The Big Promise of Big Data in Health Care case study

Attachment preview

CASE STUDY

The Big Promise of Big Data in Health Care

Healthcare spending in the United States is closing in on $4 trillion per year, with that number projected to grow at a rate of 6 percent annually. Cries to find a solution to the crisis of rising healthcare costs—while also improving quality—can be heard from across the country. The federal government, in response, has taken steps to try to bring the country closer to high-quality, afford- able health care. For example, Medicare, the largest health insurer in the country, has begun to penalize hospitals for failing to reduce hospital readmission rates, decrease the occurrence of hospital-acquired diseases and conditions, and effectively implement electronic health records (EHR). EHR systems track medical appointments, test results, health provider notes, communications, and other electronic data. The financial penalties imposed by Medicare could reduce a hospital’s Medicare revenue by as much as 6 percent.

The government has also instituted incentives to encourage the use of technology and data to decrease cost and improve healthcare outcomes. The American Recovery and Reinvestment Act of 2009 allocated $40 billion in incentive payments to healthcare providers to encourage them to implement EHR systems. The goal is to move EHR adoption, which stood at a lackluster 30 percent in 2005, to 70 to 90 percent by 2019. This goal is important: EHR systems have the potential to improve efficiency; improve patient access to their medical records; allow healthcare providers and patients to communicate more easily; increase transparency; reduce medical errors; and provide healthcare providers access to an ever-increasing amount of data about patients, medication, diagnosis, and treatments.

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.

EHR systems are just one example of the ways in which the federal government and many healthcare industry leaders are looking to information technology for ways to curtail healthcare costs while increasing quality. The hope is that with insights gleaned from mining enormous pools of healthcare data, the industry will find ways to reduce cost and identify which treatment plans are most effective. Yet, electronic health data has grown faster than insurance companies, medical research labs, hospitals, or care providers have been able to consume it. This largely unstructured data is collected from a range of sources, including lab and imaging reports, physician medical correspondence, insurance claims, and output from existing financial and customer service IT sys- tems. IT giants, IT start-ups, and healthcare providers have now begun working together to develop new technologies that make use of this data to evaluate the effectiveness of treatment plans for many diseases.

In 2012, for example, IBM partnered with the Memorial Sloan Kettering Cancer Center (MSKCC) to transform IBM’s cognitive computing technology, called Watson, into an oncolo- gist’s assistant that could diagnose and recommend treatment for cancer patients. Oncology treatment is dramatically more complex today than it was even a decade ago. IBM supplied Watson with two million pages of medical research papers. MSKCC provided 1.5 million patient records and the expertise of its oncologists. Together, they created a system that uses a patient’s medical information to synthesize the best treatment plan and display the evidence used to create the plan.

Optum, the data analytics division of health insurance giant UnitedHealth, offers a wide array of EHR and healthcare data analytics products. Optum One, for example, identifies gaps in care along with strategies to avoid patient hospitalization. Optum also offers EHR solutions that incorporate clinical performance evaluations for emergency departments and intensive care units.

In 2013, UnitedHealth launched an innovative initiative when it teamed up with the Mayo Clinic to establish Optum Labs. The new research center combined UnitedHealth’s claim’s data from 100 million patients over 20 years with Mayo’s five million clinical records covering 15 years and began mining the data for insights on how to improve healthcare. UnitedHealth also bought Humedica, a leading data analytics firm, to bring it into the project. Promising to make their research findings public, share their analytical tools, and work collaboratively, Optum Labs issued a call for partners to bring in more data. Drug companies such as Pfizer and Merck, major univer- sities, the American Association of Retired People (AARP), and many others quickly joined the project, giving the center access to vast resources.

Optum Labs now has dozens of initiatives on topics ranging from knee surgery replacement to hepatitis C to diabetes, and they are getting good results. Consider the example of metformin, the medication doctors overwhelmingly prescribe to patients when they are first diagnosed with type 2 diabetes. An Optum Labs study using data from over 37,000 patients found that sulfonylurea drugs have an equivalent effect on glucose control, quality of life, and longevity. Moreover, sulfo- nylurea drugs cost less, and patients who use this medication were able to wait longer before starting to take insulin.

Before using patient data, Optum Labs first de-identifies it, as required by HIPAA. Any links between the data set and the identity of the contributor are cut to safeguard the privacy of the contributor. Optum Labs also carefully controls data access, including preventing researchers from pulling data of an individual patient.

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.

Business Intelligence and Big Data

279

280

Still, some critics are concerned that healthcare big data analytics is becoming just another for-profit sector. They see big data as a valuable resource that large healthcare companies like UnitedHealth are now vying to control, and they argue that holders of critical data, such as clinical pathology laboratories, should consider carefully before providing access to their data to such a big company in the healthcare industry. Others, however, argue that it is vital that the healthcare industry nurture a culture of collaboration—for the betterment of all. And certainly many prominent organizations have flocked to join Optum Labs’ collaborative initiatives.

Unless considerable progress is made in increasing the efficiency of medical costs and treat- ments, health care will account for one-fifth of all spending in the United States by 2023. Big data experiments, such as those conducted by Optum Labs, if successful, may be able to do much to help reduce costs and improve health care.

What steps should Optum Labs take to ensure that its research is widely disseminated?