Analytics: The Key to the Future
Analytics:
The Key to the Future
Analytics: The Key to the Future
H
ospitals and healthcare organizations of all sizes and structures are quickly elevating analytics to the enterprise level. But according to a recent HealthLeaders Intelligence survey, some might still be lacking the expertise to perform predictive or prescriptive analytics.The July 2014 survey polled 147 senior, clinical, operations, finance, marketing, and information leaders across the healthcare spectrum. The majority of respondents were from nonprofit organizations (73%), while the remainder (27%) came
from for-profit settings.
Despite any potential infrastructure differ- ences, the general consensus of sur- veyed health systems suggests they will face similar challenges in order to thrive in an analytics-driven future. An over- whelming majority (89%) agreed that a transformational understanding of clinical
and fiscal success requires the ability to analyze clinical and cost data from all settings of care, including acute, primary, and homecare.
Doug Shaw, vice president and general manager at McKesson, says he isn’t surprised by the responses of leaders since healthcare analytics is fast becoming a performance imperative.
“The tools and capabilities and information assets available to providers are certainly more mature and
far more plentiful than they were only a few years ago,” Shaw says. “It’s a more straightforward proposition for providers who are looking to engage in healthcare analytics as opposed to dabble in it or run screaming from it.”
Healthcare has lagged behind many other industries in using data to measure and manage performance, but evidence
Importance of Analyzing Clinical and Cost Data | Please indicate the extent you agree or disagree with the following statement: To have a transformational understanding of the clinical and fiscal success of my enterprise, I need to be able to analyze clinical and cost data from all settings of care, including acute, primary care, home care, etc.
Disagree completely Disagree Neither agree Agree Agree completely
or disagree
1%
Base = 147 DOUG SHAW Vice President and
General Manager McKesson
5% 5%
35%
54%
Analytics: The Key to the Future
suggests this is rapidly changing. Additionally, the necessity of functional analytics is also pushing C-level leaders into hands-on involvement across the board.
“C-level individuals recognize that if they have an enterprise-level view of not only all that’s going on in their organization, but also global standards by which they can measure what’s going on within a functional area, they would have a more unbiased, objective understanding,” Shaw says.
In response to a question about exploring new data correlations and predictive analytics, survey responses were mixed. Approximately 41% said they were confident their organization had the expertise needed to explore new data correlations and perform predictive or prescriptive analytics, whereas 37% disagreed with the statement that they had the necessary expertise.
“There’s a certain amount of diligence that the
enterprise needs,” Shaw says, “when sourcing for that information, whether it’s through a formal RFP or in
informal conversations with vendors. The more thought- ful you are about what you need to sustain your analyt- ics initiatives, the more helpful you will be to any partner that you ultimately choose to help you get there.”
Healthcare analytics can be intimidating, so one of the associated challenges may be simply finding a place to start. According to Tina Foster, vice president of business advisor services at McKesson, it’s critical that leaders not treat analytics as simply another project.
“The more thoughtful you are about what you need to sus- tain your analytics initiatives, the more helpful you will be to any partner that you ultimately choose to help you get there.”
Exploring New Data Correlations and Doing Predictive Analysis | Please indicate the extent you agree or disagree with the following statement: Despite labor shortages in the area of data science and analysis, I am confident my organization has the experts we need to explore new data correlations and inferences and to do predictive and/or prescriptive analysis.
Disagree completely Disagree Neither agree Agree Agree completely
or disagree
7%
Base = 147
30%
22%
35%
6%
Analytics: The Key to the Future
“It’s not software that you buy and you get a go-live date,” Foster says. “It’s a program, and you have to learn that it’s a journey, it will live and breathe and change, but it will never end. Your go-live is really just the starting point, not the end point.”
Shaw goes one step further. “Just buying software does not solve your analytics needs,” he says. “You have to decide what
information sets you’re going to need, what internal governance you’re going to need, and what external support you’re going to need to bridge the gap between what your analytic competencies are and what they need to be.”
It’s clear that larger medical centers understand the com- plexity of the type of analyses that will help them in the coming year, but only 14% of survey respondents said they had the in-house capability for examining new data
correlations and inferences to do predictive analysis. Shaw finds this response rather disconcerting. “One in eight say they can do the heavy analytic lifting needed to forecast patient needs and project where their popu- lations are heading,” he says. “If that’s the case, what is the true readiness level of com- munity hospitals or smaller rural hospitals out there to prepare for future demands?”
In fact, Shaw says the industry has reached a point where software capabilities are outstripping organizations’
abilities to wield them. Even as tools grow more intuitive to use, he notes, in-house analytic competence is erod- ing as hospital budgets tighten and hiring costs for data scientists and analysts increase. This makes it all the more important for organizations to secure the right expertise to make the most of investments in analytic infrastructure (i.e., software and data resources), even if that means augmenting with or outsourcing to third parties.
An enterprise-level, cross-functional team that will have oversight across all analysis needs
Each department/function/service line will conduct its own data
The finance team The clinical leadership The IT team Outsourced consultants
51%
Base = 147
Best Model for Healthcare Analytics | As the need for transformational analysis becomes more critical to ongoing success, which model best describes how your organization will assign responsibility for healthcare analytics?
14%
12%
10%
7%
5%
TINA FOSTER Vice President of business advisor services
McKesson
Analytics: The Key to the Future
imperative that data be aggregated and governed at the enterprise level: to ensure that the entire team is using the same ‘source of truth.’ ”
Foster says an outside partner is often best equipped to navigate any cultural differences and provide refer- ences to best practices that have worked in other areas and settings.
Creating an Analytics Game Plan
From data sources to measures engines to visualization tools, the array of information resources available can be bewildering to hospital leaders and management teams. However, if an organization wants to build an infrastructure that will scale and provide quality information with organizational depth, Shaw says there are a few critical components to identify.
First, you need a platform or system that can manage all of the different data sources that might one day be incorporated. Second, “you need robust content—not just templates but real analytic guide paths,” Shaw says. He describes this as the difference between a screwdriver and a power drill. The third critical compo- nent for success is experience. “You should recognize that no matter how fluent and deep your internal people are in terms of analytics, there’s more that can be done that they can’t yet do,” Shaw says.
You also need an organization that can augment your own. “Whoever is delivering that to you, assuming you’re not building it in-house, needs to have been doing this long enough and often enough that they’ve seen your situation before and they know how to deal with it, even though you may never have worked with them before,” Shaw says. “If you’re going to change di- rection with your IT department, you need an organiza- tion that’s going to be able to support that interim need, and then you need an organization that can help you with change management, that helps your organization adopt and adapt to data-driven behaviors.” n
In another survey question, just over half of respon- dents said an enterprise-level, cross-functional team approach model best describes how their organization will assign responsibility for healthcare analysis. Another 14% identified that each department would conduct its own analysis on its own data going forward.
Foster says that for hospitals and organizations of any size, it’s imperative for analytics and targeted goals to align. “We have clients think about how data is a strategic asset in your organization and every single decision needs to be data driven,” she says. “Then, as you’re collecting data and sending it out in reports to the team, it must be aligned to your strategic goals and have some action associated with it for maximum impact.”
While healthcare has a long history with reporting, a paradigm shift must occur to one of taking action based on quality data analysis. “You have to make sure that, for the things you’re doing, you have a target and corresponding actions for when you’re above or below that target,” Shaw says.
However, Foster cautions that since there is often a history of territorialism within hospitals, it is important to build collaborative teamwork and create an enterprise- level analytics culture. “When an organization begins to get into [analytics], people tend to be territorial about their data and their numbers,” she says. “There’s usually a long history of people ‘data shopping,’ or using data to support their decisions rather than making decisions based on the data. That’s why it’s
“You should recognize that no matter how fluent and deep your internal people are in terms of analytics, there’s more that can be done that they can’t yet do.”