+1,1,1
Search

Blog


The Truven Health Blog


The latest healthcare topics from a trusted, proven, and unbiased source.


Using Big Data to Improve Quality and Reduce Costs


By Michael L. Taylor/Wednesday, July 16, 2014
Mike Taylor imageA new report on potential uses of big data for controlling cost in the hospital setting has just been published. The report, from Brigham and Women’s Hospital in Boston, appeared in the July 2014 edition of Health Affairs. Six areas of potential benefit were discussed:

  • High-cost patients
  • Preventable readmissions
  • Triage upon hospital admission
  • Decompensation of clinical condition while in the hospital
  • Adverse events, particularly renal failure, infections, and adverse drug reactions
  • Treatment optimization for those with chronic disease involving multiple organs
As the authors point out, these are six key areas for intervention to lower healthcare costs in the hospital setting, and using more diverse data sources to analyze these opportunities will be useful.

As I reflect on this report, it strikes me that this type of report would have probably not been published several years ago. Healthcare reform, particularly changes in the payment methodology, is driving this type of research. I understand the need to minimize the healthcare spend and agree these are six key areas for research. But, in my opinion, the more important clinical issue is the improvement in the quality of care and probable saving of lives from better care. This is the real issue and opportunity.

All six of these areas are a result of missed opportunities to improve care. These areas are inter-related: high-cost patients are often a result of those who are readmitted multiple times for the same condition, suffer complications, are inappropriately triaged, and have missed diagnoses or have adverse events. Some of these problems can be prevented medically, but some of these issues have broader root causes. Take readmissions – many cases are due to socioeconomic factors such as inability to pay for medications, poor access to outpatient healthcare, or inability to pay for home care. Doctors and hospitals have historically not been paid to consider and manage these non-medical factors that lead to increased medical cost. While no physician wants complications to develop in their patients, hospitals and physicians have never before been penalized if this happened, so there has not been a focus on preventing these complications. New payment incentives are driving these changes and new approaches to care are developing. The promise of higher pay for better value in healthcare of populations, not for providing more services to individuals, is leading to new solutions in these six areas. “Big data,” meaning information about socioeconomic factors, living situations and other new data sources, and then using these data in predictive algorithms, will improve our ability to care for populations, not just treat individuals. 

At Truven Health Analytics, we use data to understand high-cost medical care. As we work with the payers of healthcare, especially large employers, part of our study is high-cost patients. I consistently find these cases to be complex, often involving advanced cancer cases or complicated heart failure cases. Closer oversight of these patients, team-based care, and better methods to predict and manage complications is warranted in many of these cases. Accountable Care Organizations (ACOs), with a patient-centered focus and a population health strategy, are promising new approaches to improving care. The tragedy of many of these cases however, is the missed opportunity to prevent these cases from ever occurring. If screening guidelines were followed more universally, advanced colon cancer would almost never happen. Heart failure is usually due to multiple heart attacks that could be prevented by paying closer attention to decreasing risk factors. Not all high-cost cases can be prevented, but many could be avoided.

Why, as a nation, are we not doing a better job in managing the health of our population? The most obvious answer is because we aren’t focusing on and prioritizing disease prevention among our population. Up to 70% of healthcare costs are due to preventable disease, but our healthcare system hasn’t been paid to focus on this issue. But change is apparent. The healthcare industry is undergoing more rapid change at this time than I’ve ever seen in my 30+ years of being a doctor. The new clear message is this: the way to manage costs is to improve the quality of care for entire populations, including new ways to prevent disease. Technology in the form of implementing integrated electronic health records, using more diverse data streams, re-designing healthcare delivery, and better predictive analytics are all tools to improve the quality of healthcare in the U.S. This is the right path to reduce costs.

Michael L. Taylor, MD, FACP
Chief Medical Office

Doctors and Data: Working Toward the Triple Aim


By Michael L. Taylor/Thursday, July 10, 2014
Mike Taylor imageChange is rapidly occurring in most aspects of the delivery of healthcare in this country. One of the most promising developments is the understanding that healthcare should strive to achieve the “Triple Aim” – better care for individuals, improved overall health of our communities, and lowered costs. The Triple Aim goals are about delivering better value in healthcare, not just delivering more care. The implications for our healthcare providers are enormous and may represent a fundamental change in the way care is delivered and paid. And the data needs are far greater than before – this represents a major challenge.

Many experts are advocating for new data steams to help find people at risk for diseases, even using non-traditional types of data, such as credit card purchases or use of social media, to define risk levels. Privacy advocates are adamantly opposed, and these debates will continue. Many employers have used medical claims data to understand population risk, but even using these data is worrisome to privacy advocates. Recent federal government revelations about NSA data probes into personal lives have generated much criticism, and I think the outcome will be more controls over the use of data. I think the “new data streams” will be narrowly defined. But the good news is new healthcare delivery models are finding ways to effectively use data to improve patient care.

In the new models of delivery, as seen in the patient centered medical home concept (PCMH), a healthcare team, captained by the physician, now has the responsibility to care for a defined population, not just the patients who show up for an appointment. Physicians are financially incented to provide better care. This drives the need for data, and health records need to find “gaps in care,” such as overdue cancer screening exams and missing lab tests. A PCMH team member is empowered to reach out to patients to help them get the care that is needed. The team is responsible for (and incented to provide) the healthcare needed in all phases of a person’s life. This requires integrated data from all settings – all outpatient encounters, hospital data, and follow-up care, including rehabilitation and nursing home and hospice care. Integrating all these data together will have tremendous potential to improve care. As an HIE contractor, we have constructed platforms that are delivering this kind of integrated data, so we know it’s possible today, and we’re working with hospitals toward the same end. Data integration will be necessary in order to understand when high value care is being delivered by hospitals, physicians, and all healthcare providers.

But more than finding gaps in care; the new model incents better care. Take a simple example of diabetes: the medical evidence shows lower mortality and morbidity in those who achieve blood pressure, lipid, and glucose control compared to those who are not well-controlled. New payment methods will pay physicians at a higher rate when their patients achieve better control of their diabetes. In this scenario, payment is more complicated, and now lab data must be analyzed to determine payment.

Paying more for better value has promise, but also many challenges. Defining better care for diabetes can be done, but what metrics should be used in other conditions?  Physicians see literally hundreds of different conditions in the course of their work with patients; how should higher value be defined in other medical and surgical conditions? Is there value is ordering appropriate radiology exams and forgoing inappropriate tests? How can that be measured and compensated?

Medicare policy is driving much of the change in payment mechanisms, but large employers are also asking about value. Employers are tired of paying for medical treatments that don’t work or are unnecessary, and are looking for cooperative relationships with providers to incent better care. Hospitals are adjusting to focusing on providing better care, not more care. The transition is turbulent, but the result has the potential of achieving the Triple Aim. We will not achieve these results in a fee-for-service system. The changes we’ve seen in healthcare over the last decade are the start of real reform that is badly needed, and we need to continue driving change toward a higher value system. Innovative use of new data streams is vital to this effort.

Michael L. Taylor, MD, FACP
Chief Medical Officer

Using Data to Improve Healthcare


By Michael L. Taylor/Thursday, July 3, 2014
Mike Taylor imageAs other have pointed out repeatedly, our healthcare system is badly broken. In fact, we don’t have a healthcare system in this country – it’s a series of independent businesses, often competing with each other in the goal of making more profit. The three constituencies in the healthcare business are the customers (patients), the providers (doctors and hospitals), and the payers (health plans, employers and the government). These three groups all have perfectly misaligned incentives. Patients want care at minimal cost, providers make more money by providing more care (whether it is needed or not), and payers want to minimize payments. The payment mechanism drives more care at higher cost, and the result is the U.S. pays 18% of its GDP for healthcare – more than twice as much as any other country on the planet.

How does smarter use of data help this picture? In my opinion, more intelligent use of data is an important part of the answer. Data is a powerful tool to help physicians make better decisions. In the hospital setting, physicians should have access to ALL of a patient’s medical record, not just information gathered during a single hospital stay. In most Emergency Departments, doctors often don’t have unfettered access to outpatient medical records that may provide important clues to making correct diagnoses. Tests are needlessly repeated, incorrect medications are given and diagnostic errors are made all too often.  Electronic medical records (EMRs) should be helping this problem, but unfortunately most EMRs are simply digitized versions of the old paper record. We need EMRs to be longitudinal electronic health records, aggregating all of a person’s health information into a single record to be used by all providers of care. A unified health record then needs analytic tools to be able to use the comprehensive record to improve care, provide guidelines for evidence-based medical care, prevent incorrect medication use, stop dosing errors, and have prompts in the analytic tool to stop repeat tests and x-rays- in sum, improve the care.

A unified, single, health record for a patient would be a great tool to help improve care, but in the U.S., we have more fundamental problems than a lack of accessible data. In today’s residency training programs, physicians should be taught how to use the data and EMRs to make better decisions. An evaluation of a patient should always start with the physician sitting with the patient, taking a probing history by knowing what questions to ask, and how to elicit symptoms. This information is supplemented by knowing how to properly examine a patient and understand how to put all the information together to formulate a diagnosis. We cannot rely on an EMR or CT scans to do this job – it must start with a thorough history and a proper physical. One of the most impactful lessons I was taught in residency was that if I finished taking a patient’s medical history and yet still didn’t have a series of probable diagnoses to consider, I needed to take more history. Unfortunately, in today’s hospitals, finding a diagnosis is all too often done by ordering more testing, and in a fee-for-service payment environment, more testing means more revenue. More procedures mean more revenue. Hospitals and physicians should be paid for providing a higher level of quality, not by volume. 

I am a strong advocate of using medical data and providing better analytic tools to help physicians and patients, but tools are just tools. Physicians and other caregivers need these tools to improve care, but providers of care also need to listen to patients, think critically in making diagnostic assessments, care passionately about improving care, and use sound judgment at all times. They cannot be effective in a fee-for-service world. Providers do need to improve the care they provide, but the U.S. needs a sound healthcare strategy to solve our issues. Technology is part of that solution.

Michael L. Taylor, MD, FACP
Chief Medical Officer

Five Things Employers Want from Health Plan Reporting


By Jennifer Huyck/Thursday, June 19, 2014
Jennifer Huyck imageThese days, health plans are under pressure to deliver more comprehensive and reliable information to their employer clients.

After all, population health is on everyone’s radar, and employers are trying to keep a tight rein on rising costs. Plus, with all the talk of healthcare Big Data, employers have higher expectations of the kinds of information health plans can provide. Information transparency and combining financial and clinical data from multiple sources are becoming critical.

In other words, traditional reporting isn’t going to cut it anymore.

But what, specifically, do employers want from health plan reports?

Based on our partnerships with over 150 of the nation’s largest employers — including 25 percent of Fortune 500® companies — Truven Health experts have compiled the following list of the five most important things employers want when it comes to the health plan reporting.
  1. Acknowledge their different needs. Step away from one-size-fits-all reporting. Each employer client will want to see different slices of data and varying levels of analysis to fit their specific business questions. Reports need to be flexible enough to meet those diverse requests and stakeholders.
  2. Help them educate and inform their senior management team. Benefits managers need to be able to prove to the Powers That Be that the company’s investments in employee health are worth it, and health plan reporting is an important part of that.
  3. Provide consistent, accurate, and timely reporting. Employers want data that they can trust, and they want it quickly.
  4. Show them how to compare themselves to the outside world. Reporting solutions should allow employer clients to compare costs and other points of interest to national and regional benchmarks, so they can identify areas for improvement and recognize successes.
  5. Be consultative and creative. This is perhaps the most notable change in what employers need today versus what they needed in the past. Today, it’s not just about the numbers on a spreadsheet. Employers need those numbers to be meaningful and useful as they try to solve new challenges. And it’s now the health plan’s job to offer guidance along with the numbers.
In short, plans that can provide data and analytics that are flexible and trustworthy, and that answer the “So what?” and the “Now what?” will be the best-positioned to become problem-solving partners that employers can’t live without.

For more details about these five employer reporting needs, download our latest insights brief.

Jennifer Huyck
Vice President, Analytics and Consulting

Oncology Treatments and Care Benefit from Big Data


By Kathleen Foley/Monday, April 7, 2014
Kathleen Foley imageAn article in HealthCare IT News on March 21 discusses the latest use of IBM’s Watson computer for tackling the synthesis of complex information required to personalize cancer treatments to individual patients.   Drawing upon the medical literature, drug databases and patient genomic data, Watson will identify possible treatments for specific patients – tailored to their own genetic mutations. The application of Watson’s brain power to cull through an enormous amount of information is truly a step forward in the world of fighting cancer. The human brain can only synthesize small amounts of data at any given time, so having a computer help with the sifting and sorting is of tremendous value.

And so far, the reaction appears appropriately modest. Watson may be able to detect and identify, but Watson can’t interpret and place treatments into context the way doctors can. Watson is an aide that will hopefully free up human time for the things that human’s do best, such as interpret, understand, recommend, listen to and take into account patient emotions and family needs.

At Truven Health, our approach to big data is much the same. We use technology to simplify, organize and identify patterns of care, drivers of cost, or patient sub-groups. We draw upon many components of big data, from medical claims to hospital discharges, work productivity and oncology EMRs as well as the literature, to identify patient-level value in cancer treatments. And like our medical counterparts, we leave the heavy thinking, the place where intuition drives solutions and identifies new paths forward, to our researchers. Good technology in the hands of humans striving to treat and cure cancer is good for everyone!

Kathleen Foley
Senior Director, Strategic Consulting (Life Sciences)

RSS