Instead of the traditional path of Big Data programs leading to insightful metrics in healthcare, use insightful metrics to lead Big Data programs.
The promises of Big Data initiatives and the value they can bring to a healthcare organization are well known: a big data platform, associated analytics, and the resulting metrics will lead to predictive actionable insights for high impact decision making. However, lesser known is the success rate of these initiatives. According to Gartner, 70 percent of healthcare organizations surveyed in 2013 had already purchased or were planning to invest in big data systems; as of 2015, only 29 percent had actually de-ployed big data systems and furthermore only 36 percent believe their big data investment have or will generate positive ROI. This begs the question—why is it so hard for big data programs to succeed? And more importantly, what can you do to make sure it will be successful? Answer: follow the metrics.
Big Data programs are complex and more often than not overwhelming due to the vast amount of data companies have these days. How many of these scenarios are familiar?
■ Finding out what data repositories or data sets to start with stalls the program off the bat or activities run off track as the amount of data to access, process, and analyze becomes a seemingly impossible feat. This exacerbated by the continuing explosion in volume data fueled by EMRs, medical images, monitoring devices, and personal fitness applications.
■ The initiative is underway for several years but clear progress or value-add is difficult to demonstrate; without the demonstrated outcomes, the initiative ends up losing funding as healthcare organizations increasingly face cost pressures.
■ Collaboration and support from necessary parties such as compliance, security, customer care and clinical teams become bottlenecks as they cannot see what specific outcomes they are working towards and instead focus their efforts on other high priority projects.
We have seen a national clinical care provider see its Big Data program stall for years as the program was challenged with the volume and types of data that needed to be managed. Another initiative at a major healthcare payer / provider organization faced challenges with defining and garnering support from other groups because it was unclear what was needed of them and what it was for.
A relatively simple yet effective way of spearheading and driving Big Data programs is to define a key set of business metrics to measure. This should be a short list of 5 – 10 values which would inform decision makers on areas such as efficiency or effectiveness of their operations. For healthcare organizations, this could include percentage decrease in total office visits or percentage increase in scheduled telephone visits. Having this laser focus on generating the values for these few key metrics will serve as a clear beacon for the overall effort. The metrics guide the specific data sets and analyses required as well as clearly communicate to stakeholders what the program objectives are.
Once the initial platform and first set of metrics values are successfully determined, the value-addition and case for on-going support of the program is much easier to demonstrate. Using key business metrics to guide the program along with an agile approach for continuously developing the big data platform will keep the program on course and progressing. Focused small steps are the key to a successful Big Data program. It is these small steps that will eventually grow into greater insights such as quickly identifying high-risk patients and knowing what are the most effective forms of treatment.