Wednesday, June 16, 2010

The Massachusetts Ad Hoc HIE Workgroup

Over the past 30 days, an ad hoc multi-stakeholder group convened by the Massachusetts eHealth Institute (MeHI) worked to inventory existing state HIE capabilities, develop a list of prioritized services, and define functional/geographic gaps in services.

Here's a summary of work thus far.

Core Services needed for meaningful use in 2011

Routing Services - includes point to point delivery, filtering to ensure the right content is delivered to the right recipient i.e. do not require every edge system to decide what labs are reportable as this may change over time, de-identification if the recipient does not require patient specific demographics and translation services such as mapping of different versions of standards into common over the wire packages. Routing is likely to be a combination of local edge/community routers such as eClinical Works EHX/SafeHealth/CHAPS and statewide routing such as Surescripts/NEHEN. The role of the HIE would be to ensure that all stakeholders are covered by a service provider.

Directory services
a. Provider/Facilities Directory - A centralized service leveraging existing databases such as CAQH's provider index or MHQP's provider index. This information can be used for routing messages to the right organization for delivery to the clinician.
b. Payer/Plan Directory - A centralized which offers routing information for each Payer and details about their plans which will facilitate administrative transactions.
c. Public Health entity Directory service - A centralized service which provides routing information for all reportable lab/syndromic surveillance/immunization data.

PKI/Identity Services - A centralized service which can support authentication of endpoints (people and organizations) as well as support e-prescribing of controlled substance workflows.

Immunization/Syndromic Surveillance/Reportable Lab Registry Service - State-level repositories which are aggregated centrally and shared with cities and providers per patient consent.

Quality Data Center includes aggregation, reporting, CMS/Payer submission - a distributed service that stores multi-institutional clinical and administrative data needed to compute HEDIS, PQRI, and pay for performance measures. A single organization may offer all three services (aggregation, reporting and quality data submission) or those tasks might be divided among multiple organizations. The point was made that hospitals already have services in place to report inpatient data to CMS - this would not need to be replaced, but instead leveraged to satisfy other data requirements for the state.

Consent services. There are two distinct kinds of consents - those which are obtained by providers for clinician to clinician exchange. It is likely such consents will remain decentralized. Another type of consent is a record of consumer preferences which is managed by the patient. That would ideally be a state-wide centralized service.

Needed for meaningful use in 2013 or beyond OR a high value service that is likely to contribute to HIE sustainability. While planning for these should be started promptly, completion is not necessary prior to implementing the High Priority services.

Master Patient Index/Record Locator Services - there are three basic types of master patient indexes
a. probabilistic matching based on demographics. There is no persistent linkage of medical record numbers or any unique ID used for the patient. Lookups do matching in real time
b. deterministic - There is persistent linkage of medical record numbers based on a unique ID that is used exclusively inside the database and not used by the patient. Staff is required to fix linkage problems - duplicates, wrong patient, demographic data problems etc.
c. unique identifier - The state issues a unique identifier that is presented to all caregivers and used as the identity management key for all HIE activities. There are significant privacy concerns and costs associated with this approach.

All Payer Database - a centralized repository that might include administrative and clinical data to support utilization review, case management, disease management, opportunities for efficiency improvement etc.

Radiology Image exchange services (could possibly include EKGs too) - this is likely to be a distributed offering from multiple private companies. The role of the HIE would be to provide a list of qualified companies offering such services.

Event notification service to identify payers or providers of patient status changes. For example, a Patient Death notification service would enable providers to cancel appointments and payers to stop paying claims. Hospitals know the birthdate of their patients but rarely know when they die because the place of death and the place of birth are often different. Could also include other status changes such as: Adoption, protective services, marriage, hospice, PCP changes, coordination of benefits, etc... A distributed service. The role of the HIE would be to identify these services and connect stakeholders to them.

Routing Service for Patients which includes routing CCDs and other data to patients. Patient Routing could take the form of a portal, secure email, or interactive voice response. This is likely to be a distributed offering from multiple private companies. The role of the HIE would be to provide a list of qualified companies offering such services.

Personally Controlled Health Record Services - a distributed service (such as that provided by Google, Microsoft, or Indivo) that can gather clinical information and serve as a patient controlled repository of their lifetime health record data. The role of the HIE would be to provide a list of qualified companies offering such services.

Important to consider as service options to sustain the HIE

Pharmacosurveillance services - a distributed service that could provide early detection for such issues as the Vioxx/Heart Attack association, as well as trends in medication use such as antibiotics.

Pre-auth approval rules service (such as radiology ordering) - a distributed service that incorporates payer rules into web services, eliminating call centers and enhancing workflow.

Pharmacy directory for fax-based pharmacies, a centralized service. Surescripts charges extra for e-prescriptions that must be rendered as faxes. It may be value added to provide a list of such pharmacies so that alternatives (NEHEN, other services) could be used. More research is needed.

Advanced Directives Service - a centralized repository that would store advanced directives on behalf of the patient and make them available to caregivers as needed.

Medication safety analysis service - a distributed service providing decision support about the safety of patient medications. Commercial companies do this today.

Formulary Service - a distributed service that offers access to those formularies not provided by Surescripts.

Patient Educational materials service - a distributed service that provides disease specific educational materials. Commercial companies offer such services today.

Disclosure logging services - the HIE could log data exchanges among stakeholders. In a sense, the HIE router could become a disclosure logging service. Although a central service can be offered, there will still need to be local services for exchanges that do not pass through the HIE such as plaintiff attorney record requests.

Vocabulary Services which includes access to/mapping of LOINC, SNOMED-CT, ICD9/10, RXNorm. This should be a combination of local and central services. The role of the state HIE would be to negotiate preferred rates with companies that offer vocabulary services. The Federal Government (ONC and NLM) are also working on vocabulary services repositories.

I2B2 clinical research services which include distributed data mining/searching tools using Open Source I2B2 approaches. This is central distribution of open source code/standards with local implementation. The role of the state HIE would be to educate stakeholders about this offering and provide links to the organizations supporting its implementation.

Our next step is to ensure all priorities and functional/geographic gap analysis is vetted with all stakeholders. We'll then determine the best governance options for the services we'll offer and complete our state operating plan to ensure that high priority services are available to every patient, provider, and payer as needed to support 2011 Meaningful Use data exchanges.


Don Fluckinger said...

Given the complexity of the laundry list, is the timeframe realistic?

What is your perspective on that as of today?

Steve Waters said...

This is certainly an ambitious project to say the least. The technical aspect is relatively easy to address though a staged implementation plan to deliver basic services then subsequently adding services over time makes the most sense to me. The real challenge is how to create a sustainable business model that allows healthcare providers to choose the software services that best meet their business needs while interacting with a central (HIE). It would appear the best way to make these solutions a reality in the shortest possible time is by leveraging service models. My company, Semper Vivo, focuses on delivering HIT services directly to providers while working with software and other technology partners to promote interoperability. EHR interoperability starts with increased EHR adoption.