Prior to 2004, my exposure to information technology in healthcare had been limited to selling an early version of the Oracle database management system to a few hospitals. In 2004, our venture fund invested in a healthcare software company, RemedyMD, that was focused on helping providers and patients make better decisions. After a few weeks, the depth of the problem and the breadth of the opportunity began to manifest themselves. It was easy to see that the overall state of healthcare IT was a generation behind other industries with which I was familiar. Even our portfolio company that provided hardware and software solutions for grocery stores was more advanced than most large healthcare providers.
There are numerous reasons why healthcare has fallen far behind other industries in applying IT. In order to catch up, healthcare needs to carefully analyze how other industries have benefited from improved information technology and learn from them. I think the fi nancial industry is arguably the most advanced user of IT. So, what can healthcare learn from Wall Street? During the past 20 years, Wall Street has pioneered a concept called “predictive analytics” and invested billions to make it a reality. The objective of predictive analytics is to model past performance and use the model to predict future behavior. Those who become adept at predicting events gain a competitive advantage over those who cannot. For example, the ability to predict the outcome of a border war in Turkey could lead to significant gains in the market. Many anecdotal stories have been told about information systems that paid for themselves
in a single afternoon of heavy speculation.
There have been many attempts to apply predictive skills to healthcare informatics, because the need to recognize a pattern, predict a trend, and preempt an event is even more important in healthcare than on Wall Street. Why haven’t these efforts been as successful as they have been on Wall Street? The most signifi cant reason has to do with the way data is captured in healthcare. Data is typically collected in a non-computable format (eg, handwritten notes from a physician, free-form text entries, images, etc). Th is kind of information is virtually useless for any type of predictive analysis. In contrast to this data collection methodology, Wall Street relies exclusively on numeric data stored in a computable format (eg, stock prices, changes in the consumer price index, the popularity rating of the Prime Minister in England).
An additional data-related difference between the two industries is that Wall Street collects highly specifi c data sets and uses them to make predictions. For example, it might be that the unemployment rate in Kenya could be the most accurate predictor of the price of coffee futures. Healthcare has not collected data to this level of specificity because most clinical data is collected by “one-size-fits-all” systems. Many EMR systems are incapable of collecting and fully utilizing specialized data about a patient. For example, few, if any, EMRs can document and use a measurement like a patient’s wrist size. However, this is probably a better predictor of a person’s ideal weight than Body Mass Index (BMI).
A few weeks ago, I met with a US Senator in Washington, DC, to discuss how a new administration might change healthcare funding priorities. His agenda was diff erent than mine. He was preparing for hearings on the cost of healthcare and wanted a soundbite. He asked, “Can you just give me one or two thoughts on how we can control costs in healthcare?” I said, “It is very simple. If healthcare had the kind of pattern recognition and predictive analytics tools that exist on Wall Street today, this would squeeze billions out of total healthcare costs.” He nodded as if to acknowledge my point, but will this translate into action? Perhaps Wall Street can help us predict whether it will.Dr. Kennedy is the President and Chief Executive Officer of RemedyMD.