ConClusions PaPer

Informed decisions, improved outcomes

A summary of the Executive Leadership Panel Discussion at the 8th Annual SAS Health Care & Life Sciences Executive Conference
Panelists:

The Power of Health Analytics:

lanier M. Cansler, secretary of the north Carolina Department of Health and Human services Jan De Witte, President and Ceo of Healthcare iT and Performance solutions at Ge Healthcare H. shelton earp, MD, Director of the university of north Carolina lineberger Comprehensive Cancer Center
Moderator:

Tom Davenport, co-founder and research Director, international institute for analytics

The Power of healTh analyTics

Researchers from the UNC Lineberger Comprehensive Cancer Center found that young, African-American women in eastern North Carolina have a much higher incidence of one of the six types of breast cancer a startling finding that was later confirmed in other states and in west Africa. Knowing that ethnicity is a factor will improve screening and earlier detection. The same study found that older African-American women with estrogen receptorpositive cancer had a higher death rate than their white counterparts. Further analysis using data from the state tumor registry will show whether this disparity is due to biological differences or shortfalls in care. Are African-American women in certain areas of the state getting appropriate access to Tamoxifen or aromatase inhibitors? An expensive treatment called cetuximab is effective in treating colon cancer but only for some patients. Analytical data exploration has shown that the drug definitely does not work in about 35 percent of patients those who have a certain mutation. For those patients, the high-priced medication can be spared without compromising the quality of care. These examples just hint at the many ways data and analytics are ushering in a new age in health care. Genomic data, clinical trial data, electronic health records (eHrs), claims data and research study data terabytes and petabytes of it can be brought together to reveal important discoveries and support better decisions. For example: Analytics-based drug discovery processes identify better and safer therapies, optimal clinical trial designs and populations, and potential synergies or problems faster and earlier in the development cycle. Evidence-based medicine integrates clinical expertise and research results to support the best decisions about patient care. Protocol-based medicine draws on research results to identify best practices for specific conditions, medical histories and patient populations. Personalized medicine blends diverse data sources, including genetic profiles, with historical clinical data to lead to personalized diagnosis and treatment based on a patients specific biomarkers. in May 2011, sas brought together great minds from across the industry provider, pharma, payer and policy segments to discuss the promise of data and analytics in improving medical decisions, costs and outcomes, particularly as the industry is being reshaped in the us by the Patient Protection and affordable Care act.

The Power of healTh analyTics

More than 250 industry leaders from 26 states attended this executive summit to hear former sen. Tom Daschle and analytics guru Tom Davenport give keynote addresses, to ask questions of a panel of industry experts, to network with peers, and to attend special tracks targeted to their industry sectors. This paper summarizes the Executive leadership Panel Discussion headed up by Davenport, featuring perspectives from government, the research and provider communities, and health care iT. Participants included: Lanier M. Cansler, Secretary of the North Carolina Department of Health and Human services. Jan De Witte, President and CEO of Healthcare IT and Performance Solutions at Ge Healthcare. H. Shelton Earp, MD, Director of the UNC Lineberger Comprehensive Cancer Center.

The Paradoxical State of Health Care

The US spends more on health care than other industrialized countries, yet we have poorer health. We have technological marvels to diagnose and treat conditions, yet many business processes and systems are still on paper. We pour money into sophisticated treatments for acute conditions, while offering relatively little attention to the preventive and educational efforts that could reduce or eliminate some of those acute conditions. americans spend more on health care than any other country in the world about $2 trillion a year yet our infant mortality rates and our diabetes rates are higher, said Kecia serwin, Vice President and General Manager for Health and life sciences at sas, in her opening remarks. We rank lower on some of the safety and quality measures, and were considered overall less healthy than many of the more developed countries. The realities are discouraging. We have a cost problem. over the course of the next 10 years, this country will spend $35 trillion on health care, said Daschle in his keynote address. We now spend $8,500 for every man, woman and child in the country in taxes, premiums and out-of-pocket expenses 50 percent more than the second-most-expensive country. When i was born, health was 4 percent of the GDP. When my children were born, it had gone to 8 percent. When my grandchildren were born, it went to 16 percent, and they tell me if im lucky enough to have great-grandchildren, that figure will be 32 percent of GDP. Thats unsustainable.

The Power of healTh analyTics

according to our panelists, the cost problems have three key root causes: Waste. For every $100 spent, $20-plus is waste, said De Witte. The industry does not utilize its resources and assets very well. A typical hospital in the US will run at 60-70 percent of its capacity and still feel clogged up because the processes are not working. second, there is a lot of non-value-added clinical variation. and the third is the lack of coordination and cooperation across the system, leading to duplication of services and lack of preventative services in the system. Overuse of services. it has been estimated by rand and by the Congressional Budget office, as well as by other studies, that out of the $2.5 trillion we spend on health, upwards of $700 billion is unnecessary, said Daschle. unnecessary in part because of defensive medicine, doctors protecting themselves. unnecessary in part because of what we call proprietary medicine. in some cases, health care has become as much of a business as it is a practice, and utilization of equipment in clinics is driven by business decisions as much as by health decisions. Theres also market-driven care. People see things on television they like, they demand it of their provider, and guess what, they get it. Skewed allocation of resources. Health care looks like a pyramid in every country, said Daschle. at the base you have primary care and wellness at the very top you have the extraordinary array of technological applications unknown 20 or 30 years ago: Mris, heart transplants and so on. every [other] society starts at the base of their pyramid and works their way up until the money runs out. in the united states, we start at the top of the pyramid, and we work our way down until the money runs out. and the money runs out. We have an access problem. The last time we took up health care reform in 1993, we had 35 million uninsured, said Daschle. last year it reached a peak of 51 million people, and by 2020, if nothing changes, it will go to 64 million americans. of those who do have health insurance today, about 50 percent are underinsured. The situation is strained even for those who get government-paid health care. i currently help control about a $16 billion annual expenditure in north Carolina in Medicaid and mental health services, said Cansler. For the first time ever, in this budgetary cycle were going to actually end up with fewer dollars in our system, even though were going to have 50,000 or 60,000 more people in the Medicaid program. as a state, we face the same issues with the state employees health plan. The population is aging, and the cost of care keeps rising. Can the funding keep pace, and for how long?

Just this week, the Department

of Health and Human services reported that only 12 percent of those uninsured can afford to pay their hospital bill. That translates into about $73 billion of uncovered or uncompensated cost that becomes a burden on the system.
Kecia Serwin VP and General Manager for Health and Life Sciences at SAS

The Power of healTh analyTics

We have a quality problem. according to the Commonwealth Fund, the World Health Organization and many others who have analyzed our circumstances today, we dont match the outcomes of any other industrialized country, said Daschle. The inspector General for the Department of Health and Human services just last november said that 134,000 Medicare recipients a month experience adverse consequences as a result of their experiences in hospitals or clinics in america today; 15,000 people die each month because of medical mistakes. one out of seven people experiences a serious negative reaction due to their time exposed to our health care system today. so we have a quality problem. There is hope because there is bipartisan consensus about the goal, said Daschle. We want to build a high-performance, high-value health care marketplace with greater access, better quality and lower cost, period. The question is, how do we get from here to where we need to go, and how soon can we do it?

We know that we cannot sustain

the health care programs we have financially over the long haul. We are either going to have to cut back on the services we provide, or were going to have to better manage those services. We are convinced that the technology and the analytics and having the information are key to being able to do that over the long term.
Lanier M. Cansler Secretary of the North Carolina Department of Health and Human Services

Causes and Symptoms of an Unhealthy System

This is an industry that needs to find a new level of performance, and a combination of iT and analytics is going to be one of the key drivers to do that, said De Witte. im a very strong believer in some of the basic Deming or six sigma principles of measureanalyze-improve-control. Analytics are going to be there to go back to the root causes of poor performance in the industry, and analytics are going to be key in the control phase to bring the decision support to the places where the decisions are made. even as analytics becomes commonplace in business and government decision making including marketing, performance management, governance and compliance the high-stakes medical profession hasnt fully embraced the possibilities. our panelists discussed some of the obstacles to the data-driven, analytics-empowered ideal.

im a very strong believer in some

of the basic Deming or six sigma principles of measure-analyze-improve-control. analytics are going to be there to go back to the root causes of poor performance in the industry, and analytics are going to be key in the control phase to bring the decision support to the places where the decisions are made.
Jan De Witte President and CEO of Healthcare IT and Performance Solutions at GE Healthcare

The system itself lacks coordination.

if a system is defined as having a central administrative and decision-making authority, unlike all other industrial countries, we dont have one, said Daschle. We have a collage of subsystems about 50-50 public and private. Those collage subsystems have created a lot of stovepipes. The collaboration and coordination required to make these subsystems work in an efficient marketplace is partly what drives so much of the debate today.

The Power of healTh analyTics

There are disconnects even within a single organization. Traditional health care information systems typically reflect a process- and patient-oriented view of the business. The architecture uses a host of independent systems on different platforms, which share information in a limited way, if at all. as growth, mergers and acquisitions reshape information networks, it is common to see multiple, incompatible platforms even within a single functional area. From the private side, the industry will start to align on standards to make interoperability happen, said De Witte. if we want to create systems that bring value and that people want to pay money for, were going to have to make sure it works in the health care ecosystem. The [health care IT] industry is starting to realize that and will come together to drive it in that direction.

Sometimes key data is incompatible or unavailable. if were going to make real inroads,
When it comes to cancer treatment, for example, we really only know about how 3-5 percent of patients are treated, and those are the patients in clinical trials, said earp. Thats only a fraction of the patient population, and thats only a portion of their therapy. if were going to make real inroads, we need data and then analytics on everyone, throughout the continuum of care, about how we take care of this disease. even when clinical data is available in digital form, it is usually formatted for billing purposes rather than for analysis. Furthermore, there are several entirely different categories of data to deal with: electronic medical records that relate directly to patient care, aggregated data about organizational performance and resource utilization, statistically derived data for planning and decision support, and comparative data for research and outcome assessment. some of this data is episodic and patient-focused and some is cumulative over time, entities and populations. Data elements and coding strategies vary among specialties, such as MedDra for recording adverse events and iCD-9-CM for coding diagnoses for reimbursement. its a challenge to reconcile and cleanse all of this incompatible data, much less reap useful intelligence from it. we need data and then analytics on everyone, throughout the continuum of care, about how we take care of this disease.
H. Shelton Earp, MD Director of the UNC Lineberger Comprehensive Cancer Center

Even if the data is there, do you have permission to use it?

Patient privacy is another issue, a point that was raised by an audience member in the Q&a session: i recently started working with a regional health information exchange and was surprised to find out that the biggest challenge wasnt in aggregating the data from the various payers and providers in the area. it was actually in getting the members consent to use it.

The Power of healTh analyTics

The more participants in an information exchange, the more valuable the tool is for everybody, said Cansler, explaining why north Carolina chose an opt-out approach, rather than expecting members to opt in. notwithstanding all other aspects of the law, you can share information in the health information exchange unless the individual member opts out of the system. The hope is that people will appreciate the benefits of sharing health care data and choose to allow their information to be shared for productive benefit.

Organizations need to move beyond data to insight.

over the past decade or more, enormous amounts of money have been spent on bringing digitization to the industry, said De Witte. So today we are sitting on a lot of digitized data, but that doesnt mean you have information, insight or better decision making yet. The concept of meaningful use making reimbursement for health care iT contingent on meaningful use of data will surely drive greater adoption of analytics, our panelists agreed. That influence is already being seen in the academic world, noted earp. Because of meaningful use, our institution has seen more of a movement than ever in the last two years to bring the business hospital side and academic health services research side together and that has been very helpful to us. [Meaningful use] is the right impetus, De Witte added. after the industry has been pushed, incentivized to get basic data in digitized form, now its time to say, look, Phase Three: Youre actually going to use it to make better operational and clinical decisions.

A High-Performance Health Care Analytics Framework

We have a 21st century operating room, and a 19th century administrative room, said Daschle. only 15 percent of the administrative complexity within our health care marketplace is electronically driven today; 85 percent is still paper-driven. Paperwork today represents about 20 percent of all of our costs in health, and the source of many mistakes. However, dramatic changes are underway. The financial carrot-and-stick approach associated with meaningful use has spurred a dramatic rise in adoption of electronic medical record (eMr) systems in hospitals and physician offices, as well as the creation of health information exchanges for secure sharing of clinical data within and across organizations. EMR and health information exchanges provide the foundation for a smarter health care system.

The Power of healTh analyTics

if phase three is about putting all this data to work for better operational and clinical decisions, as De Witte noted, the remaining critical element is analytics. Providers will need analytic capabilities that can uncover useful insights from an ever-expanding universe of digital health care data.

IT infrastructure at the organizational level

an effective framework for health care analytics must include the following elements: Data management capabilities that work with the existing data infrastructure to integrate data from virtually any source, cleanse it and prepare it for analysis. Analytics of several flavors descriptive and predictive to derive meaningful insights from both traditional, structured data and unstructured data types, such as text, voice and video. Real-time analytics that support improved decision making right at the point of patient care. Customer intelligence solutions that identify the most effective engagement strategies, content and channels for different patients, based on broad knowledge of their needs and preferences. Risk-based analytics to assess todays risk and model the clinical and financial effects of planned future initiatives. Workflow management tools that deliver the necessary information to the right people, teams, roles and iT systems.

Information exchanges for cross-organizational data sharing

With a solid data management foundation in place, organizations across health care sectors and geographies can establish and/or extend valuable information exchanges. The north Carolina Health information exchange, for example, is enhancing the ability to share information for monitoring and research purposes, said Cansler. This exchange brings together data from multiple sources, such as Medicare and Medicaid claims and north Carolinas community care networks 14 networks around the state, representing approximately 95 percent of all primary care physicians. The north Carolina Health information exchange doesnt replace or compete with existing regional information exchanges, such as exchanges within the unC system, across Mecklenburg County, or spanning an 18-county region in the western part of the state. rather, it provides a forum for these exchanges to contribute and share data statewide.

The Power of healTh analyTics

Coupled with sas analytics, the north Carolina Health information exchange will make it possible to monitor the cost of care and the outcomes of chronic disease management and preventive efforts on a larger scale, uncovering insights that smaller or more narrowly focused data sources would not reveal.

We can throw money at prevention

The Promise of Analytics in Health Care

As health care organizations build data repositories and exchanges, analytics will play a bigger role in delivering care. For one, there is more data available. on the provider side, for instance, there is data pulled from operational systems that record patient admissions and discharges, bill patients and insurers, order tests and dispense medications. The data can be analyzed to evaluate alternative treatments by looking at length of stay and incidence of readmission. Data analysis can reveal which therapeutic regimens work best, uncover disease patterns, identify at-risk patients or assess the performance of individual physicians. All three panelists shared examples of how their organizations are capitalizing on analytics to improve both the business and practice of medicine.

all the time, but unless we know what were doing is actually working, its not going to get us where we need to go. The sharing of data that we will be able to do through the health information exchange is helping us identify where we need to focus the effort to be most successful, to try to keep health conditions from getting to the crisis stage where [we] spend the greatest amounts of our money.
Lanier M. Cansler Secretary of the North Carolina Department of Health and Human Services

Understand non-value-added clinical variance.

intermountain Healthcare and specifically Dr. Brent James, one of the handful of front-runners in clinical analytics has really proven that if you understand your clinical variance, particularly the non-value-adding clinical variance, you can take out 15 to 20 percent of cost while improving clinical outcomes, said De Witte. Ge is working with intermountain to create an iT platform capable of supporting that kind of study, as well as other pilots in the coming year.

Improve patient safety.

If you could analyze the circumstances around patient safety incidents falls, medication errors and such a provider organization could implement better procedures and safeguards. A new patient safety organization at GE is installing a user-friendly interface in hospitals where nursing staff can enter information about near misses, said De Witte. as we know from other industries, safety is all about understanding your near misses; that is where you learn. With sas, we take that information, run statistics, trending, root causing and benchmarking, and feed that back. Its another example of using analytics to get that measure-analyze-improvecontrol loop going. its one of the basic principles that should be applied a lot more in the health care industry.

The Power of healTh analyTics

Personalize medicine for the patients genotype.

it cost about $300 million to sequence the first human genome 3 billion Dna base pairs in every cell. now it can realistically be done for about $15,000 (and will soon drop to $8,000$10,000), leading to an exciting new era in personalized medicine. ive been in basic biomedical research for 35 years, and there has never been a time when the technology and our understanding of the genome and mutation and cancer and other diseases have been as ripe for clinical application, said earp. The information you can get out of this will trump a lot and will actually save money.

Investigate the effectiveness of serial treatment plans.

it is difficult to fully understand a patients course of treatment costs, care and outcomes because it is often serial. Patients may receive one form of care, and if they fail that, they go on to something else and something else, often in different settings. The unC lineberger Comprehensive Cancer Center has started an ambitious program to bring that serial care into a unified view. We are creating a 10,000-patient Cancer survivorship Cohort that will follow cancer patients every year, said earp. The program database will record patients germline Dna, bio-specimens from their tumors, questionnaire information about the discovery of their disease, their demographics and socioeconomic status, plus clinical information from pathology and more. Bringing together all that genomic data terabytes of it is going to be a great challenge, but with sas help, we really hope to do this, said earp. We believe that this kind of information understanding the sequence of care as parsed by the genome you inherit and the genome that is mutated in the tumor will really bring the big hits later on.

Monitor public health on a statewide level.

Within the arena of public health and monitoring, we have developed a statewide system called nC Detect which has become somewhat of a national model and nCB-Prepared, which constantly pulls data from emergency departments, eMs and other areas, said Cansler. if something is going on somewhere, we know about it quickly. We can analyze the data and know how to react much more quickly.

The Power of healTh analyTics

in the past, the hospital would send us a postcard and say, Well, weve had this flu case. We would find out about it two weeks later, and then wed start reacting. now we find out about it minutes later, and know across the state how things are happening, so we can be prepared to address broad public health or bioterrorism issues.

Influence patient behavior.

Does analytics have a role to play in shaping positive behaviors, such as a healthy diet, exercise and adherence to medication regimens? Yes, our panelists agreed. First, theres an educational component. analytic applications used for customer relationship management and marketing can also be used to model which patient communications are most effective for which types of patients. The retail marketing mantra right message to the right person at the right time is equally valid for patient education. if persuasion isnt enough, smart devices and packaging can help enforce compliance, Cansler noted. For example, Bluetooth-capable meters for asthmatics and diabetics can transmit meter readings through a cellphone connection. Bluetoothcapable pill packs can register when a pill has been removed from the pack. if a monitoring system detects that the patient hasnt taken a meter reading or a medication at the recommended intervals, he or she can get a reminder phone call or text. Have you forgotten to check your blood glucose level?

Ensure program integrity.

as we try to manage the dollars and make sure the access is there, the whole issue of program integrity is extremely important, said Cansler. Working with companies like sas, we have continued to build on our capabilities for monitoring identifying provider or consumer abuse of the system. We want to make sure that when were spending a dollar on health care, its really accomplishing the goal that we need to accomplish.

10

The Power of healTh analyTics

Who pays for health analytics?

Canslers remark leads to the inevitable elephant-in-the-room question. Who pays for all this analytic technology and innovation? not every community has an academic powerhouse such as a unC Health Care system or a heavyweight such as a Mayo Clinic. a lot of providers just dont have the resources to hire people to do this kind of analytical work they dont come cheaply, Davenport noted. Who should pay for all this data storage, data analysis and consulting services? insurance companies? Provider systems? Government?

Analytic intelligence ultimately pays for itself.

Many of the systems on the provider side have very good returns on investment, especially for workflow optimization systems and analytics, said De Witte. The industry is starting to realize that systems for operational optimization have good payback and are very worthwhile investing in. Providers will pay because they see the benefit. Where it gets more complex is on the clinical analytics, where the analytical effort and the ultimate benefit lie in different places. Complex, yes, but with the prospect of cost benefits, noted earp. For example, the ability to sequence the human genome will enable providers to forego a number of tests that are now being performed, and set up a care pattern that will enable providers to give certain expensive medications only to the patients who will respond. That has got to save money over time, said earp, and more importantly to me, it will allow this marvelous science to actually be used, rather than put on the shelf.

Information exchanges are prime candidates for cost sharing.

From the standpoint of the health care information exchange that were working on right now, and the data thats going to come from that, were looking at a process of sharing that cost, said Cansler. if were successful in building the system and we use the data wisely, everybodys going to benefit from it in one way or the other. if you have a system that creates value, the people are going to be willing to pay. insurers and payers will pay a part of it. Hospitals will pay a part of it. The physician community will pay a part of it. right now we have a great coalition of folks who are working hard from all these arenas to make this happen, to make it work, because they all see value in it. once you spread the cost over a larger population, it is not so high.

11

The Power of healTh analyTics

Key Takeaways
For all its Star Trek technological marvels CaT scans, Mris, robotic surgery and so on the practice of medicine still operates on a somewhat traditional model. Besides still being highly manual in record keeping, the practice tends to be reactive, focusing on the treatment of an existing condition an expensive proposition with suboptimal results. new data resources are improving the efficiency, cost, accuracy and outcomes of medical treatments, but the future holds even greater potential. Predictive analysis of diverse data, including genomic data, promises to revolutionize the essential model of health care delivery. our panelists closed the discussion by offering some hopeful prescriptions for positive change. Start somewhere, even if it isnt perfect. it doesnt require rocket science to start doing analytics, said De Witte. This industry today has so much low-hanging fruit that even if you dont have the perfect data warehouse yet, you can still start on that measure-analyze-improve-control loop. Its going to take some time before everybody is sitting on useful data, but the opportunity is so big that you dont want to wait until the perfect solution is there to get started. Find and eliminate waste. if you eliminate waste, you get better cost and better quality, said De Witte. its not a trade-off; you can have both. analytics will help you find some of the causes of waste, and analytics are going to help you sustain your process or decisions at the higher level. Use health care wisely, so it can be used. Part of me is really so excited about what we will be able to do with diseases like cancer on the basis of science, genomics, cell biology and technology, said earp. My greatest fear is that those advancements will be developed and not be used, because were not able to afford them. The only way were going to be able to afford them is to really segment target populations and understand how to use those kinds of things wisely so they benefit patients. Integrate data across the continuum of care. if were really going to do the job, weve got to treat the whole person and not pieces of the person, said Cansler. To make that happen, we have to have the ability to share that data.

new data resources are improving the efficiency, cost, accuracy and outcomes of medical treatments, but the future holds even greater potential. Predictive analysis of diverse data, including genomic data, promises to revolutionize the essential model of health care delivery.

12

The Power of healTh analyTics

Mental health data and physical health data; inpatient, outpatient and er; care given over time and in different locations it all must be brought together to create a 360-degree view. Thats the only way we are going to really manage the care and reduce the cost, said Cansler. The ability to share information, analyze that information, and make better decisions to avoid duplication or unnecessary utilization of services to make sure were using things that actually result in the outcome that were facing, and not just spending money trying to do things thats going to be the future for a while. The alternative is to reduce what we pay to providers, reduce the amount of services we have available, and minimize our system. And I dont think theres an American who wants to see that happen.

When we can look at not only the

patients current disease state, but all of the associated clinical data, demographic data, lifestyle behavior choices, genetic data when we can really start to get that total view and treat the health care of that person, not just the disease then we can really move the needle on improving medical outcomes.
Kecia Serwin VP and General Manager for Health and Life Sciences at SAS

13

SAS Institute Inc. World Headquarters

+1 919 677 8000

To contact your local SAS office, please visit: www.sas.com/offices

SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. indicates USA registration. Other brand and product names are trademarks of their respective companies. Copyright 2011, SAS Institute Inc. All rights reserved. 105340_S74728.0911