[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Newsgroup Home]
[news.eclipse.modeling.mdt] Re: SBVR component proposal- interest in healthcare application

I am excited at the prospect of participating in this. I'm retired, so I can spend time doing what I want to do :-).

Please bring me up the learning curve on your activities.

The RHIO market is very competitive, with IBM and GE among the major players.

I'm in South Carolina, where there is a large active and retiree military population; consequently, there's a large VA and DoD presence, with associated medical records. If appropriately swift management of infectious disease (and terrorist activity) is to be a reality, there will have to be an interchange between state, private and government health facilities. Just evaluating potential state and private RHIO information will not be adequate.

Such a mechanism does not exist today; governmental rules stand in the way. It's possible the SBVR component activity may suggest modifications to the government's rules plus design characteristics of the health records systems that would be acceptable to Congress.

Of course, all privacy concerns will need to be addressed.

RHIO implementations include rules and agents for security management. There are some tools for managing protocols, but, as far as I know, they are targeted at a research setting. If your are interested, I can dig up the url of a company I know - it may be relevant at the SBVR level because it indicates a boundary of functionality, which would lead to a piece of the SBVR semantics.

RHIOs need to contain "business = infectious disease" rules for identifying outbreaks, current treatments, and recommending the most effective (cost, efficacy) countermeasures. Such population variables as age, race, sex, and overall medical conditions must be considered in the rules. (In a multinational setting, such as the Caribbean, or ,where there is a country with semiautonomous, cultural different regions, such as Iraq, language, culture, and government regulations must be considered.)

Another use case is this: during a patient visit, the physician accumulates the patient complaint and test results data, establishes a (tentative) diagnosis, and then, in real time, this information is passed to a local agent of the primary agent. The local agent establishes the best treatment option for that patient, either autonomously, given a local copy of the recommendations from the primary agent, or, in real time dialog with the primary agent. Since the local agent is implemented by different EMR (Electronic Medical Record) vendors, I think the SBVR component is relevant in guiding implementation.

Within the primary agent, is a calculating engine, which may be proprietary by company, to establish the effectiveness metrics and prioritize the treatments.

A key critical success factor, independent of how the process is implemented- it can even be manual- is that medical practice must be changed to insure follow up and record treatment effectiveness; ie, did the patient recover + any other relevant data. In terms of implementation and the SBVR component, patient reporting might be done through a medical web portal that results in information back to both the local EMR and, eventually, to the primary agent. The SBVR component might be relevant in establishing the semantics for the dialog.

In general, it seems to me any SBVR implementation will need to satisfy a collection of use cases based on the above description.

My bottom line assertion is that a necessary condition for SBVR specification validation is that it be validated through appropriately robust medical disease control use cases and trial implementations.