Data analytics: Are U.S. hospitals late to the party?

1
Tipping point: Hospital resilience in a perfect storm
Data analytics: Are U.S. hospitals late to the party?
Written by
Using analytics to mine large datasets for insights, commonly known as Big Data,
is already transforming industries ranging from consumer goods to transportation.
Certainly, the healthcare sector has the raw information to join this group. For example,
Kaiser Permanente, a California-based health network, has an estimated 27 to 44
million gigabytes of potentially useful patient information. Expectations are that the U.S.
healthcare sector will soon have a zettabyte (1021
gigabytes) of these data.1
1 Wullianallur Raghupathi and Viju Raghupathi, “Big data analytics in healthcare: promise and potential,” Health Information Science and Systems, 2014.
TIPPING POINT:
HOSPITAL RESILIENCE IN
A PERFECT STORM

2
U.S. hospital leaders see exciting potential in this treasure trove.
According to an Economist Intelligence Unit (EIU) survey of more than
300 healthcare sector executives conducted for this research program,
63% of respondents say the value of the data that hospitals hold remains
largely untapped. In addition, 60% say that Big Data is transforming
future opportunities around the patient and 57% believe it can be a
game changer for operational performance. Several years ago, McKinsey,
a consultancy, estimated that the U.S. healthcare sector could reduce
total costs by 8% through the creative application of analytics.1
Cognitive
computing systems—which involve machines learning for themselves
rather than relying on complex algorithms to analyze data—are even
seeking to reshape how diagnoses and treatment plans are made by
“thinking” for clinicians.
1 McKinsey Global Institute, “Big data: The next frontier for innovation, competition, and productivity,” 2011.
Technology and data perspectives
% of respondents
Agree
The value of the data hospitals hold on patients and outcomes
remains largely untapped
63%
Hospitals have lagged other sectors in adopting technology and
data analytics
61%
Big Data is transforming future opportunities around patient
management and outcomes
60%
In order to get the full potential value out of Big Data in the future,
pharmaceutical companies, payors and healthcare providers
will need to collaborate on data analytics to determine the most
effective and economical patient treatments and protocols
60%
Hospital data can be a significant game changer in operational
performance
58%

3
Slow adopters of technology
However, healthcare has long had a reputation for slow adoption of
technology. Already in the mid-1990s, the imminent advances of
electronic health records (EHRs) were being touted. Yet seven years after
the HITECH Act of 2009 signalled a regulatory push for the adoption
of these tools—later given further impetus by the Affordable Care Act
(ACA)—a substantial minority are still working on the basics. Twenty-two
percent of survey respondents report that in the next three years they
will be investing to complete implementation of EHRs and 29% will put
money in software or hardware to demonstrate their meaningful use.
The picture for analytics may be similar. Dr. Kip Webb, North American
Healthcare Provider Portfolio Lead at Accenture, says that, compared
with other industries, “hospitals are already late to the party.” Survey
respondents admit as much: 61% say that hospitals have lagged other
sectors on data analytics against only 7% who disagree.
An uptick in digitization?
Some important groundwork has been laid, however. The last seven years
have seen “a huge uptick in digitization,” notes Dr. Webb, with hospitals
actively gathering electronic data, largely as part of EHR implementation.
According to the U.S. Department of Health, as of April 1995, 98% of all
hospitals had demonstrated meaningful use or at least implemented some
sort of EHR.2
As Lloyd Dean, CEO of Dignity Health, puts it, “I could
overwhelm you with data. That is not the issue.”
Yet these mountains of data bring their own unexpected challenges.
Dr. Amy Abernethy, Chief Medical Officer and SVP, Oncology at Flatiron
Health and a board member of athenahealth, two EHR and healthcare
analytics firms, says, “There used to be magical thinking around
EHRs. [The expectation was that] we would have a rise in immediately
actionable data. Then we would push a magic button so that the data
would be organized perfectly and tell us what to do. What the sector has
learned, though, is that data in its raw form is not all that useful. It needs
to be cleaned up and organized. Then analyses need to be done.”
Gathering and bringing some kind of order to the data is an ongoing
process at many hospitals. Too often, notes Dr. Webb, even collection
is overly costly because “we have made clinicians very expensive data
clerks” instead of finding more efficient ways to enter information into
systems.
Data standardization and silos, even within individual hospital systems,
also remain a widespread problem.3
James Rebitzer, Professor of Markets,
2 “Hospitals Participating in the CMS EHR Incentive Programs,” HealthIT.gov Dashboard, April 2015.
3 “Survey: Unstandardized data biggest barrier to analytics, business intelligence for healthcare,” Becker’s Health
IT and CIO Review, 1 May 2015.
“I could overwhelm you
with data; that is not
the issue.”
Lloyd Dean,
CEO, Dignity Health

4
Public Policy and Law at Boston University Business School, says that
often EHR systems “are big, cumbersome beasts that are not that helpful
yet for the data analytics challenge.” At the sectoral level, aggregation of
information—a key to unlocking Big Data for most providers—has barely
begun. “With a few exceptions,” says Charles “Chip” Kahn III, President
and CEO of the Federation of American Hospitals, “we do not have
interoperability. You need to think of each hospital or hospital system as
separate outposts, each with its own data.”
Lots of data, little insight
Despite these challenges, an increasing number of large aggregations of
potentially exploitable healthcare data now exist. That number will grow.
Rather than the ultimately solvable issues of data organization, the most
pressing concern is more fundamental—what to do with the information.
“My worry,” says Mr. Dean, “is that we are going to be overwhelmed
with data, have it all sitting there and not have the construct to make it
actionable.” Dr. Webb agrees: “We have a lot of data now, but not a lot of
insight.”
Dr. Abernethy believes that solutions to this problem have begun to
emerge. After years of focusing on the data gathering and organization,
she says, “we are now at the point where there are examples of how it can
be made actionable.” The market is certainly putting substantial money
into the capacity to provide such insight. In the first half of 2015, $6.4bn
went into venture capital funding for healthcare analytics start-ups, not
far shy of the $7.4bn invested in all of 2012.4
At the other end of the
business spectrum, in 2014 IBM put aside $1bn to create a business
unit for Watson, its cognitive computing system. One of IBM’s main target
markets is healthcare.
4 CB Insights, “Healthcare Startup Boom: 2015 Could See More Than $12B Invested Into VC-Backed
Companies,” September 2015, https://www.cbinsights.com/blog/healthcare-startup-funding-trends/
How hospitals plan to use data to produce actionable insights*
(% of respondents)
Long-term individual patient
disease prevention and care
Finding and demonstrating value
added from processes and procedures
Monitoring performance of
individual medical personnel
Improving care pathways
and processes
Improving administrative and
other non-medical processes
Individual patient diagnosis
and associated care plan
0 10 20 30 40 50
*Timeframe is within the next three years Source: The Economist Intelligence Unit survey, 2016
In the first half of 2015,
$6.4 billion went into
venture capital funding
for healthcare analytics
start-ups, not far shy of
the $7.4 billion invested
in all of 2012.

5
Only a minority of U.S. hospitals, however, are currently coming to the
party. According to the EIU survey, just 37% of respondents say their
hospital plans to invest in data analytics in the next three years. Moreover,
although every hospital expects to try to make use of analyics in some way
in that time period, only a minority plan to exploit the technology for any
given task.
The difficulty does not appear to be lack of vision. Dr. Abernethy reports
that she sees “a rapid growth in interest” in a wide range of analytics-
based tools covering fields from general population health to oncology.
Just as the existence of EHRs revealed issues around the need to organize
and analyze data, though, now the existence of software that can provide
actionable insights “has uncovered new problems,” she adds. These are
likely slowing investment in analytics by hospitals.
Realigning reimbursement
One is realigning reimbursement. Dr. Abernethy notes that, to date, use
of data in healthcare has not been rewarded financially—and determining
how to do so is far from straightforward. The Centers for Medicare &
Medicaid Services (CMS) is currently launching a pilot program to use
such incentives in cancer care, as are other insurers. Dr. Abernethy says,
“The oncology community is trying to figure out how to use the data in
order to be paid.”
A bigger, more fundamental problem is with hospitals themselves. The
existence of an actionable insight, says Dr. Abernethy, “doesn’t mean you
have the resources, bandwidth or team members that know what to do
with it. We are all going to have to figure out how to change practice in
order to continuously integrate learnings from data.”
Looking ahead, she warns that we are asking doctors to act in completely
new ways: “By 2020, the hope is that a doctor can take your personal
attributes—age, co-morbidity, biomarkers—and combine that with what
we know about management at scale to figure out exactly how to take care
of you in a completely customized, instantaneous fashion. We don’t have
technology solutions, data, health systems and doctors coordinated and
ready to do that kind of work. If we bring technology along so quickly that
it is our expectation it will be used, without figuring out how to align the
pieces, we may overwhelm things.”
A handful are starting to make the transition. Mercy Health in Cincinnati
has taken advantage of Explorsys—an analytics based system based
on EHRs from multiple American hospitals to aid with diabetes
management—to help improve comprehensive management scores for
the condition by 11% between 2013 and 2015.5
At a more general
level, Geisinger Health Systems is using analytics to determine optimal
5 “Hospitals use big-data platform to improve care,” Modern Healthcare, 22 August 2015, http://www.
modernhealthcare.com/article/20150822/MAGAZINE/308229976
“We are now at the
point where there are
examples of how it
[data] can be made
actionable.”
Dr. Amy Abernethy, Chief Medical Officer,
Flatiron Health.

6
treatment choices in a wide variety of areas and to find the best places to
intervene at a population level.
The vast majority, though, have not yet begun this part of the change
toward using analytics well. Their difficulties should not be surprising.
Professor Rebitzer notes that “a big general-purpose technology, like
analytics, often takes decades of innovation before we figure out what it
is good for.” What sets healthcare apart from other fields, he adds, is that
it has lagged in investing in analytics. If hospitals are not going to be late
for another revolution, they will need to work through the barriers as they
use data to reinvent care.
Prudential, the Prudential logo, the Rock symbol and Bring Your Challenges are service marks of
Prudential Financial, Inc. and its related entities, registered in many jurisdictions worldwide.
0288363-00001-00 RSFL1662
This article was written by The Economist Intelligence Unit and sponsored by Prudential.
For more information call Prudential Retirement at 1-800-353-2847 or visit
www.Healthcare.PrudentialRetirement.com.

Data analytics: Are U.S. hospitals late to the party?

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Data analytics: Are U.S. hospitals late to the party? (20)

More from The Economist Media Businesses (20)

Recently uploaded (20)

Data analytics: Are U.S. hospitals late to the party?