A CIO’s perspective on the options for health system analytics

03 Apr A CIO’s perspective on the options for health system analytics

Buying an EMR is an important decision, but choosing an analytics solution is far more important. In today’s healthcare marketplace, installing an EMR is table stakes. Granted, it’s necessary and expensive table stakes, but it’s still just the starting point.

The real key to transforming healthcare performance lies in analytics and the humans that use and make data-driven decisions. An EMR captures the data. Analytics uses that data to deliver the insight needed to improve the quality and cost of care.

Improving quality and cost is on everyone’s mind. At the organization where I serve as CIO, Memorial Hospital at Gulfport in Mississippi, it is a critical priority. The majority of our volume comes from Medicare and Medicaid beneficiaries and the uninsured. We are a not-for-profit, single-hospital system. We have to focus on costs and quality in order to continue to serve our community.

Fortunately, we’re advancing steadily along the path of putting infrastructure in place to drive the necessary improvement. We rolled out our integrated EMR this spring and we are now implementing our analytics solution.

I started this article by stating how important analytics is. Choosing what type of analytics solution to implement was not a decision we took lightly. I want to outline here the factors we considered as we made that choice.

I wouldn’t say that selecting our EMR solution was easy, but the fact that there were only a handful of viable options certainly simplified the process. Choosing an analytics solution was a different story. A wide variety of analytics solutions are available and they all claim to drive quality and cost improvement. We looked at BI tools. We researched multiple vendors with point solutions that address areas like capitated payments, fee-for-quality, and ACOs.

Ultimately, we decided that the right solution for our enterprise-wide analytics strategy would be an enterprise data warehouse (EDW). But even then there were several possible paths to take. We could build our own EDW, we could adopt our EMR vendor’s emerging EDW solution, or we could implement an EDW solution from a third-party analytics specialist vendor.

We quickly dismissed the option of building it ourselves. We simply didn’t have the time or resources for a trial-and-error, homegrown approach. That left us to decide between our EMR vendor’s EDW and a specialist’s solution. We went with the specialist’s solution.

Our EMR vendor’s EDW was relatively inexpensive and there was something attractive about the convenience of having one less vendor to manage. Still, I approached their EDW offering with some skepticism. I trusted their ability to handle all of the transactional functionality that is an EMR vendor’s core competency, but analytics is not part of that core competency.

Ultimately, we set three criteria as essential in an vendor. Any analytics vendor we selected would have to demonstrate the following.

A significant track record with analytics

EMR vendors really don’t have an analytics track record. Their analytics experience lies mainly in tactical operational reporting. They can easily tell me how many of my patients are on a certain medication, but my improvement initiatives will require much greater sophistication.

Specialist vendors, on the other hand, have been living and breathing nothing but analytics for years (and sometimes even decades). The best ones can share concrete examples of how their solutions have driven measurable quality and cost improvement.

The agile data architecture required to handle big data

Our EMR vendor is obviously an expert on transactional systems architecture, but that doesn’t translate to expertise in architecting a powerful analytics solution that runs on a completely different type of database. With so much volatility in healthcare today, I wanted to be sure I had a flexible architecture for analytics that could expertly adapt to new rules, standards, vocabularies, and use cases.

The ability to integrate data from multiple systems, including competitors

This was a huge consideration for us. EMR vendors are generally unwilling or unable to pull data from external sources, particularly competitive systems. We needed a solution that was source-system neutral and only the third-party analytics specialists could deliver that. Integrating data from just about any system you can imagine is their core competency. My understanding is that some EMR vendors have recognized the need to allow integration of data from beyond the EMR, but they are years behind the specialists in terms of doing this well.

I recently came across a 2013 survey by CHIME that found that 80 percent of CIOs believe analytics is an important strategic goal, but that only 45 percent feel they have a handle on it. I don’t claim to be an expert on analytics, but I hope that this brief account of my experience so far will be helpful to some.

My biggest piece of advice to any colleague that has yet to tackle analytics is to get started as soon as possible. I believe that CIOs need to change. Our focus can’t be just on the bits, bytes, databases, and servers. All of that is still an important element of what we do, and I have a staff that takes care of those details, but my focus as CIO is to provide data and information to all stakeholders—our executives, our clinicians, our patients, and more—to help drive better outcomes. That means a top area of focus for me is on analytics.