The "Lowest Common Denominator" Theory of Analytics

This afternoon we had the pleasure of listening to one of our clients describe their analytics work dealing with the Medicare bundled payment initiative. They highlighted some of the reports that they had assembled from a variety of different hospital-based systems, described in detail their use of Microsoft's PowerPivot analytics tool, and showed how an advanced analytics shop can drive an entire healthcare initiative. Both presenters highlighted the fact that they have Masters degrees in public health, not in information technology or computer science, and yet had developed the skills to integrate their healthcare business knowledge with the analytical techniques necessary for success. Having worked with this team for about a year, we know that they're absolutely fearless in exploring new technologies and techniques. Nothing is too difficult to learn – as long as a skill adds value they believe they must master it. Their executive team recognized the importance of proper hardware and software support for the analytics function, and purchased new, highly capable desktop workstations earlier this year, despite having encountered severe financial hardship resulting from hurricane Sandy. This isn't a part-time or low-level effort - they have three full-time staff members managed by a Director-level leader solely dedicated to analytics on new payment systems.

Other organizations will listen to a presentation like this and make comments like "Well, our IT department won’t let us upgrade our version of Excel, so we're still running the version that's almost 7 years old", or "We’d like to use more advanced tools, but Stanley has a hard time understanding them and he works on most of those projects, so we stick with the easier tools for everyone”. These statements seem to be accompanied by two assumptions; first, that these limitations are impossible to overcome; and second that they won’t create a significant barrier to success. And therein lies the "lowest common denominator" theory of analytics, which is to take the least capable staff member on the least capable computer and make that the standard for the entire department – and believe that it’s still good enough because… well, we don’t have any other choice.

Sorry, fellas. That's like saying "We can't afford to pay for a good second baseman, but that doesn't mean we can't win the World Series, right?” As healthcare organizations delve into more complicated payment systems, the ability to analyze data quickly and thoroughly becomes a critical capability for any organization. The best clinical teams – the best management teams – and anyone else involved in these organizations is flying blind without good data and information to lead them. The most successful healthcare organizations will compete on analytics – they’ll use their analytical capabilities to create a competitive advantage in the same way that companies like Netflix and Amazon do. They realize that the newer analytical tools are much better than the ones commonly used even a few years ago, and that a lack of current skills will create a lack of actionable information, which will ultimately create a lack of action.

Absense of these types of attitudes can totally stifle the creativity of an analytics department (or finance, or decision support, or wherever this function is located) because really good analysts can't stand working in a shop using outdated equipment and tools. Good people will leave and find jobs in environments where their skill sets are appreciated and encouraged. Lesser skilled people, however, will remain in place, satisfied to be using the same old techniques on the same old equipment that they've been using for the last decade. And management will receive the same types of static reports and dated analyses that they've been receiving for decades, which will eventually stifle any ability to proactively address their environment.

Fortunately, there's hope on the horizon in the form of the new "Masters of Business Analytics", or even the more specialized "Masters in Health Information" programs that are being introduced in many universities. These programs teach analytics theory and practice, often alongside business courses such as those covering healthcare payment systems. Graduates of these programs should be able to understand business problems that healthcare organizations will face, and have the skills to perform the types of analyses that will produce a type of information required for success. Hopefully these new graduates will find homes in forward-looking healthcare organizations that will provide an environment in which they can thrive.

But the real catalyst for analytics must come from higher levels of management. Managers must recognize and embrace the need for new skills and technologies, even if they don't posess those skills themselves. Organizations that are unwilling to accommodate these new skill sets will find themselves increasingly less successful in the new healthcare environment.