Healthcare Analytics – A Proven Return on Investment: So What’s Taking So Long?

So what do you get when you keep all your billing data in one place, your OR management data in another, materials management in another, outcomes and quality in another, and time and labor in yet another? The answer is…………..over 90% of the operating rooms in America!

That’s right; the significant majority of operating rooms DO NOT have an integrated data infrastructure. In the simplest terms, that means that the average OR Director/Administrator CAN’T give answers to questions like, “of all orthopedic surgeons performing surgery within your organization (single or multi-facility), what surgeon performs total knee replacements with the lowest case duration, least number of staff, lowest rate of complication, infection, and re-admission rate at the lowest material and implant cost with the highest rate of reimbursement?” In other words, they can’t tell you who their highest quality, most profitable, least risky, least costly, best performing surgeon is in their highest revenue surgical specialty. Yes, I’m telling you that they can’t distinguish the good from the bad, the ugly from the, well, uglier, and the 2.5 star from the 5 star. Are you still wondering why there is such a strong push for transparency of healthcare data to the average consumer?

You’re sitting there asking yourself, “Why can’t they answer questions like whose the least costly, most profitable and highest quality surgeon?” and the answer is simple, “application-oriented analysis”. Hospitals have yet to realize the benefits of healthcare analytics. That is, the ability to analyze information that comes from multiple sources in one location, instead of trying to coordinate each individual system analyst and have them hand their spreadsheet off to the other analyst that then adds in her data and massages it just right to hand it off to the next guy, and then….ugh you get the point. If vendors like McKesson, Cerner, and Epic could make revenue off of sharing data and “playing well with others” they would, but right now they don’t. They make their money off of deploying their own individual solutions that may or may not integrate well with other applications like imaging, labs, pharmacy, electronic documentation, etc. They will all tell you that their systems integrate, but only once you’ve signed their contract and read that most of the time, it requires their own expertise to build interfaces, so you’ll need to pay for one of their consultants to come do that for you – just ask anyone who has McKesson Nursing Documentation how long it takes to upgrade the system or how easy it is to integrate with their OR Management system so floor nurses can have the data they need on their computer screen when the patient arrives directly from surgery. Out of the box integrated functionality/capability? Easy-to-read, well documented interface specifications that a DBA or programmer could script to? Apple plug-n-play convenience? Not now, not in healthcare.

Don’t get too upset though, there are plenty of opportunities to fix this broken system. First, understand that organizations such as Edgewater Technology have built ways to integrate data from multiple systems and guess what – we integrated 5 OR systems in a 7 hospital system and they saved $5M within the first 12 months of using the solution, realizing a ROI 4 times their original cost.  Can it be done? We proved it can. So what is taking so long for others to realize the same level of cost savings, quality improvement and operational efficiency? I don’t know, you tell me? But don’t give me the “it’s not on our list of top priorities this year” or the “patient satisfaction and quality mandates are consuming all our resources” or don’t forget the “we’re too busy with meaningful use” excuses. Why? Because all of these would be achievable objectives if you could first understand and make sense of the data you’re collecting outside the myopic lens you’ve been looking through for the past 30 years. Wake up! This isn’t rocket science, we’re trying to do now what Gordon Gekko and Wall Street bankers were doing in the 80’s – data warehousing, business intelligence, and whatever other flashy words you want to call it – plain and simple, it’s integrating data to make better sense of your business operations. And until we start running healthcare like it’s a business, we’re going to continue to sacrifice quality for volume. Are you still wondering why Medicare is broke?

Cloud Computing Trends: Thinking Ahead (Part 3)

cloud-burstIn the first part of this series we discussed the definition of cloud computing and its various flavors. The second part focused on the offerings from three major players: Microsoft, Amazon, and Google. The third and final part discusses the issues and concerns related to the cloud as well as possible future directions.

A company may someday decide to bring the application in-house due to data security or cost related concerns. An ideal solution would allow creation of a “private in-house cloud” just like some product/ASP companies allow option of running a licensed version in-house or as a hosted service. A major rewrite of existing applications in order to run in a cloud is probably also a non-starter for most organizations. Monitoring and diagnosing applications in the cloud is a concern. Developers must be enabled to diagnose and debug in the cloud and not just in a simulation on a local desktop. Anyone who has spent enough time in the trenches coding and supporting complex applications knows that trying to diagnose complex intermittent problems in a production environment by debugging on a simulated environment on a desktop is going to be an uphill battle to say the least. A credible and sophisticated mechanism is needed to support complex applications running in the cloud. The data and meta-data ownership and security may also give companies dealing with sensitive information a pause. The laws and technology are still playing catch-up when it comes to some thorny issues around data collection, distribution rights, liability, etc.

If cloud computing is to truly fulfill its promise the technology has to evolve and the major players have to ensure that a cloud can be treated like a commodity and allow applications to move seamlessly between the clouds, without requiring a major overhaul of the code. At least some of the major players in cloud computing today don’t have a good history of allowing cross-vendor compatibility and are unlikely to jump on this bandwagon anytime soon. They will likely fight any efforts or trends to commoditize cloud computing. However, based on the history of other platform paradigm shifts they would be fighting against the market forces and the desires of their clients. Similar situations in the past have created opportunities for other vendors and startups to offer solutions that bypass the entrenched interests and offer what the market is looking for. It is not too hard to imagine an offering or a service that can abstract away the actual cloud running the application.

New design patterns and techniques may also emerge to make the transition from one cloud vendor to another easier. Not too long ago this role was played by design patterns like the DAO (data access object) and various OR (object relational) layers to reduce the database vendor lock-in. A similar trend could evolve in the cloud based applications.

All of the above is not meant to condemn cloud computing as an immature technology not ready for the prime time. The discussion above is meant to arm the organization with potential pitfalls of a leading edge technology that can still be a great asset under the right circumstances. Even today’s offerings fit the classic definition of a disruptive technology. Any organization that is creating a new application or over hauling an existing one must seriously consider architecting the application for the cloud. The benefits of instant scalability and “pay for only what you use” are too significant to ignore, especially for small to mid size companies. Not having to tie up your cash in servers and infrastructure alone warrants a serious consideration. Also not having to worry about setting up a data center that can handle the load in case your application goes viral is liberating to say the least. Any application with seasonal demand can also greatly benefit. If you are an online retailer the load on your website probably surges to several times it average volume during the holiday shopping season. Having to buy servers to handle the holiday season load which then remains idle during rest of the year can tie up your capital unnecessarily when it could have been used to grow the business. Cloud computing in its current maturity may not make sense to pursue for every enterprise. However, you should get a solid understanding of what cloud computing has to offer and adjust the way you approach IT today. This will position you to more cost effectively capitalize on what it has to offer today and tomorrow.