Query+Health+-+Project+Charter

include component="page" wikiName="siframework" page="Query Health Header"



toc

Challenge
The nation is reaching a critical mass of Electronic Health Records (EHRs) that comply with data and vocabulary standards. The wide deployment of EHRs creates an opportunity to aggregate health care data to provide a broad range of benefits that can contribute towards improved health of individuals and the population as a whole. A standardized clinical information model and a common method for querying data sources are critical to enabling and simplifying data aggregation across widely distributed EHR systems.

There are a number of important uses for distributed queries, including quality measures, disease outbreaks, comparative effectiveness analysis, efficacy of drug treatments and monitoring health trends. These are largely supported today by extracting data from source systems and integrating it into a centralized database where queries and analysis are managed. The Query Health Use Case moves away from the centralizing tendency to “bring the data to the questions” to distributed population queries that “bring questions to the data.” Distributed queries provide access to data, for analysis purposes, while maintaining patient privacy and security by keeping protected health information safely behind healthcare organization firewalls. This will reduce the complexity of managing patient consent and authorizations, audit logs and access lists requirements.

The Query Health project will establish requirements for the clinical information model, distributed queries and results expression, with the objective of giving providers, consumers, researchers and others insight into prevention issues, healthcare research and disease outbreaks. Standards and services for distributed population queries will be selected to enable widespread adoption of the clinical information model and query capabilities. Use of these standards and services will result in increased speed and reduced transaction costs for stakeholders to analyze and apply information, ultimately reducing the cost of healthcare and improving the health of citizens.

Scope
The Query Health Initiative aims to identify the standards and services for distributed population health queries to certified EHRs and other patient data sources, such as HIEs, originating in the routine course of patient care. As a result, information requestors will be able to create and securely distribute queries to data sources directly or via optional network data partners, who serve as intermediaries. The information requestors can distribute queries to data sources, or network data partners, who support the distributed queries; however the data sources ultimately retain control over the decision whether to respond to a query as well as maintain control over the data to be released. Network data partners, when used, may examine queries and pass them on to data sources, and may aggregate and modify the data returned, performing such tasks as masking of provider organization, etc. Data sources, such as a provider organizations, will execute the query against a standard clinical information model, securely return the results of the query directly to the requester or via the network data partner. The Initiative will develop models for the technical and financial sustainability as well as best practices for organizations, management and coordination, data use, data sharing; giving consideration to privacy, security and consent requirements.It will also address methods for extensibility of the clinical information model; specifically those data elements, terminologies, and code sets that enable the queries and results expression.

The Initiative will align with and leverage other initiatives of the S&I Framework (e.g. Transitions of Care); EHR certification criteria; Meaningful Use requirements; and other health IT initiatives. The Initiative will evaluate and leverage where possible existing investments, technology and thought leadership in distributed query, implement the specifications through an open source reference implementation, and evaluate these findings through demonstrations and pilots. The evaluation will help refine these standards.

To support the goals and objectives of this Initiative, the Use Case and its requirements will evaluate and address several outcomes, such as the following (to be refined and further validated by the community):
 * EHRs and other clincial data sources have the capability to transform and map data (or a view of its data) to a common clinical information model;
 * Network data partners and data sources have the ability to participate with selected information requestors and specific queries;
 * Requestors will have the ability to create and securely deliver “well formed queries” to selected network data partners and/or data sources; and
 * Network data partners are able to examine and to pass the requestor information to the clinical data sources. Such information (depending on the functional requirements) may include confirmation of the query; security information of the requestor; and timeliness of the results being sought.

The Use Case scenarios focus on querying against an extensible common information model and returning a final result set. For the purpose of the ONC pilot, the initiative will prioritize and select a few User Stories, based on national priorities and existing research and public health network infrastructure. Additional User Stories, developed by the community, will be analyzed and evaluated to ensure that the architecture framework, standards and services for distributed queries are robust and extensible.

The pilot/reference implementation will be built based on the scenarios described in the Query Health Use Cases. As described in the figure below the reference implementation will involve multiple distributed EHRs being queried, with the return of aggregated results to the requestor.

The pilot/reference implementation may also include establishing requirements for an Intermediary to review the query, review the results, edit/blur the results, and to pre-aggregate across organizations depending on the requester and query. The pilot/reference Implementation sites will be selected based on national priorities and existing research and health information management infrastructure.

Value Statement
As evidenced by the Query Health Summer Concert Series, distributed queries have been used to support a number of healthcare and health related operations, including quality monitoring, public health surveillance, and clinical research. Approaches to date have relied upon dedicated experts exploring standards and services to best utilize data from distributed systems. The value of the Query Health Initiative will be to lower the barrier using consensus-based standards and specifications to support queries for population based/aggregated data from certified EHRs and other community records. The initiative will provide a standardized clinical information model to support implementable, high-value user stories, based on available, shareable and standardized information from EHRs and other patient care systems. The initiative will also provide extensible ‘Query’ and ‘Return Results’ standards and services, enabling interoperability between and among information requestors and data sources. Specification of these standards will assist the development and implementation of software and pilots, and will facilitate analysis of results. Use of these standards and services can result in increased speed and reduced transaction costs for healthcare providers to analyze and apply information regarding prevention activities, healthcare research, and disease outbreaks. These benefits can be combined to reduce the overall cost of healthcare and to improve the health of all citizens.

Target Outcomes
The Query Health Initiative will provide a standardized clinical information model targeting implementable, high-value user stories. Available and shareable information supported by standardized terminologies will be targeted by ‘Query’ and ‘Return Results’ services, enabling interoperability between data partners. Specification of these standards will assist in the development and implementation of software and pilots, in addition to facilitating analysis of results.

The key to realizing the value of query health will be targeting several supporting activities which will include:
 * Identification of **functional requirements** describing key conditions and business rules to enable the query and return of results, while protecting privacy and confidentiality;
 * Identification of consensus approved **standards** and **services** for the distribution of queries and results in support of the functional requirements;
 * When gaps have been identified where existing **standards** don't address a topic area, we will 1) identify an appropriate solution and then 2) propose such a solution to the relevant standards body.
 * Design and validation of technical architectures to meet requirements specified in the Use Case.
 * Development of concise architectural guidance using easy-to-understand documentation, user-friendly tooling and formal models to assist implementers in implementing technical requirements.
 * Development and publication of an **Open Source Reference Implementation.** The standards and specifications within the initiative may ultimately be incorporated into the NIST certification and testing suite. The reference implementation will be provided via an open source license to promote adoption and pilot of the standards and services.
 * Conducting one or more **Query Health Pilots.** Community partners will utilize the reference implementation in the field, testing the utility of Query Health standards, specifications, terminologies, and services in the context of high value user stories. The initial and subsequent pilots of the reference implementation will be used to evaluate the Query Health standards, specifications and services, and provide a feedback mechanism to incrementally enhance and improve the standards in meeting the functional requirements.

Strategic Alignment
The Query Health Initiative will enable population analyses to inform clinical and payment strategies for health care in alignment with both the Affordable Care Act (ACA) and Health Information Technology for Economic and Clinical Health Act (HITECH). From a HITECH perspective, Query Health will leverage the standards and policies that enable the Meaningful Use of patient care and health information exchange.

Roadmap
The Query Health Initiative will progress through an agile and iterative approach, with key milestones and deliverables to be defined by the Community through the initiative Workgroups. From a high-level perspective, the following phases and timelines are laid out to guide the community and the initiative as it begins its work. Once the workgroups have identified specific options for moving forward, there will be further refinement of the timeline and key milestones by the initiative participants and the participant workgroup leads

Stakeholders
The ONC and the Query Health Support Team have engaged leaders and gained commitment from providers, health IT vendors, states, federal partners, and the research community to gain insight and support this effort. As with each of ONC’s Standards and Interoperability Initiatives, the Query Health Initiative is an Open Government Initiative that will be consensus-based, transparent, and open to all interested parties. The following stakeholder groups will be essential to the success of the Query Health initiative. The Coordinator of the Query Health Initiative is Rich Elmore.
 * **Stakeholders** || **Representative Profile** || **Targeted Contributions** ||
 * **Providers, Infection Control Practitioners** || Providers within Health Systems and Hospitals, Accountable Care Organizations, Academic Medical Centers and others. || Prioritized User Stories and Functional Requirements. Pilot Implementations. ||
 * **Informaticists, Research Community, Distributed Query Thought Leaders** || Organizations leading distributed query initiatives such as Mini-Sentinel, OMOP, hQuery, i2B2, SHARP, Universal Public Health Node, PopMedNet Electronic support for Public Health, caGRID, and others. || Leadership in definition of Clinical Information Model, Query Standards, Results Standards. Pilot Implementations. ||
 * **Government Agencies** || Federal, State, HIOs, Local agencies and other organizations responsible for population based analysis, reporting, monitoring and others. || Prioritized User Stories and Functional Requirements. Pilot Implementations. ||
 * **Standards Community** || Clinical Standards Experts. || Support for definition of SDO-independent Clinical Information Model, ‘Query’, and ‘Return Results’ Standards. ||
 * **Health Information Organizations** || State, regional, local or affiliation-organized HIOs and HISPs. || Prioritized User Stories, Clinical Information Model, Query Standards, Pilot Implementations ||
 * **Health IT Vendors** || Mix of Acute, ambulatory and analytics vendors - large and small. || Reference Implementation development. Support Pilot Implementations. Support for definition of SDO independent Clinical Information Model, ‘Query’, and ‘Return Results’ Standards. ||
 * **Privacy and Security Experts** || Consumer/patient and academic experts who represent the public's interests for privacy and security. || Neutral/non-industry privacy and security experts. ||
 * **Patient Advocates** || Patient advocates who act as liaisons between a patient, Health Care Provider(s) and research institutions. || Patient advocacy organizations. ||

Standards
Standardized models and associated terminologies will be essential for retaining data integrity for both the query and the return of results. Vocabulary value sets supporting the prioritized Query Health user stories will be identified by the community participants within the initiative. These value sets will be derived from accepted standards associated with HITECH and will likely include – problems - SNOMED-CT; medications - RxNorm; Immunizations - SNOMED; results- LOINC among others.

The clinical information model enabling query health will be derived by the user stories and functional requirements that the community agrees to in the Discovery phase of the initiative. Several previous modeling and distributed query efforts will provide insight into how this clinical model could work, including, but not limited to ONC’s Transitions of Care Initiative, i2b2, Clinical Element Models, C32 Laika / hQuery, PopMedNet, PCAST Data Element Access Services and others.

Within the initiative, a gap analysis of these and other community identified models will be performed to build on existing work, and leverage best practices and successes from other clinical modeling efforts. The key outcome of the clinical modeling efforts will be to leverage the expertise and experience of Standards Development Organization (SDO) members to create an independent comprehensive Clinical Information Model (CIM) to support the Query Health Use Case and subsequent User Stories. More detail for these models and other inputs can be found in Appendix I.

Guiding Principles
To guide the initiative from conception through design to reference implementation, we will aim to:
 * Minimize burden to providers by leveraging data captured during the routine course of care without adding new workflows or data capture requirements on the provider;
 * Maintain data at or close to the source;
 * Develop a standard clinical information model for use as a common platform for data sharing;
 * Bring population analysis questions to where data are naturally managed, archived, and maintained;
 * Leverage existing intellectual property, technology, and standards for distributed queries to enable rapid implementation of a reference implementation;
 * Leverage internet technologies to enable national scale and performance;
 * Protect privacy by reporting results in aggregate or in "minimum necessary" form;
 * Recommend standards and services that facilitate integration with existing EHR and other clinical data sources; and
 * Keep demonstrations, pilots, and implementation simple (see below)

Key Assumptions – Keeping it Simple
Assumptions are general assertions of requirements that must be met in order for the specific technical information interchange to be implemented. They generally refer to policy, business and technology requirements.
 * The Clinical Information Model (CIM) will use standardized vocabulary and computable data elements, and will attempt to leverage the logical data model and data element definitions from the Transitions of Care Initiative CIM to the greatest extent possible. The ToC logical data model and data element definitions leveraged may be extended and modified to support the selected User Stories and broader requirements.
 * The outputs of other S&I Framework Initiatives (Transitions of Care, Provider Directory, Direct, etc.) will also be given priority as part of the design of the Query Health outputs
 * Each data repository or EHR supporting a Query Health data interface and/or Clinical Information Model will contain protected health information which will be kept and maintained behind each organization’s firewall
 * The data obtained from the Clinical Information Model and returned as the query result will always meet  applicable laws and regulation regarding treatment of PHI. Most commonly, query results will return only aggregated “count” data.
 * Query requests and responses shall be implemented in all ONC sponsored pilots to use the least identifiable form of health data necessary in the aggregate within the following guidelines:
 * ** Disclosing Entity: ** The results returned will be under the control of the disclosing entity (e.g., manual or automated review of queries and/or results).
 * ** Data Exchange: ** Data will be either 1) mock or test data, 2) de-identified data sets or limited data sets each with data use agreements or 3) a public health permitted use under state or federal law and regulation.
 * ** Small cells: ** For other than regulated/permitted use purposes, cells with less than 5 observations in a cell shall be blurred by methods that reduce the accuracy of the information provided.
 * Addressing directories, data rights, and rights for execution of services are to be managed “out of band”, not within the defined information interchanges
 * A query health network of information requestor(s), network data partners and providers (data sources) has been established including (//This will be addressed during later stages of the project)//
 * Agreement of roles of each participant including coordination
 * General protocols for responses will be addressed in the Query Health technical specification (timing, masking, declining, etc.)
 * Any financial considerations
 * Actual data use and subscription agreements between Provider Organization (Data Source), Intermediary (Query Network Data Partner) and Information Requestor (Query Source) are in place
 * 1:1 relationship between each clinical data source and its Clinical Information Model
 * The 1:1 relationship between the clinical data source and the Query Health data interface is a simplifying assumption that for each data source there is one Query Health data interface for the pilot. It does not suggest that there is a 1:1 relationship between the fields in the data source and the Query Health data interface. It is likely that additional fields and data sources (like claims) will be added over time.
 * The CIM will reside as a view of the data present within a clinical data source or will reside in a clinical data repository
 * Appropriate use and/or disclosure rules will be applied to protected health information, de-identified or limited data sets being used as part of the Query Health data interface
 * The Query Health Data Interface will maintain an audit log of patient level identifiers (e.g., numerators of a MU quality measure) for a query response, However the audit log need not include a listing of the relevant patients in the population (e.g, denominators of a MU quality measure
 * Data elements/structures of summary reports such as CCD's produced by EHRs under MU will be leveraged in defining the Query Health Clinical Information Model; however this does not imply it is a patient centric information model
 * Data Extensibility: As new certification and other query requirements evolve (e.g. for MU Stage 2) the base Clinical Information Model will need to be kept up to date and consistent across the nation while allowing for pluggable information models, across multiple networks to meet new or local needs
 * Appropriate privacy and security policies are in place to:
 * Prevent unauthorized disclosure of protected health information
 * Obtain, create and maintain consents, authorizations, audit logs and access lists required to support Query Health implementations
 * Support authorized follow-up by Information Requestors to examine specific patient records or to obtain patient level data as de-identified or limited data set
 * Specify best practice reporting of Small Cell of Results
 * The Query Health architecture will be designed to allow for future phases that may include:
 * Patient matching and the ability to de-duplicate
 * Claims data source
 * Capture of historical data through changes in patient information

media type="custom" key="11138434"

include component="page" wikiName="siframework" page="space.template.inc_contentleft_end"