Tag Archives: UMass
As I indicated in Competing on Analytics, if you ask CIOs today about the importance of data to their enterprises, they will likely tell you about their business’ need to “compete on analytics”, to deliver better business insights, and to drive faster business decision making. These have a high place on the business and CIO agendas, according to Thomas H. Davenport, because “at a time when firms in many industries offer similar products and use comparable technologies, business processes are among the last remaining points of differentiation.” For this reason, Davenport claims timely analytics enables companies to “wring every last drop of value from their processes”.
So is anyone showing the way on how to compete on analytics?
UMass Memorial Health Care is a great example of an enterprise that is using analytics to “wring every last drop of value from their processes”. However, before UMass could compete on data, it needed to create data that could be trusted by its leadership team.
Competing on analytics requires trustworthy data
At UMass, they found that they could not accurately measure the size of their patient care population. This is a critical metric for growing market share. Think about how hard it would be to operate any business without an accurate count of how many customers are being served. Lacking this information hindered UMass’ ability to make strategic market decisions and drive key business and clinical imperatives.
A key need at UMASS was to determine a number of critical success factors for its business. This included obviously the size of the patient population but it also included the composition of the patient population and the number of unique patients served by primary care physician providers across each of its business locations. Without this knowledge, UMASS found itself struggling to make effective decisions regarding its strategic direction, its clinical policies, and even its financial management. And all of these factors really matter in an era of healthcare reform.
Things proved particularly complex at UMass since they act as what is called a “complex integrated delivery network”. This means that portions of its business effectively operated under different business models. This, however, creates a data challenge in healthcare. Unlike other diversified enterprises, UMASS needs an operating model–“the necessary level of business process integration and standardization for delivering its services to customers”– that could support different elements of its business but be unified for integrative analysis. This matters because in UMass’ case, there is a single denominator, the patient. And to be clear, while each of UMASS’ organizations could depend on their data to meet their needs, UMASS lacked an integrative view into patients.
Departmental Data may be good for a department but not for the Enterprise
UMass had adequate data for each organization, such as delivering patient care or billing for a specific department or hospital, but it was inadequate for system wide measures. And aggregation and analytics, which needed to combine data across systems and organizations was stymied by data inconsistencies, incomplete population of fields, or other types of data quality problems between each system. These issues made it impossible to provide the analytics UMass’ senior managers needed. For example, UMass’ aggregated data contained duplicate patients—people who had been treated at different sites and had different medical record numbers, but who were in fact the same patients.
A key need for UMass creating the ability to compete on analytics was to measure and report on the number of primary care patients being treated across their entire healthcare system. UMass leadership saw this as a key planning and strategy metric because primary care patients today are the focus of investments in wellness and prevention programs, as well as a key source of specialty visits and inpatients. According to George Brenckle, Senior Vice President and CIO, they “had an urgent need for improved clinical and business intelligence across all our operations, we needed an integrated view of patient information, encounters, providers, and UMass Memorial locations to support improved decision making, advance the quality of patient care, and increase patient loyalty. To put the problem into perspective, we have more than 100 applications—some critical, some not so critical—and our ultimate ambition was to integrate all of these areas of business, leverage analytics, and drive clinical and operational excellence.”
The UMASS Solution
The UMass solved the above issues by creating an integrated way to view of patient information, encounters, providers, and UMass Memorial locations. This allowed UMass to compute the number of primary care physician patients cared for. In order to make this work, the solution merged data from the core hospital information applications and processed this data for quality issue that prevented UMass from deriving the primary care patient count. Armed with this, data integration helped UMass Memorial improve its clinical outcomes, grow its patient population, increase process efficiency, and ultimately maximize its return on data. As well UMASS gained a reliable measure of its primary care patient population, UMASS now was able to determine an accurate counts for unique patients served by its hospitals (3.2 million), active patients (i.e., those treated within the last three years—approximately 1.7 million), and unique providers (approximately 24,000).
According to Brenckle, data integration transformed their analytical capabilities and decision making. “We know who our primary care patients are and how many there are of them, whether the volume of patients is rising or decreasing, how many we are treating in an ambulatory or acute care setting, and what happens to those patients as they move through the healthcare system. We are able to examine which providers they saw and at which location. This data is vital to improving clinical outcomes, growing the patient population, and increasing efficiency.”
Thomas Davenport Book “Competing On Analytics”
Competing on Analytics
The Business Case for Better Data Connectivity
CIO explains the importance of Big Data to Healthcare
The CFO Viewpoint upon Data
What an enlightened healthcare CEO should tell their CIO?
If you ask a CIO today about the importance of data to their enterprises, they will likely tell you about the need to “compete on analytics” and to enable faster business decisions. At the same time, CIOs believe they “need to provide the intelligence to make better business decisions”. One CIO said it was in fact their personal goal to get the business to a new place faster, to enable them to derive new business insights, and to get to the gold at the end of the rainbow”.
Similarly, another CIO said that Big Data and Analytics were her highest priorities. “We have so much knowledge locked up in the data, it is just huge. We need the data cleaning and analytics to pull this knowledge out of data”. At the same time the CIOs that we talked to see their organizations as “entering an era of ubiquitous computing where users want all data on any device when they need it.”
Why does faster, better data really matters to the enterprise?
So why does it matter? Thomas H. Davenport says, “at a time when firms in many industries offer similar products and use comparable technologies, business processes are among the last remaining points of differentiation.” A CIO that we have talked to concurred in saying, “today, we need to move from “management by exception to management by observation”. Derick Abell amplified upon this idea when he said in his book Managing with Dual Strategies “for control to be effective, data must be timely and provided at intervals that allow effective intervention”.
Davenport explains why timely data matters in this way “analytics competitors wring every last drop of value from those processes”. Given this, “they know what products their customers want, but they also know what prices those customers will pay, how many items each will buy in a lifetime, and what triggers will make people buy more. Like other companies, they know compensation costs and turnover rates, but they can also calculate how much personnel contribute to or detract from the bottom line and how salary levels relate to individuals’ performance. Like other companies, they know when inventories are running low, but they can also predict problems with demand and supply chains, to achieve low rates of inventory and high rates of perfect orders”.
What then prevents businesses from competing on analytics?
Moving to what Davenport imagines requires not just a visualizing tool. It involves fixing what is allying IT’s systems. One CIO suggested this process can be thought of like an athlete building the muscles they need to compete. He said that businesses really need the same thing. In his eyes, data cleaning, data security, data governance, and master data management represent the muscles to compete effectively on analytics. Unless you do these things, you cannot truly compete on analytics. At UMASS Memorial Health, for example, they “had four independent patient registration systems supporting the operations of their health system, with each of these having its own means of identifying patients, assigning medical record numbers, and recording patient care and encounter information”. As a result, “UMass lacked an accurate, reliable, and trustworthy picture of how many unique patients were being treated by its health system. In order to fix things, UMASS needed to “resolve patient, provider and encounter data quality problems across 11 source systems to allow aggregation and analysis of data”. Prior to fixing its data management system, this meant that “UMass lacked a top-down, comprehensive view of clinical and financial performance across its extended healthcare enterprise”.
UMASS demonstrates how IT needs to fix their data management in order to improve their organization’s information intelligence and drive real and substantial business advantage. Fixing data management clearly involves delivering the good data that business users can safely use to make business decisions. It, also, involves ensuring that data created is protected. CFOs that we have talked to say Target was a watershed event for them—something that they expect will receive more and more auditing attention.
Once our data is good and safe, we need to connect current data sources and new data sources. And this needs to not take as long as it did in the past. The delivery of data needs to happen fast enough that business problems can be recognized as they occur and be solved before they become systemic. For this reason, users need to get access to data when and where they it is needed.
With data management fixed, data intelligence is needed so that business users can make sense out of things faster. Business users need to be able to search and find data. They need self-service so they can combine existing and new unstructured data sources to test data interrelationship hypothesis. This means the ability to assemble data from different sources at different times. Simply put this is all about data orchestration without having any preconceived process. And lastly, they need the intelligence to automatically sense and respond to changes as new data becomes collected.
Some parting thoughts
The next question may be whether competing upon data actual pay business dividends. Alvin Toffler says “Tiny insights can yield huge outputs”. In other words, the payoff can be huge. And those that do so will increasingly have the “right to win” against their competitors as you use information to wring every last drop of value from your business processes.
Solution Brief: The Intelligent Data Platform
I had the good fortune to work in the information services department at UMass Memorial Healthcare for several years prior to joining Informatica. It was pretty clear when I was there that the investments UMass Memorial was making in information systems was the future direction of healthcare everywhere, and that the lessons being learned there had applicability across the broader healthcare market. Since joining Informatica, I have had the opportunity to meet with a wide cross section of our healthcare customers and prospects, and I can confirm that this is in-fact absolutely true. A good case in point is the recent discussion I had with Karen Marhefka, Associate CIO at UMass Memorial, about the challenges of poor data quality and the adverse impact this can have on migrating existing data to new applications. (more…)