There are three different models for health information exchanges, with each varying according to how the patient’s data is stored and accessed by participants. Randy McCleese, FHIMSS, CHCIO, vice president of information systems and CIO at St. Claire Regional Medical Center, Morehead, KY, offers this summary of the three types in use across the country:
- Because all data exists in a single warehouse, it is very easy and fast to perform queries against it.
- The model facilitates communitywide data analysis for research and public health purposes since the data is available centrally.
- There may be better expertise in managing central resources due to their scale and the class of products used.
- The entire system is dependent on good management and maintenance of the database.
- Timeliness of data can be an issue because slow data submission from participating systems to the central database will result in inaccurate consolidated records.
- Duplication of demographic and clinical records will be a constant concern because the data on any patient will be collected and submitted by multiple providers.
- A larger up-front investment in central resources is required.
“The biggest drawback with centralized data is physically managing the data of a million-plus patients in one system,” McCleese says. “Feeding that data in to the system in an efficient way, keeping the record updated and making sure you’re matching data to the right patient, is crucial to creating an effective resource. That’s a challenge when you’re handling millions of messages a month, which is typical for an HIE.”
- Since the healthcare provider stores the data , there is no conflict concerning who owns the data.
- The data is always current because the provider does not have to send updates to a central data bank.
- The failure of a single system doesn’t cripple the whole network.
- The provider must ensure that only authorized third-party systems connect to the network.
- Standards and guidelines for inter-provider access to patient data are still being defined.
“There is a trade-off. If we keep the record within our system here and our network is down for some reason, then the providers in the exchange are not going to be able to access that information,” McCleese says. “But if you have the HIE maintain all that patient data, there is a cost associated with that. Our hospital serves roughly 160,000 people in this area, but an HIE could have data on millions of people.”
Data discrepancy also can be an issue with the federated model, McCleese says. If a patient is identified differently at two facilities, it may be difficult to match up the records and get a complete EHR.
- Faster access for queries, when allowed by providers in control of the information.
- More support from the HIE and less administrative or technological burden for the participants.
- Potentially more costly than the decentralized or federated model.
This article appears in the September 2011 issue of HealthLeaders magazine.