Using Electronic Medical Records to Their Full Potential
The NY Times carries a story about IBM and Google investing in universities to help fund research and train students in how to deal with massive volumes of data. Apparently these companies can’t find people to help them deal with and capitalize on the mountains of data they generate. This will soon be a problem in the healthcare field, especially in research. Hopefully the current administration’s investment in health technology will pay off and result in much wider adoption of electronic medical record systems. Having electronic forms of such records means programs can be designed to scour these files for information to include in a research study. For example, if I want to study all patients who had pneumonia and see if they got Drug X and measure how long their hospital stay was, then I could use a program to find all those patients and automatically spit out their length of stay instead of pulling mountains of paper charts from the medical records department. More importantly, I can do this on a large scale. Bigger numbers means higher quality research.
Some forms of this currently exist. A consortium of pediatric hospitals (all of whom have EMR systems) pools their data together. Currently, over 19 million patient encounters are included in this database. This means that rare diseases can be studied more efficiently and studies of common diseases have much better power.
However, no training in capitalizing on such data sources currently exists (that I’m aware of). The amount of data is going to grow exponentially over the next decade, just like it has over the previous decade with the internet and growth of electronic systems in commerce. We are going to need epidemiologist, biostatisticians, and clinical researchers that can deal with these massive amounts of data and use them to create a better, more efficient healthcare delivery system.