string(7) "English"

#PutData2Work

“The innovations will be the result of finally gathering our data. Data will be healthcare’s next great breakthrough.” — Leonard D’Avolio, CEO of Cyft (adapted from Healthcare IT News)

#PutData2Work was the hashtag at the 2017 HIMSS Big Data & Healthcare Analytics Forum in San Francisco, CA. I had an opportunity to present a session titled Surviving and Thriving in a Big Data World at the session on May 15, 2017. Here are some of the takeaways from the session and conversations with participants and practitioners thereafter.

Apathy is sticky

I saw the image to the right at the conference, and kind of liked it — I call it: “The do-nothing approach…”

So, let’s definitely NOT do nothing (double-negative intended). Also, let’s NOT keep doing what we’ve been doing in the past. Instead…

Let’s get STRATEGIC

When you distill the things that matter the most to healthcare organizations, you’ll typically find some or the other variant of the following 2 overarching goals:

  1. Better patient care & outcomes
  2. Better financial health & margins

Healthcare IT organizations need to become a strategic partner to the rest of the organization. Fullstop. To do so, the IT organization must focus their energies on helping with the top 2 or 3 overall priorities — each, in turn, supporting one or both of the aforementioned overarching goals, all while continuing to keep the lights on and provide a stable, performant operating environment for clinicians, staff, and patients. It’s easy to get swept up with the promise of a better tomorrow, but we need to, as an industry, focus on keeping the main thing the main thing, viz. let’s make the existing operational environment {better, cheaper, faster, stronger} while continuing to make gradual improvements towards a future-state. We need the IT organization to stop being a cost center, and instead help drive improvements to both the top- and bottom-line for the business.

We need to eliminate waste from all aspects of healthcare, including wasted dollars and resources on sub-optimal infrastructure, software, and processes. We need to be strategic: we need efficient; we need effortless; we need evergreen. We need the power of simplicity.

Data matters

In a previous blog, I talked about an inflection point in our industry.

With the big emphasis on cost remediation, waste elimination, and risk reduction, we find ourselves at an inflection point as an industry: we need to leverage all of the digital health data to generate value from our EHR investments.

In healthcare, all data matters — exhaust data (i.e., data that’s generated as a byproduct of working with other data) matters; unstructured data (i.e., data in clinical notes, flat-files, images, BLOBs, etc.) matters; structured data (i.e., data in EMRs, OLAP, and OLTP databases) matters; IoT data (sensors, wearables, trackers, etc.) matters. It all matters.

We’re at our infancy of harnessing value from some of the data, but in reality, there’s a lot of yet-to-be-mined value from all of the data. With over 80% of the data in healthcare being unstructured, it’s no wonder that Deep Learning — things like Natural Language Processing (NLP), Machine Learning (ML), and Artificial Intelligence (AI) — is getting a significant amount of dollars, press, (hype, too!), and attention these days. At the Big Data & Healthcare Analytics Forum, for example, we were hard pressed to escape a conversation or presentation (mine included!) without the mention of one or more of these sciences.

To truly be successful at harnessing all the data, we need to have 2 things: interoperability, and a data platform.

Data integration and interoperability is really, really hard. The simple fact is that there’s not a one-size-fits-all approach to doing this. Consider this: an average organization (here, I’m including organizations who have <10 beds, as well as mega-large organizations with multiple hospital groups) has over 350 applications running, each peripherally talking to each other, and each typically having their own database and data schema. It’s most definitely not easy to bring this data together. However, before we get too much into the doom-and-gloom glass-mostly-empty Eeyore approach, let’s keep in mind that there’s a lot of work that’s being done to make this data integration and interoperability better. We’re not going to delve into those bodies of work here, but I’m generally optimistic that we, as an industry, are taking steps in the right direction.

We’ll talk about the characteristics of a data platform later in this blog.

Big Slice, it’s what’s for dinner!

Instead of “Big Data”, I propose “Big Slice”. Let me explain by way of example.

Let’s look at customer service — what would your experience be if, before the customer service representative answered your call, she had a complete snapshot of all of your most recent interactions and transactions with this business, and she had relevant data blending (i.e., data from your most recent relevant social media interactions), so she could greet you with relevant-to-you, actionable knowledge? Wouldn’t this experience be bespoke, Jarvis-like (or for those who, like me, have a love for P. G. Wodehouse, Jeeves-like)?

The idea is simple — formulate analysis on just-in-time, relevant data, where the relevance is specific to a particular customer (patient), for a particular scenario (encounter, visit, diagnosis, etc.), and generate relevant insight from this data. To do this correctly, once again, you need interoperability, and a data platform.

Big Themes

We often talk about Precision Medicine, but, I propose that we should instead focus on Precision Health. Instead of trying to define this myself, let’s see what Dr. Lloyd Minor, scientist, surgeon, and the dean of Stanford Medicine, has to say about Precision Health:

If the amazing scientific advances of recent years can help us more effectively treat disease based on individual factors, shouldn’t we also put them to work by helping us keep people from getting sick in the first place?

The vision would be to go beyond Precision Medicine: instead of a frantic race to cure disease after the fact, we can increasingly focus on preventing disease before it strikes. By focusing on health and wellness, we can also have a meaningful impact in reducing healthcare costs. At Stanford, we call this idea Precision Health, where we focus on helping individuals thrive based on all the factors that are unique to their lives, from their genetics to their environment.

It is this vision and goal of actually trying to focus on the long game (to borrow a medical analogy — treating the disease instead of the symptoms) that will eventually get us to a state where we accomplish both top-level goals, viz. better patient care / outcomes and better financial health / margins.

Dr. Lloyd continues on to say:

Bringing the promise of Precision Health to patients will require a fundamental shift in our view of medicine, one which combines two seemingly different approaches – high tech and high touch.

It’s the high tech items that I’d like to focus on for the rest of this section.

The goal of Precision Medicine is to correctly identify the cause and process for a particular individual’s disease state, or epidemiology, and subsequently specifically treat the specific disease by creating and delivering a bespoke treatment targeted at the specific underlying cause and process.

To make this bespoke treatment happen, we need to leverage data: via -omics sciences, viz. genomics, phenomics, exposomics, radiomics, etc.; via unstructured data; via data blending (incorporating relevant information from non-traditional means, such as social media, psychological disposition, behavioral data, environmental data, etc.).

We need to bring deep learning — Machine Learning, Artificial Intelligence, and Analytics — to bear against this data so that we can unlock meaning and actionable insights from this data in the here and now.

Further, we need to learn about patterns and comorbidities and commonalities and heuristics across wider swaths of people so that we can continue to keep populations of people healthy. By the way, we now refer to this as Precision Public Health, and the term “Population Health” seems to be heading the way of the dodo. Why? Because we really need to apply the personalized nature of Precision Medicine to keep people healthier longer, and focus on prevention.

We need to be able to deal with a significant volume and variety of data, all for a single goal of building the bespoke treatment. We need big slice. We need a data platform.

Data Platform

I recently did an interview where I covered some thoughts around a data platform. To recap, nominally we need a few things:

  1. An agile, cloud-like, simple, cost-effective {i.e., better, cheaper, faster, stronger} operating and analytical environment, and
  2. (Micro) Services, visualization, near-real time interaction, and just-in-time actionable insight capabilities

In order to be able to do a lot more with a lot less, we need a robust Data Platform to be able to put data to work, and the Pure Storage Data Platform is perfectly poised to do just this. We need to switch from treating our systems like a library — i.e., a repository or archive of data — and instead start to drive meaning, action, and insight from the data.

Quite simply, Pure Storage helps put your data to work so you get the speed, insight, and agility you need to drive industry-defining innovations. Our unique data platform is built for the cloud era and powered by software and hardware that’s effortless to use, efficient from end-to-end, and evergreen to upgrade.

You can read more about our Data Platform here, and you can watch a video of my interview here.