Anticlue

HIMSS MSHUG

It is going to be a busy week, and I’m looking forward to my first HIMSS experience. I must say San Diego is breath takingly beautiful, and having lunch yesterday with pals by the arena was a pleasure. In February the temperature is above 40 and there is sun! A true joy for a new Englander like me.

Started the day attending the MS-HUG Tech Forum, the welcoming address clarified Microsoft’s strategic view of HIT focused on changing the health organization from a vendor centrix to a technology centric growth. Seems very high level out of the box thinking from my narrowed prespective, imaging a technology centered view of when evaluating vendor selection.

There was a good pyramid of the different architectural views, amazing what can be done when we focus away from the fires of the day.

A cool idea present is to have share point for portals made on the different user communities, clinician, student, researchers, employees, and most importantly patients. Imagine an interactive chronic disease portal and the enhanced benefits of that informational exchange.

Blackford Middleton, MD performed the keynote presentation. It was truly representative of the clinician’s side of the problem, and the frustration exhibited. Due to our disparate, unintegrated systems, it is almost impossible to deliver quality of care at the time the patient presented. It’s impossible to assess the facts, correlating and diagnosis, when one just can not get at the information.

The truth of the matter is that the ROI on any clinical data capture benefits are soft, and it is difficult to prove the value until the system is in place. One example was Brigham’s which saw a 1 million dollar a month benefit
in improved charge capture. Another is the benefits of embedded analytics and decision support, a continuum of clinical decision support. Once the data is reported in an informative manner, clinical documentation will improve by leaps and bounds.

For implementations several good ideas were shared, one of which was displaying a running tally of radiology and lab charges within the longitudinal medical record. Longitudinal medical records refer to the record that is captured over the life of the patient. Another good point was to create medical reminders for consequent or corollary orders.

Along with reminders, there was an example of ten year old business logic hidden within edits on the front end of a system, with no authorizer or owner, and created serious charge capture implications. So a point was to identify who requested what rules, and a yearly review process for rules was recommended.

One surprising point is to ask the question why was that the least beneficial and most costly integration is HL7. The best end result is anticipated to be the symantec web with SOA architecture.

A note to self is to check out documentum document management systems and get a copy of the presentation.

The next presentation was on the New Zealand Health IT Cluster. To summarize New Zealand had a state mandate to implement a collaborative system that provides an integrated view of information from disparate systems. In order to do this all systems collaborate through a health collaboration engine using HL7 Messaging and a Health Services Directory which contains where the patient has been seen and by whom. A good concept shared was to indicate previously chronically diagnosed diseases as a part of the enterprise MPI. One would think allergies maybe beneficial here also.

The collaborative approach indicates that there are several participants in the care of a person, including the individual. All should have access to at least some level of the data to improve quality or care. Obviously it would be important that the exchange be standards based and secure. Also the information is displayed in the UI in an easily digestable. (bulleted) format.

The lessons learnt from our friend across the sea was that a federated model does indeed work, and project management is crucial. A key point was the legacy db was fox pro, not a mainframe wonder.

Next we listened to a case study on the Ann Arbor Community Health Data Exchange, which had 4 separate EMR updating a centralized repository. Before exchanging paperwork was hindering the quality of care, and engaging in a information exchange positions one well strategically for the pay for performance changes later on down the road.

One of the key success factors was community agreement as to what components will be utilized and which components will be excluded. For example substance abuse is not pushed to the centralized repository.

A lesson learned and passed onto others was that workflow is very important and needs to be captured before hand. Also having workflow be closed loop has benefits. The example that really proved this was the automation of the referral process.

Kevin Carr, MD, gave one of the best presentations of the day on developing trust in a health information exchange. The main point of the presentation is that software must reflex what happens in reality, and how will IT support the real world.

The Waterbury health access program is a healthy community federal grant that has two phases first how to get the low income and / or uninsured into health care and then once they are being seen how to keep them in a care and manage that care appropriately.

One key component is patient access and spreading the wealth across family practice physicians. This is managed by a case management combonent so the distribution is evenly dispensed.

The other component is how to manage patients within the system for disease management. The idea was that sharing certain pieces of the clinical record across competing institutions. The solution is that selected information from the disparate longitudinal patient records are exported to a federated web portal. Also inorder to move into the future, we may need to forget past indiscretions and move forward. Wise Words in the politically charged healthcare environment.

Another Key idea was that because this is a community a regulatory body with equal participation from all community data inputs owns the data, having membership from the state and board of health as tie-breakers, if needed. For example a report request and resultant format is sent to a central advisory board, this board then justifies the creation of the report in the format. Any scope changes of the report must also go through the committee.

In a community setting with competing operations, a tack to manage the project is transparent management with unbiased physician and administration leadership. It is absolutely necessary not to have favorite or a “golden operation”, because that will only begin to have the competing operations sparing again.

As with all startups, a sustainability plan needs to be established for the cost of participation after the “seed” grant money is exhausted. The tactic take here was to spread the maintenance cost evenly across the competing operations.

Final a technical infrastructure or architect needs to be establish. Is a centralized or federated model applicable? Surprisingly, this decision is more politically charged then technology charged.

And finally what is Trust?

Treaties and teamwork
Regulatory body
Unbiased PHYSICIAN leadership
Sustainability plan
Technical infrastructure

The next presentation covered a LHIO (localized spin on RHIO). The problem was there were two local paperless independent organizations, Saint Francis Heart Hospital and Cardiology of Tulsa, so before the LHIO the practice was to print, fax to the other institustion, so the other place could scan and import. Basically the automation created two digital islands with humans serving as the bridge.

So to bridge the chasm an interoperability was the pavement, with security as the foundational gravel underneath.

A quick break for lunch and then we were all awoken with the need for a National Health Information Network. Currently we are one of the few remaining industries that does not have all the necessary information available for a customer when they present. And the cost of healthcare is transferring from the employers to the government to finally land on the consumer or patients. Even day to day seemingly simple efforts are bearish, ie immunization records.

The other benefit would be improvement on medical errors, statistics provided in the presentation on Medical Errors.

Due to Medical Errors currently 44,000 – 78,000 Americans die and 770,000 people are injured. In asking the freakonomics question, I wonder how that relates to other portion of statistics. How many injuries are there of handguns?

The only solution is to share data across everyone, but the question is how to get there. IMHO, the cost is very high, and a visionary is needed or a government mandate.

Finally the evening ended with a dinner among my colleagues over at Trattoria la Strada. All in all a great way to spend the day.

Although secretly, I’m itching to hit the convention floor, maybe tomorrow…..