTION 8: COMPETITION PROCESS, CONTENT SUBMISSION REQUIREMENTS, AND

EVALUATION CRITERIA

8.1 PHASE I -FULL TECHNICAL AND COST PROPOSAL

8.1.1 Proposal Submission Instructions

8.1.2 Proposal Format

8.1.3 Proposal Content

8.1.4 Proposal Evaluation Criteria

8.1.5 Cost Evaluation

8.1.6 Proposal Selection

8.2 Phase II - Data Linkage Platform (DLP) Capabilities Demonstration and Evaluation (Invitation Only)

8.2.1 Meeting Structure

8.2.2 Data Linkage Approach

8.2.3 Evaluation

8.3 Phase III - OT Award (Invitation Only)

SECTION 9: ADDITIONAL INFORMATION

9.1 Non-Government Advisors

9.2 Proposal Costs

9.3 Protests or Appeals

9.4 Additional Requirements

9.4.1 Organization and Governance Structure

9.4.2 Data Security and Management

9.4.3 Human Subject Requirements

9.5 Organizational and Personal Conflict of Interests

9.5.1 Responsibilities of Recipients regarding Organizational Conflicts of Interest

9.5.2 Responsibilities of Recipients regarding Financial Conflicts of Interest (FCOI)

9.6 Lower Tier Agreements

9.7 Terms and Conditions

SECTION 1: RESEARCH OPPORTUNITY ANNOUNCEMENT

INFORMATION

1.1 Purpose

The All of Us Research Program is a historic effort to gather data from at least 1 million people, with the goal of accelerating health research and medical breakthroughs, and enabling individualized prevention, treatment, and care. Due to the scale of the program – specifically the number of individuals enrolled and longitudinally retained – data linkages are required to minimize participant burden and enrich the data resource, the All of Us Researcher Workbench (RW). NIH aims to establish a platform and partner to acquire, process, and link data to All of Us program participants’ records without recontacting them (i.e., through passive means).

Consequently, the purpose of this announcement is to solicit and review proposals, evaluate potential partners, and fund the future All of Us Center for Linkage and Acquisition of Data (CLAD). The intention of CLAD is to expand the type and research utility of program data by acquiring new passive data streams and providing additional methods to acquire data for existing data streams, such as electronic health records (EHRs).

CLAD will provide the technology and services necessary to implement this vision by centralizing capabilities within the program to: 1) implement a new cloud-based Data Linkage Platform (DLP), 2) implement a set of priority data linkages, and 3) align data with interoperability standards and clinical terminologies. DLP system security and protection of participant privacy and data will be of the highest priority and must adhere to NIH and All of Us data security and privacy requirements. Simultaneously, CLAD will lead to the development of analytic tools for these linked datasets that can be used within the All of Us RW, which is operated by the Data and Research Center (DRC). This will serve to jumpstart analyses using the linked data. Moreover, CLAD shall incorporate a disclosure review activity to reduce the identifiability of linked data, in preparation for delivery to the All of Us RW. The DLP must be ready for deployment and be capable of integrating with necessary All of Us partners and existing platforms, within 12 months of award. The base period linkage datasets, tools, and documentation must be ready within 18 months of award.

1.2 Overview

This Research Opportunity Announcement (ROA) includes three phases. These phases aim to identify an offeror that has or can readily assemble a promising technology platform and services with capabilities to acquire and link passive data streams with existing All of Us participant data.

This performance also includes delivering related documentation, research and analytic tools, training, and partner technical assistance relative to the scope described herein.

The three phases are as follows:

● Phase I – Submission and Evaluation of Full Technical and Cost Proposal

● Phase II – Data Linkage Platform (DLP) Capabilities Demonstration and Evaluation

(Invitation Only)

● Phase III – Negotiation and OT Award (Invitation Only) https://www.researchallofus.org/data-tools/workbench/ https://allofus.nih.gov/protecting-data-and-privacy/precision-medicine-initiative-data-security-policy-principles-and-framework-overview https://allofus.nih.gov/protecting-data-and-privacy/precision-medicine-initiative-data-security-policy-principles-and-framework-overview

The NIH intends to down-select at each phase resulting in making one OT award in Phase III.

Depending on program requirements, the NIH reserves the right to omit Phase II, or to combine the Phase submissions as needed.

SECTION 2: ISSUING OFFICE AND POINTS OF CONTACT

2.1 Contact Information Agency

National Institutes of Health (NIH) 6710B Rockledge Drive, 4th Fl. Bethesda, MD 20817

Program Name: All of Us Research Program

Opportunity Title: All of Us Center for Linkage and Acquisition of Data

Opportunity No.: OTA-23-003

Contact Information:

All of Us Awards Management Branch (AMB) Email: AllofUsDTPDAwardRequirements@nih.gov Subject: OTA-23-003 - Center for Linkage and Acquisition of Data

2.2 Questions

Questions about this Research Opportunity Announcement (ROA) may be submitted and should be concise and direct. In addition, the relevant section, page, and paragraph of the ROA must be referenced for each question. Questions must be submitted to the email address listed in section

2.1 at least 14 calendar days prior to the full technical and cost proposal submission deadline. If submitted less than 14 calendar days from the deadline, they may not be answered.

2.3 Amendments

Updates to this ROA will be posted to SAM.gov: https://sam.gov/content/opportunities

The ROA can also be accessed via the All of Us Funding Opportunities webpage:

https://allofus.nih.gov/funding-and-program-partners/funding-opportunities

SECTION 3: AWARD INFORMATION

3.1 Award Type and Authority

The funding instrument for the anticipated award is the Other Transactions (OT) awarding mechanism. An OT award is not a grant, cooperative agreement, or contract, but instead uses Other Transactions Authority (OTA). The anticipated OT award will be made pursuant to the Other Transactions Authority (OTA) granted to the NIH Director under the 21st Century Cures Act, as specified in Section 402(n) of the Public Health Service Act, 42 U.S.C. 282(n).

https://sam.gov/content/opportunities https://allofus.nih.gov/funding-and-program-partners/funding-opportunities https://allofus.nih.gov/funding-and-program-partners/funding-opportunities https://grants.nih.gov/funding/other-transactions.htm

3.2 Funding and Period of Performance

3.2.1 Anticipated Funding

The program anticipates making one award in Phase III, based on the description of work in section 5, subject to funding availability. NIH encourages partnering to provide the best solution within a single proposal to the capabilities and milestones outlined in section 5. The submitting entity (as defined in SECTION 6: ELIGIBILITY INFORMATION) may submit alone or as a primary entity in a submission that includes one or more other collaborating entities. An entity is limited to a single proposal submission as a prime, but may be included as a collaborating entity on one or more other submissions.

Total costs for CLAD, if all milestones are funded, is anticipated not to exceed $30 million in aggregate for the base period. Specific milestones to be funded at award will be determined during the negotiation process.

3.2.2 Period of Performance

The period of performance anticipated for this OT award will be an 18-month base period, and may include up to four (4) twelve (12) month optional periods. The NIH reserves the right to extend the period of performance beyond the negotiated period. The estimated start date is subject to the agreed upon date of the final award and the availability of fiscal year (FY) funds.

SECTION 4: PROGRAMMATIC BACKGROUND

4.1 All of Us Research Program

The All of Us Research Program is a historic effort to gather data from at least 1 million people living in the United States, with the goal of accelerating research and medical breakthroughs and enabling individualized prevention, treatment, and care. The All of Us Research Program is guided by a set of core values, including diversity, transparency, and accessibility. The program is creating a national research resource to inform thousands of studies, covering a wide variety of health conditions. Researchers use data from the program to learn more about how individual differences in lifestyle, environment, and biological makeup can influence health and disease.

The All of Us Research Program is committed to enrolling a cohort of participants who reflect the diversity of the U.S., including groups historically underrepresented in biomedical research.

The program characterizes populations that are underrepresented in biomedical research as groups that historically have low representation rates in biomedical research studies (including clinical trials). Groups underrepresented in biomedical research include (1) individuals identifying as from racial and ethnic minority groups and/or (2) sexual and gender minority groups; (3) children and older adults; and people with (4) disabilities, (5) barriers to accessing health care, (6) lower incomes, (7) limited educational attainment, and/or (8) residents of rural areas.

The All of Us Research Program has enrolled more than 596,000 adult participants since 2017, including over 520,000 who have consented to sharing EHR data. All of Us comprises a partner network of more than 100 organizations supporting community outreach and engagement, https://allofus.nih.gov/about/core-values https://allofus.nih.gov/about/diversity-and-inclusion https://doi.org/10.1371/journal.pone.0234962 https://www.researchallofus.org/data-tools/data-snapshots/ participant enrollment and retention, study operations, and longitudinal cohort management. The majority of program participants have enrolled in the All of Us Research Program with support from a network of Health Care Provider Organization (HPO) partners, including Regional Medical Centers, Federally Qualified Health Centers, and U.S. Department of Veterans Affairs Medical Centers. In addition to a strong facility-based enrollment network, the All of Us Research Program supports nationwide virtual enrollment of eligible adults, ages 18 and older.

Participants can complete initial protocol activities and provide biosamples remotely from 50 states and most U.S. territories.

4.2 Purpose of CLAD

Currently, All of Us Research Program participants contribute a range of data for research use, including information from surveys, EHRs, physical measurements, wearable devices, and biosamples. Due to the scale of the All of Us Research Program, including the number of individuals enrolled and longitudinally retained, data linkages are essential to minimize participant burden and enrich the data resource. Thus, the primary purpose of CLAD is to provide new sources of linked data related to All of Us participants.

By integrating linked data, the program aims to expand analytic pathways within and across lifestyle, environment, and biology. The CLAD cores and platform will enable data linkages across foci, such as, but not limited to, climate and environment (built, physical, and social);

medical records including EHR, image and waveforms, and medical claims; registries; and sensor data. Fundamentally, data linkages will be selected based on the principles of expanding the understanding of a participant’s health, increasing research utility of the data, aligning data collection with the expectations of consented participants, quality, and to increase retention metrics.

While the geographic distribution and cost of recontacting program participants further highlights the need to acquire new data through linkage, there are existing data streams that require alternative methods to reduce data missingness. For example, the majority of participant EHR data is shared through records transmitted from the All of Us Health Care Provider Organization network. This results in EHR data gaps caused by factors such as participant mobility, primary or specialist care received outside of the enrollment HPO, and lack of EHR interoperability. As a result, CLAD will pursue new approaches to augment existing EHR data for All of Us Research Program participants that are aligned with the consent and HIPAA authorization that participants provide. This will serve as an important function to reduce the missingness of EHR data in the All of Us research dataset.

4.3 Relationship Between CLAD and the Data and Research Center (DRC) The All of Us Research Program DRC maintains the data repository and supports the RW platform, enabling researcher data access for the program. The CLAD DLP is envisioned to acquire, process, store, and deliver data to the DRC and the All of Us RW. The CLAD DLP will not provide access to outside researchers for the purpose of analyzing the linked data.

https://allofus.nih.gov/funding-and-program-partners/health-care-provider-organizations https://allofus.nih.gov/about/protocol/all-us-consent-process https://allofus.nih.gov/about/protocol/all-us-consent-process https://allofus.nih.gov/funding-and-program-partners/data-and-research-center

An important aspect of the CLAD awardee’s efforts will be to forge a strong working relationship and partnership with the DRC for successful integration of CLAD data linkages, tools, and documentation into the data processing pipelines and for dissemination to the All of Us RW. It is expected that the CLAD awardee will work as a partner to support the DRC on challenges, errors, or questions that may arise on the dissemination of data linkages, documentation, and tools to the RW.

4.4 Awardee High-Level Tasks

The awardee will provide program support to:

● Manage the overall award including project management and reporting.

● Provide scientific and methodological expertise to acquire and process new data linkages.

● Design, develop, and implement a cloud-based DLP.

● Acquire linked data on behalf of the All of Us Research Program.

● Link new datasets to All of Us program participants without recontact.

● Develop analytical tools for processing linked data to ‘jumpstart’ analyses in the RW.

● Document data acquisition, processing, and tools.

● Collaborate with the All of Us Research Program staff and the DRC to process and disseminate newly linked data, documentation, and tools into the RW.

● Collaborate with other All of Us Research Program partners as required.

● Ensure all work is done consistent with laws, rules, policies, principles, and regulations that govern the All of Us Research Program.

● Ensure that the DLP and related systems adhere to NIH and All of Us security and privacy requirements, and maintain the trust of All of Us participants through the security of participant data.

4.5 Resources and Previous Announcements

Offerors should familiarize themselves with the All of Us Research Program by reviewing public resources which includes:

● Program Overview: https://allofus.nih.gov/about/program-overview

● Partners: https://allofus.nih.gov/funding-and-program-partners

● Protocol: https://allofus.nih.gov/about/all-us-research-program-protocol

● Enrollment Website: https://www.joinallofus.org/

● Research Hub: https://www.researchallofus.org/

● Informed Consent Process: https://allofus.nih.gov/about/protocol/all-us-consent-process

● Informed Consent:

https://allofus.nih.gov/sites/default/files/Consent_to_Join_AoU_English.pdf

● Authorization to Share EHRs for Research:

https://allofus.nih.gov/sites/default/files/AoU_HIPAA_Authorization_23June2022.pdf

● Precision Medicine Initiative: Privacy and Trust Principles:

https://allofus.nih.gov/protecting-data-and-privacy/precision-medicine-initiative-privacy-and-trust-principles

Previous and current funding announcements related to the planned CLAD award:

● RFA-PM-16-001: Precision Medicine Initiative Cohort Program Coordinating Center

(U2C) https://allofus.nih.gov/ https://allofus.nih.gov/about/program-overview https://allofus.nih.gov/funding-and-program-partners https://allofus.nih.gov/about/all-us-research-program-protocol https://www.joinallofus.org/ https://www.researchallofus.org/ https://allofus.nih.gov/about/protocol/all-us-consent-process https://allofus.nih.gov/sites/default/files/Consent_to_Join_AoU_English.pdf https://allofus.nih.gov/sites/default/files/AoU_HIPAA_Authorization_23June2022.pdf https://allofus.nih.gov/protecting-data-and-privacy/precision-medicine-initiative-privacy-and-trust-principles https://allofus.nih.gov/protecting-data-and-privacy/precision-medicine-initiative-privacy-and-trust-principles

● OTA-20-001: All of Us Participant Technology Systems Center (OT2)

● OTA-23-002: All of Us Participant and Partner Services Center (OT)

SECTION 5: CLAD AWARD STRUCTURE, CORES AND FUNCTIONS,

BASE PERIOD DATA LINKAGES, BASE PERIOD MILESTONES, AND

OPTIONAL PERIODS

5.1 Award Structure

The CLAD OT is structured such that it consists of four primary cores. Offerors are encouraged to propose partnerships to provide expertise across the four primary cores within their proposal.

There will be a single award for this OT. Offeror’s proposals must cover the work of all four cores as well as the milestones and objectives outlined in this section. Partial solutions will not be accepted. The CLAD awardee will be required to work collaboratively as a team and across All of Us Research Program partners to achieve CLAD and program objectives.

The four cores in the CLAD award are:

1. Research Management Core (RMC): this core will have overall responsibility for CLAD activities, including managing cross-core and cross-partner coordination, cost management, schedule delivery, risk mitigation, partner management, and the overall methodological, scientific, and technical direction of the work.

2. Data Platform Core (DPC): this core will be responsible for the methodological, scientific, and technical activities related to designing, building, testing, deploying, and maintaining the cloud-based DLP.

3. Data Linkage Core (DLC): this core will be responsible for the methodological, scientific, and technical activities related to data linkages.

4. Program Compliance Core (PCC): this core will be responsible for the alignment of CLAD activities with laws, rules, policies, and regulations that govern the All of Us Research Program.

5.2 Cores and Functions

5.2.1 Research Management Core (RMC)

The RMC, working with the All of Us CLAD Other Transactions Program Official (OTPO), leads CLAD, and will provide scientific, technical, and program management services supporting data linkages for the All of Us Research Program. The RMC will engage and collaborate with NIH and program partners to provide scientific, program operations, and technical expertise necessary to leverage the CLAD data linkage platform to facilitate the expansion of program data streams available on the All of Us RW. The RMC will support the following CLAD functions:

● Leadership, management, administration, coordination, and staffing

● Oversight of platform development, linkage efforts, and compliance

● Contributing scientific and methodological expertise

● Alignment with All of Us Research Program strategic and innovation planning

● Award management, monitoring, and reporting coordination with NIH https://allofus.nih.gov/sites/default/files/funding_opportunity_ptsc.pdf https://sam.gov/opp/f9fba45eefaf41a6915e3c7ef784ab37/view

5.2.2 Data Platform Core (DPC)

The DPC will be responsible for designing, building or customizing, testing, and deploying a cloud-based DLP. The DLP must be compliant with the All of Us Research Program Data Security Framework (see Appendix A) which will exceed the FISMA Moderate baseline.The DPC may utilize an existing platform such as a commercial data integration platform or a platform used for data linkage on other projects. These examples may require modification and configuration to support CLAD. A pre-existing platform, or a platform consisting of two or more distinct, yet integrated software products that have been implemented together to serve a similar need, may also be proposed. Platforms should be configurable to the CLAD requirements, and may have custom built extensions based on CLAD requirements. The program expects that All of Us data is not commingled with non-All of Us data (e.g., data from other studies). Conversely, a completely custom-built platform based on CLAD requirements may also be proposed where such a solution can support the required milestone schedule.

During the development of this platform and until the platform receives Authorization to Operate (ATO) from NIH, the awardee will use synthetic data acquired from the All of Us Research Program, when available, or sourced by the awardee to enable development and testing of the DLP capabilities.

Data practices will need to comply with the All of Us Research Program Core Protocol and applicable All of Us data use policies. A collaborative relationship with the All of Us DRC and other program partners will be required to ensure that all required data transmission protections and disclosure practices are adhered to. During the course of the All of Us Research Program, some participants will withdraw from the program. Under such conditions, the CLAD awardee must be prepared to remove all data related to these program participants from the DLP.

The DPC will support the following DLP functions:

● Develop the DLP

○ Design, build and/or customize, test, and deploy the DLP.

● Data Management and Production

○ DLP Database Management: The software system designed to store, retrieve, query and manage all CLAD data.

○ Extraction, Transformation, and Load (ETL): Combining data from differing sources into the DLP using repeatable pipelines.

● Identity Management

○ Participant Identity: Utilization of a Master Study Participant Index (MSPI) for positive participant identification, authentication, data provenance, tracking of sourced data, and related functionality.

○ Participant Linkage: Linkage of new data streams to All of Us Research Program participant records in a secure and privacy-preserving manner.

○ Partner Identity: Provision access and permissions for CLAD to the awardee, program partners, and NIH staff as needed.

https://allofus.nih.gov/protecting-data-and-privacy/precision-medicine-initiative-privacy-and-trust-principles

● Integration Services

○ Ingestion and Exchange: Intake and exchange of diverse data types and formats through methods such as APIs, plugins, point-to-point, HL7® Fast Healthcare Interoperability Resources® (FHIR), secure file transfer, or other types of interfaces. Data may be acquired from partners or data providers; data will be exchanged with the DRC.

○ Platform Configuration: Configure DLP for ingestion of acquired data and matching to All of Us Research participants.

○ Ontology and Terminology services: Ensure incoming CLAD data is compliant with agreed upon healthcare data/interoperability standards. Map data to the Observational Medical Outcomes Partnership (OMOP) common data model (CDM) or another format that eases the ingestion of the data by the DRC. Provide terminology services as needed to fulfill relevant milestone objectives..

● Monitoring

○ Provide modules for monitoring CLAD activities through data dashboards, reports, and visualization. Access will be provided to the Other Transactions Program Official (OTPO) and designees.

● Sustainability

○ Ensure proper handling of data and decommissioning of the DLP to include:

backup, transfer, and destruction of data; approaches to transfer all data, tools, software, documentation, and other artifacts to another All of Us Research Program partner at the conclusion of this effort; and DLP decommission.

The All of Us Research Program is open to offerors proposing additional platform functionalities if they achieve CLAD goals, milestones, and objectives.

5.2.3 Data Linkage Core (DLC)

The DLC will be responsible for linking externally acquired datasets to All of Us Research Program participant records. These data will be used by investigators on the RW. It will also help fill in gaps in program participants’ longitudinal record and their lived experience. The DLC will use the DLP to execute the work required to produce linked datasets. All linked datasets will be stored within the DLP. These datasets will be transferred to the DRC. Each linkage will be guided by methodological innovation, data quality, and data governance. Details on the specific linkages during the base period are described in section 5.3.

Generally, quality refers to agreement with requirements, dependability, fitness for use, or meeting expectations. Data quality is important as it impacts generalizability, power, and the accuracy of conclusions drawn from the data. The All of Us Research Program uses several steps in the data curation pipeline to assess and improve EHR data quality. These steps include quality control checks by HPOs prior to sending data to the DRC; use of automated tools to flag data issues; and further cleaning, de-identification, and de-duplication checks. Two additional data quality elements are layered onto these steps: 1) production of periodic data characterization reports and 2) application of a All of Us created data quality framework of five data quality dimensions (DQD) that capture the fitness of use for All of Us EHR data (see Appendix B).

Data governance is an important aspect of CLAD as it refers to the authority and control over the management of the linked data. Governance enables partner collaboration; informs data architecture and structure; allows secure data access, control, lifecycle management, sharing, transfer, and use; operationalizes policies, procedures and standardization; and facilitates analytics and monitoring. It also covers data loss prevention and incident response. Offerors will develop a data governance plan to include these elements.

During the base period of this award the DLC will produce several data linkages as described in section 5.3. Linked datasets will be delivered to the All of Us DRC in the OMOP CDM standard format or another format that eases the ingestion of the data by the DRC. In some cases, a CLAD data linkage dataset may contain data elements that necessitate an extension to the OMOP CDM or other standardized vocabularies. As such, offerors should make a recommendation on these extensions to support the representation of a specific CLAD linkage in the RW. Simultaneously, the DLC will produce documentation for each linkage and tools to be deployed to the All of Us RW to “jumpstart” analyses of these data.

Linkages will be conducted in a manner that protects and preserves the privacy of All of Us Research Program participants. Linked datasets must be compliant with the All of Us Research Program core protocol, informed consent, research authorization for EHR data (where appropriate), other applicable state and federal regulations regarding the protection of participant privacy, and the RW data tiering structure. Of note, the RW has three tiers: 1) a Public Tier with data that anyone can browse; 2) a Registered Tier for registered data for authorized users; and 3) a Controlled Tier that may be accessed only by researchers who have completed all data access requirements and trainings. To better understand the alignment of the RW tiers with the All of Us informed consent see Appendix D.

An important aspect of each data linkage is to match All of Us Research Program participants to data sources with high certainty. This is of vital importance as it affects research utility and potentially the return of data to study participants. Note, linkages may be done with combinations of participant personally identifiable information, and/or, through methods that utilize privacy preserving techniques. The matching efforts should be done such that the offeror’s efforts result in an acceptable level of sensitivity, specificity, positive predictive value, and negative predictive value.This will be reviewed and approved by the OTPO.

The DLC will support the following functions for data linkage:

● Data Governance and Quality

○ Data Governance: Develop and use or identify and adapt a best practice data governance framework to create and execute a CLAD data governance plan.

○ Data Quality: Use the existing All of Us data quality steps and elements in whole or in part and adapt it as necessary to create and execute a CLAD data quality plan, or replace this approach with another data quality assurance framework, as defined by the offeror.

● Data Linkage

○ Requirements: Gather requirements and analyze alternatives for linkage data sets.

Make recommendations on the preferred dataset and related approaches to meet requirements.

https://www.researchallofus.org/data-tools/data-access/

○ Acquisition and Linkage: Acquire and link data to participant records using appropriate methods (e.g., privacy preserving record linkages) and match with high certainty.

○ Data Modeling: Define and analyze the data requirements needed to support the process of data linkages. Produce visual representations of the data and information system, modeled at various levels of abstraction.

○ Dataset Production: Conduct manual and automated “cleaning,” edits, and quality checks for linked datasets.

○ Transcoding: Support transcoding of data to standards such as the OMOP CDM, claims data / X12, Logical Observation Identifiers Names and Codes (LOINC), RxNorm, Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT), and United States Core Data for Interoperability (USCDI).

○ De-identification: Execute the PCC-developed processes for data obfuscation or degradation, such as statistical methods for disclosure protection and de-identification.

● Data Documentation

○ Documentation: Produce a methods report cataloging each data linkage by acquisition tradeoffs, processing, de-identification procedures, linkage accuracy, and other related elements.

○ Statistical analyses: Produce analyses to identify patterns and trends in the linked data.

● Tool Development

○ Tool Production: Develop tools in R, Python, and Jupyter notebooks for basic analyses and modeling to provide an analytic jumpstart for researchers using linked data on the RW. Future tools may be developed for other programming (e.g., SAS or STATA) and non-programming (e.g., Excel, Power BI) environments. DLC developed tools must be well-documented.

● Dissemination and Support

○ Dissemination: Transfer linked datasets, documentation, and tools to the DRC for deployment on the RW.

○ Training and Technical Assistance (T&TA): Develop learning materials and/or deliver virtual training in the use of the linkage datasets and related tools for DRC personnel and/or RW users. Assist the DRC in answering questions from RW users about the linked datasets, documentation, and tools.

5.2.4 Program Compliance Core (PCC)

The PCC will be responsible for the alignment of CLAD with laws, rules, policies, and regulations that govern the All of Us Research Program and those that govern the potential data sources for linkage. The PCC will proactively identify internal and external policy, regulatory, and legal requirements, as well as propose solutions, and manage these challenges.

An important aspect of this work is the protection of participant privacy and the prevention of disclosing information about All of Us study participants that may result in harm. Disclosure review refers to proactive measures the program takes to maintain participant confidentiality and prevent disclosure of confidential information. The offeror shall develop policies and procedures for disclosure review of CLAD data. Examples of procedures include data degradation, using derived items, collapsing categories, and not releasing data where the cell size is below a specified number.

The PCC will work with the DRC to ensure that linked data is consistent with existing data de-identification, degradation, and obfuscation techniques. The PCC will develop new techniques with the DRC as needed, and ensure that data is produced in a manner that is consistent with the RW data access tiers.

In addition, the PCC will be responsible for preparing materials for review by the All of Us Institutional Review Board (IRB) and responding to IRB questions and comments, as needed.

These activities will be done in collaboration with the All of Us Research Program Policy Office, Research Compliance Branch, and DRC.

The PCC will support the following functions:

● Compliance: Ensure CLAD is compliant with relevant laws, regulations, and policies that govern the All of Us Research Program. The PCC will regularly report on compliance activities to the All of Us Research Program and activate procedures in the event of unexpected or deleterious events.

● Disclosure Review: Offeror will collaborate with the DRC in disclosure review activities to align linkage datasets with the All of Us informed consent and RW tiers (see Appendix D).

● Institutional Review Board (IRB) Activities: Prepare CLAD-related IRB materials and respond to IRB questions. Provide support, as needed, for the assessment, evaluation, development, and monitoring of research protocols.

● Privacy and Security: Assure and safeguard the All of Us Research Program participant privacy and the security of the DLP as this relates to CLAD.

5.3 Base Period Data Linkages

The data linkages to be implemented during the base period are outlined in this section. These data linkages have been selected following principles described in section 4.2. In addition, these linkages were chosen to address EHR data missingness, initiate the inclusion of environmental data for exposure science research, and to support current and future linkages which rely on multi-level geocoded addresses.

The linked data will be de-identified, degraded, or obfuscated consistent with the All of Us RW tiers as required. A methods report, basic descriptive statistics, and analytic tools will be produced for each data linkage as described in the DLC. Data will be securely transferred to the DRC. The awardee will be responsible for compliance with program policy requirements.

Offerors will include in their proposal candidate dataset sources to be acquired for the mortality, residential history, geocoding, and health care claims data linkages. Conversely, for the environmental health data linkage, the All of Us Research Program is specifying a dataset from the Centers for Disease Control and Prevention (CDC) as described in section 5.3.4.

Additionally, a Health Information Network (HIN)/ Health Information Exchange (HIE) pilot study will be proposed by the offeror to acquire EHR data through HIN/HIE as described in section 5.3.6.

5.3.1 Mortality Data

The DLC will acquire and link mortality data to All of Us Research Program participants. These data will include the date of death, the underlying causes of death, and other related data elements.

5.3.2 Residential History

The DLC will acquire the residential history for All of Us Research Program participants. The residential history is the list of postal addresses a participant has lived at over the course of their life. The degree of confidence in the accuracy of the addresses and matching metrics will be documented. Only addresses in the U.S. and U.S. territories are to be considered. The collected addresses will conform with an appropriate standard(s), to be determined through a thorough analysis of different address standards. This data will be delivered to the DRC. Some of this data may not be releasable to the RW.

5.3.3 Geocoding

The DLC will convert All of Us participant residential addresses, derived in section 5.3.2, to enable multi-level geospatial data linkages. Each address will be geocoded to levels including the latitude and longitude, census tract, census block, zip-3, zip-5, county, and state. The offeror will specify one or more rural/urban classification schemes (e.g., Rural-Urban Commuting Area (RUCA)) for additional coding. Each participant address will be coded to these schemes. This data will be delivered to the DRC. Some of this data may not be releasable to the RW.

In addition, a strategy for the preparation, implementation, and maintenance of geographic data for the All of Us Research Program data linkage activities will be developed. It must comply with HHS Geospatial Data Act Reporting requirements and standards promulgated by the Federal Geographic Data Committee. The strategy will be updated on a yearly basis. The offeror will support the program in responding to the yearly HHS Geospatial Data Act Reporting activity with aspects related to CLAD.

5.3.4 Environmental Data

The All of Us Research Program is interested in understanding the relationship between where a program participant lives and their related health outcomes. These relationships can be analyzed across three broad focal areas, including the built, physical, and social environments. Examples of data associated with these environments and the related health concerns may include air quality and asthma, access to parks and physical fitness, or water sources and oral health.

Linking a program participant’s residential address to a specific location on the Earth’s surface, and then mapping it to a variety of geospatial levels, provides the mechanism to make geospatial based data linkages. These linkages are typical to geospatially related information databases, often referred to as “geomarkers.”

It is important to note that incorporating geomarkers into the All of Us RW relies on the use of personal identifiers (addresses or geocodes) for the actual linkages. These linkages are subject to IRB approval.

In order to capture a broad range of well-characterized measures related to health, the offeror will implement a data linkage to the Environmental Justice Index (EJI), developed by the CDC.

This linkage will include associated sub-indexes and data. The EJI incorporates datasets at the census tract level from federal and non-federal entities. The EJI ranks each census tract for 36 factors, grouped into three modules.

These three modules are:

a) Environmental Burden: air pollution measures (e.g., ozone, PM2.5), potentially hazardous and toxic sites, built environment, transportation infrastructure, and water pollution;

b) Social Vulnerability: racial/ethnic minority status, socioeconomic status (e.g., poverty, housing tenure, lack of health insurance), and household characteristics (e.g., 65+, civilians with disability) measures; and

c) Health Vulnerability: estimates of asthma, cancer, high blood pressure, diabetes, and poor mental health.

The DLC will acquire the EJI data and link it to All of Us participant records based on geospatial information acquired in section 5.3.2 and section 5.3.3.

5.3.5 Health Care Claims Data

The All of Us Research Program is interested in acquiring standardized health care claims data from care participants received. Health care claims data provides a focused clinical dataset in cases where EHR and other clinical data are unavailable. Acquisition of these data will allow investigators to answer research questions unique to health care claims. For example, questions related to health care utilization or revenue codes and other cost information. Moreover, it may serve as a metric to evaluate the completeness of All of Us Research Program EHR data.

Health care claims and EHR have some overlap. However, health care claims data differs as it contains data elements such as primary diagnosis/principal diagnosis, National Provider https://www.atsdr.cdc.gov/placeandhealth/eji/index.html

Identifier (NPI), charges/revenue codes/total charges, place of service, and insurance plan name or program name.

The DLC will acquire health care claim data in the X12 837 format, specifically data conformant to the following claims data reporting implementation guides: 837 Health Care Claim:

Professional, 837 Health Care Claim: Institutional, and 837 Health Care Claim: Dental.

However, there may be instances when the 837 Post Adjudicated Claims Data Reporting:

Professional, 837 Post Adjudicated Claims Data Reporting: Institutional, and 837 Post Adjudicated Claims Data Reporting: Dental, or other 837 Health Care Claim formats are the only formats available and/or need to be acquired.

The DLC will acquire available health care claim data on All of Us Research Program participants, including those available on Medicare and Medicaid beneficiaries. Health care claims data may be acquired from a combination of sources, including the Centers for Medicare & Medicaid Services (CMS), all-payer claims databases, data aggregators, and other sources of participant health care claims data.

5.3.6 Pilot for Health Information Network (HIN) / Health Information Exchange (HIE) HINs and HIEs are an emerging approach to acquiring EHR data for research purposes. HINs and HIEs primarily support the exchange of health data, including EHR data, for patient care, public health surveillance, emergency services, and individual access requests from patients.

Acquiring EHR data from HINs and HIEs may partially fill gaps in the EHR data currently received for All of Us Research Program participants. All of Us obtains a signed HIPAA authorization form from participants who agree to share their EHR data. This authorization meets the requirements of HIPAA, as well as relevant state regulations (based on the self-reported state where care was delivered to the individual participant), for disclosure of protected health information (PHI) to All of Us.

The DLC will design, develop, and implement a pilot program to collect EHR data from HINs and/or HIEs. EHR data acquired from HINs and/or HIEs will initially reside within the DLP and be transferred to the DRC. It should be noted that although the All of Us Research Program believes that the participant HIPAA authorization allows us to collect EHR data though HIEs/HINs, providers’ sources that connect to HIEs/HINs, and the HIEs/HINs themselves, may have questions about participant authorization, identity proofing, and data restrictions which will need to be resolved. Therefore, the DLC will advise and make recommendations to the OTPO (and/or designees) to resolve these policy and legal considerations.

Currently, about 80% of enrolled participants meet the criteria for inclusion in one or more groups historically underrepresented in biomedical research. All of Us Research Program participants are broadly distributed throughout the U.S.and territories (see https://www.researchallofus.org/data-tools/data-snapshots/ for more information). Consequently, offerors will include as part of their ROA response the design of a pilot to acquire EHR data with the following criteria:

a) Aligns with All of Us Research Program’s benchmarks for enrolling participants who identify with groups historically underrepresented in biomedical research https://x12.org/flow/health-care https://allofus.nih.gov/about/diversity-and-inclusion https://www.researchallofus.org/data-tools/data-snapshots/

b) Results in sufficient data quality

c) Includes one or more candidate HIN/HIEs

d) Ensures broad geographic coverage to increase the volume of matching to All of Us

Research Program participants.

5.4 Base Period Milestones

The OT award made in Phase III must meet the milestones detailed in this section. CLAD base period milestones will be implemented within a phased project schedule. The schedule and milestones below account for key tasks for the CLAD awardee and all collaborating partners, including, but not limited to, onboarding, systems integration, security approvals, data acquisition, testing and quality assurance, pilot testing, platform launch, and transfer of linked datasets to the DRC. The CLAD milestones are applicable to the work of the RMC, DPC, DLC, and the PCC. All work under the milestones and objectives will be reviewed and approved by the OTPO (and/or designees), unless otherwise specified. The OTPO (and/or designees) may request revisions based on these reviews.

Associated with each milestone is an estimated timeframe for which the work is expected to be completed. Some of these time periods are lengthy because they are dependent on other milestones being completed or because of objectives inherent to the milestone. For example, the objectives for a data linkage milestone may begin early after award. However, the linkage may be dependent on the transfer of data (e.g., personally identifiable data or privacy preserving tokens) from the DRC, a platform ATO, and IRB approvals. Nevertheless, these time periods are meant to be a guide. Offerors, based on their understanding of the requirements, may adjust the timeline for these milestones in their proposal. The work must be completed within the base period.

Milestone 1 - Award Onboarding and Access Estimated Timeframe: Month 1

● Objective 1.1 - Participate in the award kick-off meeting with the OTPO and Awards Management Team. Establish an ongoing management meeting schedule with the OTPO (and/or designees) and Awards Management Team.

● Objective 1.2 - Establish a leadership plan and award management structure supporting NIH requirements that will serve to administer and lead the scientific, administrative, and project delivery activities.

● Objective 1.3 - Provide documentation of award staff requiring access to key program tools, documentation, resources, and program meetings, and verify access and establish offboarding procedures.

● Objective 1.4 - Complete necessary All of Us Research Program training for award staff.

● Objective 1.5 - Establish processes to support required milestones and financial reporting to the OTPO and Awards Management Team on a monthly basis.

Milestone 2 - Regulatory and Policy Estimated Timeframe: Months 1-4

● Objective 2.1 - Review All of Us Research Program Institutional Review Board (IRB) Core Protocol, Protocol Appendices, Informed Consent, HIPAA Authorization, and associated protocol and program governance materials.

● Objective 2.2 - Familiarize staff and review existing program integrations, services, tools, processes, and functional capabilities of NIH and program partners, including the Data and Research Center (DRC), Participant Technology Systems Center (PTSC), and other awardees.

● Objective 2.3 - Establish a plan for integrating the DLP, services, tools, and processes into the CLAD environment and developing security boundaries as required to meet program security requirements.

● Objective 2.4 - Prepare and deliver a monthly report on compliance activities.

Milestone 3 - DLP Security, Privacy, and Authorization To Operate (ATO) Estimated Timeframe: Months 2-12

● Objective 3.1 - Conduct NIH and CLAD Security Team introductions, communications planning, ATO requirements, and timeline review.

● Objective 3.2 - CLAD ATO

○ Objective 3.2.1 - Support iterative development, documentation, and review of required ATO materials or as directed by the All of Us Research Program Information Systems Security.

○ Objective 3.2.2 - Prepare the CLAD final ATO package for review(s) and approval.

○ Objective 3.2.3 - Submit the ATO package for review and approval.

● Objective 3.3 – Finalize a System Security Plan (SSP) and supporting FISMA artifacts in advance of engagement with third-party controls assessor.

● Objective 3.4 - Identify, contract, and coordinate with Third-Party Assessment

Organization (3PAO) for assessment of the CLAD security controls and penetration testing.

● Objective 3.5 - Support review, mitigation, and retesting of security controls assessment and penetration testing results.

● Objective 3.6 - Support integration of all CLAD systems, tools, and processes, with existing All of Us platforms, as needed, as well as ongoing collaboration with the All of Us Security Operations Center (SOC).

● Objective 3.7 –…