Big promise and big challenges for big health care data

The American College of Surgeons (ACS) has long valued the benefits of collecting and analyzing clinical outcomes data as a means of improving surgical care. In the early 20th century, for example, Ernest A. Codman, MD, FACS, proposed The End Result Idea—his premise that a hospital staff should follow every patient treated long enough to determine whether the treatment was successful.1,2 Physicians would then use any failures as learning opportunities to improve care and avoid similar situations or outcomes in the future. This pattern of collecting data and using it to drive improvements in the quality of surgical care has been repeated and refined for more than a century.

Dr. Codman and his peers could scarcely have imagined the explosion of data pertinent to quality improvement efforts available today or the insights derived from clinical data—information that will, hopefully, become widespread in the near future. Integration of data from clinical registries, electronic health records (EHRs), and other primary sources holds great potential as a means of improving the quality and efficiency of health care and lowering costs. This article explores the growing demand for health care data and introduces readers to the concept of “big data” and its implications for surgical practice.

Growth in HIT

In the 21st century, significant efforts have been made to incorporate modern computer technology, mobile devices, software, and other technological advances into the production and analysis of medical data. Examples of growth in health information technology (HIT) include both the move from paper to EHRs, as well as the growth in clinical data registries. More than 120 clinical data registries were in operation or in development by specialty organizations in the U.S. as of November 2014, according to a clinical registry inventory produced by the National Quality Registry Network (NQRN), a voluntary network of registry sponsors and other interested parties.3

Many surgeons and other physicians had anticipated that EHRs would better leverage clinical information within their daily workflow and support direct data feeds to clinical registries and, in turn, better guide clinical care. However, early products have largely failed to collect, analyze, and return useful information at the point of care. The College has tried to improve surgeons’ experience with EHRs by providing tools to help them choose the right EHR product, offering guidance to enable a better understanding of the EHR incentive program and empowering users to meet meaningful use requirements.4,5 But to fully attain the goal of leveraging clinical data at the point of care, surgeons must push for optimal use of the digital information in EHRs and registries by appropriating more information into the surgical team’s workflow.

Federal efforts to expand HIT

Early federal efforts at using data to reduce costs and drive improvements in outcomes focused largely on administrative claims data, as these were the data readily available to the government through the Centers for Medicare & Medicaid Services (CMS). Administrative claims data are routinely collected for payment purposes and are relatively easy to analyze. However, claims data do not capture the nuances of comorbidities, severity, conditions present on admission, complications, patient experience, or other socioeconomic factors critical to understanding health outcomes.

Unlike administrative data, which aggregate experience for system management requirements, clinical data are patient-specific and can be more precisely stratified to define best practices.6 In fact, administrative claims-based performance measures quickly proved inadequate to fully achieve the dual goals of improving health care outcomes and slowing growth in health care spending. In 2009, Congress passed the Health Information Technology for Economic and Clinical Health (HITECH) Act as part of the American Recovery and Reinvestment Act (commonly referred to as the stimulus package).7 The HITECH Act contained incentives for providers to adopt EHRs and laid out meaningful use requirements with the goal of ensuring that the federal funds were being spent wisely and in a way that would improve the provision of health care. Although the meaningful use requirements are far from perfect, the EHR incentive program has helped to expand the use of EHRs, increasing the amount of clinical data potentially available for analysis.

Many obstacles must be overcome on the path to meaningful use of digital information in health care records before patients and surgeons will feel the beneficial effects of digital clinical information. At first, federal lawmakers seemed to anticipate that simply digitizing the paper record would provide a return that would satisfy surgeons and other health care providers. However, many surgeons saw the rollout of EHRs largely as an additional administrative burden. Due to a limited information exchange, a lack of data standards and interoperability, and virtually no real-time clinical analytics, time spent entering data into EHRs may seem like a poor use of resources. For many clinicians, the EHR is simply an expensive means of recording data previously stored in a paper record, and extracting information from these digital files has proved to be an inefficient tool for meeting the needs of patients or surgeons.

Federal programs have since taken incremental steps to encourage the use of clinical data registries. This effort has been aimed at increasing the clinical value of data collected and reducing administrative burdens, but it has also put further pressure on EHR users to feed information back into registries. One such federal action was attached to the so-called “fiscal cliff bill” that prevented a government shutdown in January 2013 and delayed sustainable growth rate-related cuts in Medicare physician payments.8 This provision provides an opportunity for Medicare eligible professionals (EP) to simultaneously use existing high-quality clinical registries for quality improvement and for meeting Physician Quality Reporting System (PQRS) requirements. Beginning in 2014, EPs were also able to report to PQRS with the qualified clinical data registry (QCDR) reporting option. QCDRs offer more flexibility than other PQRS reporting options, allowing EPs to report on a variety of measure types. In addition, QCDRs must have the capacity to track outcomes, provide timely feedback reports, and risk-adjust when appropriate. All of these capabilities are intended to result in the reporting of measures that are more relevant, clinically appropriate, and actionable for surgeons than the measures currently available as reporting options through PQRS.In April of 2014, CMS approved the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) data registry, as a QCDR. This combined ACS-American Society for Metabolic and Bariatric Surgery registry was  one of approximately 30 to receive such a  designation. By becoming a QCDR, the MBSAQIP is able to develop its own quality measures—in effect, this will enable metabolic and bariatric surgeons to choose what is reported to CMS. The ACS maintains that the measures within the MBSAQIP QCDR are more relevant, clinically appropriate, and actionable when compared with traditional PQRS measures.

The steps that the federal government has taken have been well-intentioned and, in some instances, have provided opportunities to leverage high-quality clinical data for use in federal incentive programs, such as the QCDR. However, many programs such as meaningful use have been poorly coordinated and missed the big picture, moving the ball only a modest distance toward the goal of using robust clinical data to achieve higher quality health care and slowing growth in health care spending, while arguably adding to the administrative burden that surgeons typically face. However, when factored together, the multitude of primary sources of data paired with powerful modern data analysis techniques and hardware are greater than the sum of their parts and show great potential for improving care for patients while reducing the administrative burden on surgeons.

This is where big data comes in. There are many differing definitions of big data, but in its simplest terms, it involves the analysis of large amounts of complex data from both structured and unstructured sources to derive valuable insights. For health care, this could mean combining information from registries, EHRs, claims databases, and other sources to improve the provision of care to the public, and some organizations are doing just that today.

HIT in a big data environment

Private sector firms are working together to create standards and build systems that can harness the massive amounts of data collected on a daily basis and use it to promote real-time measures for leading and lagging indicators that improve care and patient safety.

For example, the Louisiana State University Health Care Services Division, Baton Rouge; Intermountain Healthcare, Salt Lake City, UT; and the Beth Israel Deaconess Medical Center, Boston, MA, have currently set up systems that are capable of pulling data from multiple EHR vendors, registries, payors, and other sources and using it to provide physicians with real-time tools to improve the quality of care provided and improve population health, allowing for a focus on preventative care. These tools can include leading indicator alerts that, for example, point out that a colonoscopy may be recommended for a given patient, or automatically notice when emergency department visits for flu spike and remind primary care providers to recommend flu vaccines to patients who are being seen for other reasons. Lagging indicators, such as outcomes or compliance with best practices, also benefit as additional information can be garnered from the wide variety of sources and fed back into registries to provide a more comprehensive view of performance.

These pioneering health care systems have had to overcome a number of barriers to be able to put their data to work. The data available are currently siloed and housed in a number of public and private systems. Additionally, data use agreements have had to be worked out with each system. Another challenge is that there is no single standard for health data and metadata to allow for easy analysis across multiple sources, so data must be processed and formatted into a standardized structure; this process is sometimes referred to as cleaning or scrubbing. And, of course, all patient data must be protected to ensure privacy and comply with federal laws and regulation. All of these actions are accomplished through the use of advances in technology that exploit service oriented architecture (SOA) and platform as a service (PaaS). SOA and PaaS can create a stack of informatics which sits across many primary data sources (including EHRs from multiple vendors) and supply the analytics needed to provide researchers, health care providers, and even patients with valuable, real-time information.

SOA and PaaS can be thought of as similar to the apps running on mobile devices, specifically smartphones. Data are output to the end user through application programing interfaces (APIs), commonly referred to as apps. The smartphones do not store the data. The apps on the phone use common data definitions, such as global positioning system (GPS), to create a host of applications from mapping your location to directions to the local grocery store or gas station. In this way, data are entered once and repurposed many times. In health care, the GPS function could be replaced by a set of apps that track hemoglobin A1c or surgical site infections, for example. Since this is all accomplished in the cloud through virtual analysis, the data can remain housed at its source, obviating the need for data aggregation. The figure below gives an overview of how such a system might be structured.

Big data platform structure

Big data platform structure

Barriers to expanding the use of big data in health care

Creation of a national library of data standards and a library of defined cohorts for conditions (such as definitions for diabetes, congestive heart failure, or asthma) would be a leap forward in enabling advanced health analytics. The NQRN is attempting to create data standards for registries to define aspects of care such as a deep vein thrombosis, surgical site infection, an acute myocardial infarction, or diabetes. In addition, several large health care systems have come together to form the Healthcare Services Platform Consortium (HSPC) with the goal of setting standards for platforms (a sort of “cloud,” such as an Android-like platform) that stack atop and reach across EHRs, using NQRN data standards, to pull data and repurpose them for use in Maintenance of Certification (MOC) programs, registries, and so on. Creation of a set of common standards, analogous to the one used by the banking industry to communicate seamlessly across a multitude of independent entities, would go a long way toward facilitating adoption of these new tools. Ultimately, given the right environment, a separate marketplace for clinical apps on health care platforms will develop in which applications can be made available to multiple end users.

Access to all pertinent sources of clinical data is essential to appropriating the full benefit from big health care data, and for this to happen, the primary data held in EHRs will need to be fully leveraged. Unfortunately, health systems that have created and implemented this type of big data platform are already seeing private EHR vendor attempts to govern and control the information flow by blocking access to clinical information for analytics so that they can control and monetize the use of data housed in their products. Such attempts by various private companies to block information could prevent physicians from continuing to use these powerful tools for improving patient care and quality, or, at the very least, greatly limit their value. If left unchecked, this trend will have a chilling effect on innovation and will slow progress toward interoperability.

The advantages of the interoperability provided by PaaS and the improvements to health care it promises merit further investment in this technology as well as regulatory support from the federal government. In an August 14, 2014, letter to Sen. Ron Wyden (D-OR) and Sen. Charles Grassley (R-IA), ACS Executive Director David B. Hoyt, MD, FACS, expressed the need for information held in EHRs, registries, and other primary sources to remain available or be made available in a secure, cloaked or de-identified, standard format for the purpose of data analysis. This letter was sent in response to a request for information on how to advance the availability and utility of health care data from then Senate Finance Committee Chairman Wyden and Finance Committee member Grassley.

In fact, the issue of information blocking is receiving growing attention in Washington. In July 2014, two physician members of the House Energy and Commerce Committee, Rep. Phil Gingrey (R-GA) and Rep. Bill Cassidy (R-LA) spoke out on the issue, going so far as to call out individual EHR vendors for the practice and suggesting that the lack of interoperability between EHRs that have been supported with billions of taxpayer dollars is equivalent to fraud.9 The Energy and Commerce Committee has undertaken an ambitious multi-year effort to ensure that federal laws and regulations keep pace with advances in medicine in many areas, including “unleashing the power of digital medicine.”10 The ACS has tried to support these efforts by providing information on multiple occasions.

More recently, the fiscal year 2015 spending bill enacted by Congress in December included report language urging the Office of the National Coordinator for Health Information Technology (ONC) to certify only EHRs that do not block health information exchange.11 The report language further called on the ONC to provide a detailed report to Congress on the problem of information blocking, including a comprehensive strategy on how to address the issue. Although such report language is nonbinding, it is seen as a strong statement on the intent of Congress. Federal agencies, which receive their funding from Congress, frequently use these reports as guideposts for developing rules and regulations. If that proves to be the case and ONC follows Congress’ instructions, the report and strategy recommendations could be released in the spring.

How surgeons can support big data ecosystems

ACS Fellows and members can take a number of College-supported steps to realize the full potential of a data ecosystem. Actions surgeons may take to support the data ecosystem include the following:

  • Surgeons must first understand and engage in the design of data ecosystems larger than the EHR “space.” This understanding ideally means surgeons will have a role in optimizing data and its multiple repurposing—both on the input and the output side of the ecosystem.
  • The data ecosystems require a library of data definitions and cohorts defined by leaders in specialty medicine and surgery so that all platforms and EHRs will have the necessary metadata needed for use in analytic tools.
  • Apps or data outflows for numerous purposes need to be defined such as to identify a specific moment in clinical care during a patient visit or for more broadly tracking a surgical population outcome, such as in cancer care or results from total joint replacements. Surgeons need to be involved in designing clinical alerts, outputs for MOC, EHR to registry feeds, and more. The data ecosystem must also fit the payor system for assuring outcomes, driving improvement, optimizing costs, and creating public transparency.
  • The governance of the data ecosystem must ensure that EHR vendors do not limit the availability of data needed by patients and clinicians to optimize care. Silos of information with data standards defined by EHRs would not create the broad data ecosystems needed for digital data to reach its full potential.

Big data in health care—if allowed to flourish in an environment conducive to the secure analysis and use of that data—holds great promise for improving outcomes and informing decisions at the point of care. Surgeons have the experience and expertise to create this environment and to use it to improve efficiency and outcomes. This could represent the next evolution of the College’s century-long cycle of continuous quality improvement, putting into the hands of individual surgeons the capability to meaningfully analyze amounts of data that would have been inconceivable to Dr. Codman and his peers. It is up to the College and today’s Fellows to continue to lead and innovate in this realm.


References

  1. American College of Surgeons. History of surgery: 100 year timeline. Available at: http://timeline.facs.org/1913.html. Accessed February 2, 2015.
  2. Donabedian A. The end results of health care: Ernest Codman’s contribution to quality assessment and beyond. Milbank Q. 1989;67(2):233-256.
  3. American Medical Association. An inventory of national clinical registries. November 2014. Available at: www.ama-assn.org/resources/doc/cqi/x-pub/nqrn-national-clinical-registry-inventory.pdf. (Available on members-only website.) Accessed February 2, 2015.
  4. American College of Surgeons. Guide for EHR vendor selection. Free online EHR comparison resource for members. Available at: www.facs.org/advocacy/regulatory/ehr/vendor-selection. Accessed January 30, 2015.
  5. American College of Surgeons. Electronic Health Records (EHR) Incentive Program. Available at: www.facs.org/advocacy/regulatory/ehr. Accessed January 30, 2015.
  6. Surgical Quality Alliance. Surgery & Public Reporting: Recommendations for Issuing Public Reports on Surgical Care. 2014. Available at: www.facs.org/~/media/files/advocacy/sqa/2014sqa_publicreportingdocument.ashx. Accessed February 2, 2015.
  7. Library of Congress. Public law 111-5. Available at: www.congress.gov/111/plaws/publ5/PLAW-111publ5.pdf. Accessed January 30, 2015.
  8. Library of Congress. Public law 112-240. Available at: www.congress.gov/112/plaws/publ240/PLAW-112publ240.pdf. Accessed January 30, 2015.
  9. Pittman D. Dems, GOP call for investigation of federal health IT problems. Politico. July 24, 2014. Available at: www.politico.com/story/2014/07/federal-health-it-problems-investigation-democrats-republicans-109351.html. Accessed February 16, 2015.
  10. U.S. Energy & Commerce Committee. House of Representatives. Energy and commerce cures. Available at: http://energycommerce.house.gov/cures. Accessed January 30, 2015.
  11. Library of Congress. Explanatory statement regarding the House amendment to the Senate amendment on H.R. 83. Available at: www.congress.gov/congressional-record/2014/12/11/house-section/article/H9307-1. Accessed February 17, 2015.

 

Tagged as: , , , ,

Contact

Bulletin of the American College of Surgeons
633 N. Saint Clair St.
Chicago, IL 60611

Archives

Download the Bulletin App

Apple Store
Get it on Google Play
Amazon store