Health care initiatives put big data clichés to the test

Profile picture for user jreed By Jon Reed June 19, 2014
Summary:
Does big data innovation in health care live up to the hype? Our recent batch of Sapphire Now videos brings promising stories to light - and warns of obstacles that lie ahead.

datasurgery
At the Sapphire Now/ASUG 2014 show, Den Howlett and I filmed several videos with organizations pursuing health care initiatives, all with an analytics bent. The conversations provided a chance to push through stifling big data clichés and dig into use cases.

Health care and data innovation are an uneasy mix from the get-go. Privacy concerns are paramount. Life-or-death scenarios are heavy with legal ramifications for data misuse. Even those with access to health care struggle with a system that is chock full of soul-draining paperwork and extended wait times. There's a reason why the 'customer experience' crowd doesn't focus on the health care vertical.

The organizations we filmed are not backing away from these challenges. But hold on - why is data suddenly a focal point for innovation? After all, health care is no stranger to tedious data gathering and reporting - electronic records are hardly a new development. However: a review of our Orlando videos indicate that past approaches to managing health care data were insufficient:

  • data is too often trapped in silos, preventing identification of important patterns
  • delayed processing times are a bugaboo, sometimes with high stakes consequences for patients
  • 'legacy' analytics systems are unable to handle the data load for innovations like genome sequencing
  • archaic, cumbersome IT systems drain resources and take time from innovative use cases

So what's being done about it? To what extent can in-memory solutions and simplified IT lead to better outcomes? And can access to the right information lead to a better patient experience? Let's have a look.

Dartmouth: using health care tech to personalize patient education

During our first health care shoot, we talked with Dartmouth's Dr. Sohail Mirza, chairman of the Department of Orthopedics and professor of orthopedics at the university's Geisel School of Medicine. With the goal of reducing unnecessary operations through patient education, Dartmouth will soon be embarking on a pilot program using SAP Care Circles. The focus?  Educate knee replacement candidates about their medical choices, via smart phones or tablets.

I tend to lean skeptical about tech for its own sake. I kicked off the video by asking Mirza if tech could really impact our health care dilemmas:

Mirza is bullish on tech's potential to drive health care into new, more transparent models:

Technology can change most things, if not everything about health care, and we at Dartmouth believe it has a critical role and we are pushing for that. In fact, Dartmouth has been very involved with health care reform and leading the nation in new models of care, more evidence-based, more patient-centered, and we think technology's crucial for that.

Perhaps more encouraging, Mirza thinks tech can lead to a more informed patient, and a better doctor-patient relationship:

We struggle currently measuring things directly from patients, like what their pain level is, what their activity level is. This technology can make that easier. Even more, we want to be able to turn that information back to patients, feed it back to them in a way that makes it meaningful for them.

Mirza also sees the potential for group feedback and group pattern detection (e.g. determining that low blood pressure readings increase chances of infection during surgery). He is hopeful that working with SAP HANA will allow clinicians to look at large sets of patients, evaluating data in a 'meaningful, interactive way, close to real-time.' Ideally, even changing the patient's course of treatment prior to outcome, based on fresh insights.

Stanford: tackling date silos in pursuit of genome sequencing

Next up was Stanford genetics professor Carlos Bustamante, who is tackling the issue of data silos head-on. You can't tolerate data silos when your ambition is human genome sequencing, which requires large scale data integration and real-time analysis. Stanford and SAP are part of a 150 member organization, the Global Alliance of Genomics and Health, tasked with sharing innovative method of processing genome-type data, without the crushing limitations of silos.

At the time of our video with Bustamante, Stanford was moving from a one year pilot on SAP Hana to a new, three year contract to use Hana for real-time data analysis. During the pilot, the speed increases were notable. But Bustamante knows that speed gains will mean nothing if there isn't better global collaboration:

Bustamante brought up a need for globalizing genome data I hadn't considered:

For far too long, much of the genetic data that's been collected in the studies that have been done have focused almost exclusively on populations of European descent. We think that's problematic. We really need to go global. Figuring out how to create the right incentive structure, the networks of collaborators so that everybody can benefit from the genomics revolution is of upmost importance to me.

Bustamante concedes that we 'don't have a complete picture' of genome data yet, but he sees powerful implications:  'When you go to the doctor and you have your genome sequenced - which I bet you you will in the within your lifetime - we can really give you back meaningful information so that if you carry a mutation in a gene that may predispose you to cancer, we'll know that.'

MemorialCare: blowing doctors' minds with real-time dashboards

Dan Exley of MemorialCare Health System brought the issue of outdated data systems to the fore. As Exley explained, MemorialCare has been logging all kinds of medical data since it first began electronic regards in 2006. Yes, there is plenty of 'very detailed, very fine grained data', but can the analytics systems of the past handle it? Exley thinks not. Traditional analytics reporting just isn't enough to allow for real-time interventions:

Exley used an emergency room example to drive the point home:

A physician in the emergency department who is caring for 50 patients right now  - how could we get information to him? [What if] some new information has come to light, and it gets fed automatically into this formula and we present it back to that physician, and say really quickly: 'You should go and take intervention now - because an hour can make a big difference.'

Knee deep in a number of pilot projects, Exley points optimistically to a Hana use case where all drugs given to any of MemorialCare's hospitalized patients were loaded into Hana. That's 22 million drug administrations which can be sliced and diced and presented to physicians in a dashboard view.

This project has the doctors' attention: 'It blows their mind. You can even drill down to of all the drugs a particular doctor gave to that type of patient last week or yesterday in a matter of seconds. We've never been able to do that before, and it really is revolutionary.'  Exley adds that providing patient access to this information is on the project list.

Verdict

These use cases add field evidence to those who believe data innovation can transform health care. But it will not be easy:

  • During a video with SAP Chief Medical Officer David Delaney, he painted a picture of health care CIOs 'drowning in data,' stuck on complex analytics stacks supported by 1980s disk-based relational databases.
  • Perhaps more concerning, Delaney confirms that IT departments are still bogged down in resource-intensive maintenance chores: 'A lot of internal resources are focused and really focused, often 60% to 80% just keeping the lights on. It prevents them from being agile and really understanding how the business can run.'
  • Of course, SAP thinks it can help companies solve these issues via HANA-powered analytics, in-memory, and stack simplification.
  • For this piece, I was less concerned with the distinctions of SAP's technology, and more interested in how modern data initiatives can improve health care - without running roughshod on data privacy.

The early results in this video series seem promising, but we'll need to watch as stories unfold. Health care has a way of humbling the most powerful leaders - and software vendors.

Image credit: heart rate monitor in hospital © fivepointsix - Fotolia.com

Disclosure: SAP paid the bulk of our travel, shipping and accommodations at Sapphire Now Orlando and arranged these video shoots, which we filmed and produced on our own time. SAP is a diginomica partner as of this writing.