Data Mining in Healthcare
As said by many in the industry, healthcare is broken and needs to be fixed ASAP. Many proposals to fix healthcare are presented daily, but a consistent theme remains: you have to know what's wrong with healthcare to fix healthcare. To understand what's wrong, you must present empirical data for your theory on the cause of healthcare's issues.
The history of healthcare in America is rich. Therefore, there is ample data supporting theories on where the most need and attention in fixing healthcare is. But the concern often presented is that of data isolation. Have we collated, stratified and reviewed the data? Have we mined the data appropriately?
Data mining is the process of analyzing data to make sense out of it. It's descriptive statistics that helps researcher's understanding the situation at hand. Data mining turns random numbers that to the ordinary eye mean nothing into meaningful expressions and analysis. There are many tools involved in the process of data mining. One of the tools used in the discipline of Lean Six Sigma and adopted in data mining is the Pareto Chart. The Pareto Chart is used to determine most frequent data. Such a tool helps stratify data that is then further analyzed for mining.
In this TEDx from 2013, Joel Selanikio shares the importance of data mining. His talk presents the surprising seeds of a big data revolution in healthcare. He shares some haunting images of stacks of paper being used to collect health information in local communities in parts of Africa for global health issues. What's alarming is how much time is spent gathering data and the unfortunate reality that not all of it can/will be used in the decision-making process for the health of that community. Such information could benefit the healthcare system if properly mined.
There's no point collecting data for it not to be used. In public health, we are taught the concept of Assessing, Planning, Implementing and Evaluating programs. Data collection falls under the assessment part. If we collect data and fail to use it to plan interventions that could be implemented, we've only achieved 25% of our goal. It is of utmost importance that healthcare leaders follow the process all the way through. Indeed, there will be times where data tells you not to intervene, but you must review the data to know what it's attempting to share.
The benefits of data mining include relevant information to help improve the quality of care provided, roadmaps on the path's to avoid, and an improved cost of care. With a focus on cost, McKinsey and Company presented that 12-17% of healthcare cost could be saved through proper data mining. How's that for benefits.
Turn those numbers into words, let it tell you a story.
