The good and the bad (Score 1)

This is my job. I've built EHR software with exposure to the internals, interfaces, and connectivity with external systems. I'm now a healthcare analytics consultant and considered and expert in my field. I feel the need to point out that patient-centric EHR systems are very good for system usage by the physicians and staff at the hospitals because they are optimized to retrieve and work with an individual patient's records (orders, results, etc). The flip side of this coin is that the system which works well at the patient level is not appropriately indexed for enterprise-wide reporting. For example, the database indexes make it so that querying to see if a patient is a smoker is very fast, but querying to see what percentage of the patients smoke is a much bigger task... not because of the size of the data returned, but because the tables are set up to retrieve the data in this fashion. As such, it makes reporting a much bigger task than predicted by the people requesting the reports. I'll go on to say that the hierarchy of hospitals is more complex than many people imagine. Hospitals are often owned by a parent organization which divides the hospitals into clusters (usually by geographical location) that share a common EHR system. Reporting overall numbers requires manual calculation or the use of an EHR or data mart with ETL jobs running nightly to capture the data which can't be queried real-time. There's a LOT of overhead and we haven't even touched on the fact that hospitals are bought and sold all the time so an organization will own many hospitals using many different EHR systems so normalizing the data is a task in itself. Welcome to my world. :)