he healthcare industry historically has generated large amounts of data, driven by record keeping, compliance & regulatory requirements, and patient care. While most data is stored in hard copy form, the current trend is toward rapid digitization of these large amounts of data. Driven by mandatory requirements and the potential to improve the quality of healthcare delivery meanwhile reducing the costs, these massive quantities of data (known as ‘big data’) hold the promise of supporting a wide range of medical and healthcare functions, including among others clinical decision support, disease surveillance, and population health management [2-5]. Reports say data from the U.S. healthcare system alone reached, in 2011, 150 exabytes. At this rate of growth, big data for U.S. healthcare will soon reach the zettabyte (1021 gigabytes) scale and, not long after, the yottabyte (1024 gigabytes). Kaiser Permanente, the California-based health network, which has more than 9 million members, is believed to have between 26.5 and 44 petabytes of potentially rich data from EHRs, including images and annotations.
By definition, big data in healthcare refers to electronic health data sets so large and complex that they are difficult (or impossible) to manage with traditional software and/ or hardware; nor can they be easily managed with traditional or common data management tools and methods . Big data in healthcare is overwhelming not only because of its volume but also because of the diversity of data types and the speed at which it must be managed .The totality of data related to patient healthcare and wellbeing make up “big data” in the healthcare industry. It includes clinical data from CPOE and clinical decision support systems (physician’s written notes and prescriptions, medical imaging, laboratory, pharmacy, insurance, and other administrative data); patient data in electronic patient records (EPRs); machine generated/sensor data, such as from monitoring vital signs; social media posts, including Twitter feeds (so-called tweets) , blogs ,
Big Data Analytics in Healthcare
Health data volume is expected to grow dramatically in the years ahead. In addition, healthcare reimbursement models are changing; meaningful use and pay for performance are emerging as critical new factors in today’s healthcare environment. Although profit is not and should not be a primary motivator, it is vitally important for healthcare organizations to acquire the available tools, infrastructure, and techniques to leverage big data effectively or else risk losing potentially millions of dollars in revenue and profits. What exactly is big data? A report delivered to the U.S. Congress in August 2012 defines big data as “large volumes of high velocity, complex, and variable data that require advanced techniques and technologies to enable the capture, storage, distribution, management and analysis of the information”. Big data encompasses such characteristics as variety, velocity and, with respect specifically to healthcare, veracity. Existing analytical techniques can be applied to the vast amount of existing (but currently unanalyzed) patient-related health and medical data to reach a deeper understanding of outcomes, which then can be applied at the point of care. Ideally, individual and population data would inform each physician and her patient during the decision-making process and help determine the most appropriate treatment option for that particular patient.
Advantages to Healthcare
By digitizing, combining and effectively using big data, healthcare organizations ranging from single-physician offices and multi-provider groups to large hospital networks and accountable care organizations stand to realize significant benefits. Potential benefits include detecting diseases at earlier stages when they can be treated more easily and effectively; managing specific individual and population health and detecting health care fraud more quickly and efficiently. Numerous questions can be addressed with big data analytics. Certain developments or outcomes may be predicted and/or estimated based on vast amounts of historical data, such as length of stay (LOS); patients who will choose elective surgery; patients who likely will not benefit from surgery; complications; patients at risk for medical complications; patients at risk for sepsis, MRSA, C. difficile, or other hospital-acquired illness; illness/disease progression; patients at risk for advancement in disease states; causal factors of illness/disease progression; and possible comorbid conditions (EMC Consulting). McKinsey estimates that big data analytics can enable more than $300 billion in savings per year in U.S. healthcare, two thirds of that through reductions of approximately 8% in national healthcare expenditures. Clinical operations and R & D are two of the largest areas for potential savings with $165 billion and $108 billion in waste respectively
The conceptual framework for a big data analytics project in healthcare is similar to that of a traditional health informatics or analytics project. The key difference lies in how processing is executed. In a regular health analytics project, the analysis can be performed with a business intelligence tool installed on a stand-alone system, such as a desktop or laptop. Because big data is by definition large, processing is broken down and executed across multiple nodes. The concept of distributed processing has existed for decades. What is relatively new is its use in analyzing very large data sets as healthcare providers start to tap into their large data repositories to gain insight for making better-informed health-related decisions. Furthermore, open source platforms such as Hadoop/MapReduce, available on the cloud, have encouraged the application of big data analytics in healthcare. While the algorithms and models are similar, the user interfaces of traditional analytics tools and those used for big data are entirely different; traditional health analytics tools have become very user friendly and transparent. Big data analytics tools, on the other hand, are extremely complex, programming intensive, and require the application of a variety of skills.
For the purpose of big data analytics, this data has to be pooled. In the second component the data is in a ‘raw’ state and needs to be processed or transformed, at which point several options are available. A service oriented architectural approach combined with web services (middleware) is one possibility . The data stays raw and services are used to call, retrieve and process the data. Another approach is data warehousing wherein data from various sources is aggregated and made ready for processing, although the data is not available in real-time.
While several different methodologies are being developed in this rapidly emerging discipline, here we outline one that is practical and hands-on. There are main stages of the methodology. In Step 1, the interdisciplinary big data analytics in healthcare team develops a ‘concept statement’. This is a first cut at establishing the need for such a project. The concept statement is followed by a description of the project’s significance. The healthcare organization will note that there are trade-offs in terms of alternative options, cost, scalability, etc. Once the concept statement is approved, the team can proceed to Step 2, the proposal development stage. Here, more details are filled in. Based on the concept statement, several questions are addressed: What problem is being addressed? Why is it important and interesting to the healthcare provider? What is the case for a ‘big data’ analytics approach? (Because the complexity and cost of big data analytics are significantly higher compared to traditional analytics approaches, it is important to justify their use). The project team also should provide background information on the problem domain as well as prior projects and research done in this domain. Next, in Step 3, the steps in the methodology are fleshed out and implemented. The concept statement is broken down into a series of propositions. (Note these are not rigorous as they would be in the case of statistical approaches. Rather, they are developed to help guide the big data analytics process). Simultaneously, the independent and dependent variables or indicators are identified. The data sources, as outlined in Figure 1, are also identified; the data is collected, described, and transformed in preparation for analytics. A very important step at this point is platform/tool evaluation and selection. There are several options available, as indicated previously, including AWS Hadoop, Cloudera, and IBM BigInsights. The next step is to apply the various big data analytics techniques to the data. This process differs from routine analytics only in that the techniques are scaled up to large data sets. Through a series of iterations and what-if analyses, insight is gained from the big data analytics. From the insight, informed decisions can be made. In Step 4, the models and their findings are tested and validated and presented to stakeholders for action. Implementation is a staged approach with feedback loops built in at each stage to minimize risk of failure.Methodology
The next section describes several reported big data analytics applications in healthcare. We draw on publicly available material from numerous sources, including vendor sites. In this emerging discipline, there is little independent research to cite. These examples are from secondary sources. Nevertheless, they are illustrative of the potential of big data analytics in healthcare.
Premier, the U.S. healthcare alliance network, has more than 2,700 members, hospitals and health systems, 90,000 non-acute facilities and 400,000 physicians and is reported to have data on approximately one in four patients discharged from hospitals. Naturally, the network has assembled a large database of clinical, financial, patient, and supply chain data, with which the network has generated comprehensive, and comparable clinical outcome measures, resource utilization reports and transaction level cost data. These outputs have informed decision-making and improved the healthcare processes at approximately 330 hospitals, saving an estimated 29,000 lives and reducing healthcare spending by nearly $7 billion . North York General Hospital, a 450-bed community teaching hospital in Toronto, Canada, reports using real-time analytics to improve patient outcomes and gain greater insight into the operations of healthcare delivery. North York is reported to have implemented a scalable real-time analytics application to provide multiple perspectives, including clinical, administrative, and financial . Another example, reported by IBM, is that of the large, unnamed healthcare provider that is analyzing data in the electronic medical record (EMR) system with the goal of reducing costs and improving patient care. (Data in the EMR include the unstructured data from physician notes, pathology reports and other sources). Big data analytics is used to develop care protocols and case pathways and to assist caregivers in performing customized queries
At minimum, a big data analytics platform in healthcare must support the key functions necessary for processing the data. The criteria for platform evaluation may include availability, continuity, ease of use, scalability, ability to manipulate at different levels of granularity, privacy and security enablement, and quality assurance In addition, while most platforms currently available are open source, the typical advantages and limitations of open source platforms apply. To succeed, big data analytics in healthcare needs to be packaged so it is menu driven, user-friendly and transparent. Real-time big data analytics is a key requirement in healthcare. The lag between data collection and processing has to be addressed. The dynamic availability of numerous analytics algorithms, models and methods in a pull-down type of menu is also necessary for large-scale adoption. The important managerial issues of ownership, governance and standards have to be considered. And woven through these issues are those of continuous data acquisition and data cleansing. Health care data is rarely standardized, often fragmented, or generated in legacy IT systems with incompatible formats. This great challenge needs to be addressed as well.
Big data analytics has the potential to transform the way healthcare providers use sophisticated technologies to gain insight from their clinical and other data repositories and make informed decisions. In the future we’ll see the rapid, widespread implementation and use of big data analytics across the healthcare organization and the healthcare industry. To that end, the several challenges highlighted above, must be addressed. As big data analytics becomes more mainstream, issues such as guaranteeing privacy, safeguarding security, establishing standards and governance, and continually improving the tools and technologies will garner attention. Big data analytics and applications in healthcare are at a nascent stage of development, but rapid advances in platforms and tools can accelerate their maturing process.
Raghupathi W: Data Mining in Health Care. In Healthcare Informatics: Improving Efficiency and Productivity. Edited by Kudyba S. Taylor & Francis; 2010:211–223.
Burghard C: Big Data and Analytics Key to Accountable Care Success. ID Health Insights; 2012.
Bian J, Topaloglu U, Yu F, Yu F: Towards Large-scale Twitter Mining for Drugrelated Adverse Events. Maui, Hawaii: SHB; 2012.
Raghupathi W, Raghupathi V: An Overview of Health Analytics. Working paper; 2013.
Zikopoulos PC, DeRoos D, Parasuraman K, Deutsch T, Corrigan D, Giles J: Harness the Power of Big Data. McGraw-Hill: The IBM Big Data Platform; 2013.
Zikopoulos PC, Eaton C, DeRoos D, Deutsch T, Lapis G: Understanding Big Data – Analytics for Enterprise Class Hadoop and Streaming Data.
McGraw-Hill: Aspen Institute; 2012. 32. Bollier D: The Promise and Peril of Big Data. Washington, DC: The Aspen Institute; 2010.