Monday, February 9, 2009

Setting Expectations

A few weeks ago when I presented to government staffers, one person used a recent experience with uncoordinated care to suggest that data standards harmonization has not been successful. Standards are very important, but standards alone cannot reform the US Healthcare system.

Thus, as we evaluate the work of ONC, AHIC (now NeHC), HITSP, CCHIT, HIPSC, and the NHIN pilots, we need to be very specific about the criteria for success and how we'll measure it. Here's my sense of the performance of each of these groups and the next steps required to align expectations for the work ahead.

There are hundreds of stakeholders for healthcare interoperability including payers, providers, patients, employers, government agencies, vendors, and standards development organizations. Over the past 20 years, these groups have worked together informally, but there has not been a process to create a single list of priorities or a coordinated project plan. ONC was charged with creating a single strategic plan for all these stakeholders to create a uniform set of data standards, pilot architectures for data exchange, document privacy principles, and certify compliant systems which adhere to functional criteria created collaboratively. It has coordinated the work of many subgroups that have successfully executed this work, as described below.

AHIC prioritized standards harmonization use cases for the country based on multi-stakeholder input. It produced 3 use cases in 2006, 4 use cases in 2007, 6 use cases in 2008, and a prioritized list of standards gaps to fill in 2009. The successor to AHIC, the National eHealth Collaborative will continue this work by prioritizing value cases that, in addition to standards, will include model processes (such a model of the “ideal” care coordination process) best practices (such as incorporation of ePrescribing into provider workflow or managing the communication of results out to the referring physician) and frameworks, (a service oriented architecture for health information exchange).

HITSP received 13 use cases plus a privacy/security standardization request from AHIC. All of these use cases led to unambiguous interoperability specifications that were delivered within 9 months of HITSP receiving the request. The standards were chosen by consensus in an open transparent way. Many controversies were resolved along the way. At this point, standards for personal health record exchange, laboratories, biosurveillance, medications, quality, emergency first responder access to clinical summary data, home care device monitoring, immunizations, genomic data, hospital to hospital transfers of records including imaging data, public health reporting, and patient-provider secure messaging are finished. There is still some discussion in the industry about the Continuity of Care Record verses the Continuity of Care Document, but all the other harmonization has broad consensus among the majority of stakeholders. This is A+ work. However, if broad and immediate implementation of these standards is the expected metric, then there are many factors outside of HITSP that need to be aligned including widespread implementation of products which use the standards, incentives to exchange data, and the resources to do so.

CCHIT has certified products based on hundreds of detailed functional and standards conformance criteria. It has achieved broad industry recognition as the place to develop a roadmap for the features and interoperability requirements to include in the yearly revisions of Healthcare IT products. This is A+ work. However, if the metric for success is plug and play data sharing between all certified products, then there are many factors outside of CCHIT that need to be aligned such as a very specific expectation of the data to be shared, by whom, and under what circumstances. CCHIT certified systems today are excellent at laboratory and e-prescribing data exchange. They can also import and export standardized clinical summaries, and this year will add network-interface features that let them share them via Health Information Exchanges. But there is not a live nationwide grid of healthcare data exchange to plug into as of 2009, so data exchange among CCHIT certified products is dependent on local and regional data exchange infrastructure.

The NHIN pilots demonstrated a successful architecture for pushing data between stakeholders, for query/response to pull data, and appropriate security protections. Many of these pilots have become production systems in their localities. This is A+ work. However, if the NHIN is to be judged on the number of transactions exchanged among cities, regions, and states, there are many factors outside of the NHIN, such as variations in state law and privacy policies that are still a work in process.

The HISPC groups documented the privacy practices in 34 states and territories. There was great sharing of experiences and best practices for polices, opt-in/opt-out models of consent, and technical security protection. This is A+ work. However, if HISPC is to be judged on the creation of a uniform national privacy policy there are many outside factors, such as HIPAA pre-emption by state laws, that are still a work in process.

Thus, to me all the efforts to date have been very successful, as measured by the goals each organization was given. Their work continues and every year brings new accomplishments.

However, the public, Congress, and the healthcare industry may have different expectations. I believe the work ahead is to set achievable expectations based on what can be widely implemented, with clear milestones and metrics for success. Here are the measurable goals I would suggest:

1. Require an EHR for every patient in the country by 2014 through the use of incentives, penalties, and regional implementation teams. ONC should monitor progress and help overcome resource and policy roadblocks.

2. In parallel with the EHR rollout, ONC/NeHC/HITSP/CCHIT/NHIN/HISPC/CMS, and the private payer community should work together to ensure 90% compliance with e-Prescribing, electronic laboratory exchange, and clinical summary exchange for all existing EHR users by the end of 2011, aligning this work with existing CMS incentives for adoption and adding new ones where needed.

3. NeHC should add data exchange goals in a stepwise fashion after 2011, based on stakeholder priorities, success of goals 1+2, and resources available.

These goals are focused, require that we ensure the interoperability specifications for the standards are easy to implement, require that we finalize the specifics of the data exchange architecture to be used, and require that we make some determination of the privacy policy that is good enough (patient controlled via a PHR or opt-in consent are my recommendations).

We can publish quarterly statistics of our progress and all stakeholders can objectively evaluate our success or failure based on well defined metrics.

By setting expectations, allocating the resources, and monitoring our progress, I am confident that we as a country can do this. We put a man on the moon with 7 years of effort. When we align our strategy, resources, and strength of will, we can do anything.


Unknown said...

I agree with all that Dr Halamka has said, but it worries me that our reimbusement system does not yet provide the incentives for coordination of care that is facilitated by the exchange of clinical data. As long as our health care services are reimbursed on an incident basis (unrelated to outcome) rather than on an episode basis, there is not substantial financial reward for the expense or the changes in behavior need to promote the improvements that can be gained by access to information across the many settings of care.

Anonymous said...