Big Data, skinny data, smart data, actionable data — as more healthcare activities are made electronic and structured, these are the terms being thrown around about the information being captured. The emphasis for most Health IT organizations and national endeavors is on "Big Data," that is the fact that more pieces of information are being coded and stored electronically on patients than ever before. Behavioral actions, diagnoses, procedures, medications, demographic information is all being captured by EMRs and other such programs. But the problem is what to do in this tidal wave of data? How can people find what is essential, what is useful, and what will help further and improve healthcare for patients? The answer may be found in an unlikely place — a diagram about the hype cycle of technological innovations.
The Gartner Hype Cycle is a way of looking at how new technologies are greeted, spurned, and eventually accepted by users. There's the need and invention that leads to the technology being announced and slowly implemented in select areas. Next follows the early anticipation, the belief (however untethered to reality it may be) of all the possibilities this new technology will bring to the industry adopting it. Next comes the "trough of disillusionment" — that moment when people realize that all of their dreams are hampered by realistic limitations and issues that will always stem from adopting a new technology. This is the rough time when people are deluged with possibilities but unfortunately it may not be what they expected. Those expectations are tempered though, and people then see what can actually be accomplished by this new technology. Lastly, once pragmatic constraints have been accepted while understanding the actual ways that this new tech can help, a new era of productivity can begin. This is seen all the time as a new iPhone launches or some new EMR comes out for doctors: promises of a complete game changer turn into soured realities, and eventually people figure out the best way of using this new piece of technology to get what they want out of it.
I'd argue that we are currently in the midst of the "trough of disillusionment" with Big Data. Sure there's a deluge of information now available to doctors that is growing by the day...but so what? If it's not being used or implemented in a way to make better the workflows of physicians, the lives of patients, or the tasks of healthcare, then how is it truly benefiting anyone? I've written in the past about the need to differentiate between Big Data and Actionable Data, and that while it's great to capture all of this information in a structured way, it's far better to do so with a plan in mind and in such a way that will have a meaningful impact on the lives of others. Others have taken up this call as more and more people are discarding Big Data for a smarter, more focused approach.
In the Harvard Business Review article titled "You May Not Need Big Data After All," the authors point out that outside bodies will have to form to determine what data is worth capturing and in what format. These could be facility-based groups, national organizations, or even professional associations like College of American Pathology or the American Surgical Association. By having these groups meet, discuss, and decide on what should be captured, physicians and their institutions can now have guidelines for what to capture in their reporting. This makes the data not just easier to deal with on scale, but more pertinent to offering up improved healthcare. As the article states:
"But over time, quality matters, so companies will want to initiate processes for improving data capture. Invariably, that means reviewing business processes and identifying where mistakes enter systems. People required to use data will take an active interest in governance processes designed to clarify data definitions and in learning how information flows through the organization."
Another article, published on Healthcare IT News, is entitled "Data analytics top concern, but industry stumped about where to start" and similarly outlines the problem of determining what should be captured. Big Data is not helpful as it's a deluge of information that still requires programs to sift through to find the pertinent data. The real key is Smart or Actionable Data, that is setting out to define what should be captured, by whom, and when, and then using that pre-determined data set to look closely at whatever problem (Cancer, Obesity, etc.) that researchers and clinicians are interested in.
"Organizations feel they need to jump on the big data bandwagon," said Shane Pilcher, vice president of Stoltenberg Consulting, in a press statement. "Yet they approach this emerging issue reactively versus proactively. Healthcare IT leaders should instead focus on collecting smart healthcare data, monitoring what data they're saving and concentrating on the quality, quantity and validity of data needed to answer future questions for organizations."
At mTuitive, we have some experience with this form of proactive data capture. We partnered with the College of American Pathologists to develop CAP eFRM, a reporting solution that is focused on capturing specific cancer checklists with fields determined by a governing body at CAP that can be later used. In fact, by using CAP eFRM, St. Joseph's Health - Northern California is now able to transmit that cancer data directly to the California Cancer Registry. That is smart data, that is actionable intel that is not part of some nebulous cloud of Big Data but instead a specific set of information that can be used for specific purposes in research, or in examining demographic changes in a disease, or the efficacy of a particular treatment. Simply casting a wide net and hoping to gather the information you need is an expensive and inefficient process, one that must be supplanted by more focused reporting techniques that are proactively designed by organizations that know the information they want to capture. And yes, it's possible that such information may change, or expand, which is why it's important that Health IT companies take an iterative and agile development approach to incorporating the changes and expanding the scope once it is necessary.
Furthermore, mTuitive is involved in the Canadian Partnership Against Cancer's synoptic reporting initiative for cancer surgery. Once again, groups determined what should be included in reports, what will be important to know and examine on a macro level, and what is possible for surgeons to capture at the time of surgery. These groups set out their needs for data capture and we met them and allowed them to compile data on a provincial scale, which is then used on a national scale. Once national scale gets involved, it sounds like Big Data again — but remember that this national level is comprised of smaller bits that are predetermined by physicians and proactively sought out. The parts end up informing the whole, rather than having a mass of information that is nigh impossible to decipher.
Imagine having access to every operative report done in the US or Canada. Contrast the task of gleaning useful, actionable knowledge from a predetermined, consistent set of structured data points on each procedure, with doing the same using every procedure's narrative dictation. Even IBM's Watson can't produce data that was never recorded.
There is a lot of data out there in healthcare. Some of it is useful, some of it isn't. And the treatments of tomorrow are based on the data being captured today. So it is up to Health IT and healthcare organizations to decide what it is they want captured, how to capture it, and how best to use it. Big Data may be a buzzword, but it is also a distraction from the true goal: better data to ensure better treatment.