The concept of “Big Data” conjures up all sorts of Sturm und Drang in the healthcare information technology world. Roll out a fleetingly fashionable moniker to describe a nebulous category to attract the attention-deficit-addled social media set, mix in the notion of analytics to give it purpose and sobriety, and you have a strategic recipe to tackle the tactics of accountable care in a reform-minded industry.
Clearly, Big Data evokes great hope and passion among the healthcare IT set. In fact, Health Management Technology readers exude mixed emotions about the premise and promise of Big Data, with opinions sashaying between money pit and money pool.
Big Data keeps getting bigger, particularly in the healthcare industry, where the adoption of electronic solutions and increased connectivity are driving higher levels of information capture than ever before. Storing data, however, holds little value unless it can be accessed, analyzed and put into action.
The quandary of today’s healthcare environment is that there is a massive volume of data, yet few viable automated processes to extract meaning from data that is diverse, complex and often unstructured. Advanced analytics provide organizations with powerful tools to assess current performance metrics and model “what if” scenarios to make data-driven decisions that impact financial stability and care quality.
According to a recent study, If U.S. healthcare were to use Big Data creatively and effectively to drive efficiency and quality, the sector could create more than $300 billion in value every year. Two-thirds of that would be in the form of reducing U.S. healthcare expenditures by about 8 percent.
Healthcare organizations that leverage the power of analytics have the potential to gain deeper insight into the factors that impact clinical quality and costs. They can also better understand the root causes of undesirable clinical and financial outcomes, and drive improvements in processes and practices to elevate care quality while avoiding unnecessary costs. These are the components necessary for the industry to focus on rewarding high-quality care, instead of paying for the quantity of care delivered.
Electronic data availability and processing power have sparked a rapid evolution in clinical analytics. Clinical data has greater inherent complexity than financial or billing data, and traditionally, large data-set analysis has been performed using billing codes, which are ubiquitous and easy to use; however, they do not reflect clinical detail. While not all clinical information can be accurately reduced to structured and codified data, ever-growing amounts of clinical data are available in discrete electronic formats. As a result, we now have the capability to analyze enormous volumes of electronic clinical data, so statistical validity is not a barrier and analysis can provide insights not available with traditional medical research methods.
The results of such analysis can be applied in real time to patient care. Readmission predictors are a good example of this analytic evolution. Over the years, many readmission models have been described with varying degrees of validity and usefulness. Unfortunately, many have relied on post-discharge billing information and therefore were not applicable at a patient’s hospital arrival or during the hospital stay. But today, new statistically valid measures of readmission risk that are based on electronic data are emerging that can be used at the time of admission. These measures provide the clinician with subsequent readmission probabilities to guide current interventions. Point-of-care intelligence on other risks will be available soon as well as data-driven best practice guidelines to improve care for high-risk individuals and populations. Today’s exciting advances will pave the way for mind-boggling possibilities tomorrow.
Think of all the information collected every day by providers, payers, pharmas and researchers. Bring all that data together in intelligent, actionable ways, and you have incredibly powerful insight for managing population health, improving outcomes and reducing costs.
To fully appreciate the potential of analytics – both for healthcare organizations and for the patients in their care – consider the possibilities of adding genomic data into the mix. Organizations are already looking at how genomic profiles impact metabolism of medications. Imagine what this could mean to the healthcare industry and our delivery of care.
Being able to predict at the genetic level how a specific dose will affect a patient’s metabolism of medication could be crucial in preventing an emergent situation. Similarly, think of all the conditions that present similar risks, such as any infectious diseases, and how genomic profiling could help clinicians fine-tune therapies and medications to deliver safer, faster results and better care.
That’s the impact of Big Data. Through analytics, providers have powerful tools for spotting potential health risks across their patient populations. Combined with payers’ ability to bring in financial data, you’ll see the true impact of how providers and payers can work together to create value that neither could achieve individually.
We expect Big Data to get even bigger in healthcare analytics because so many areas of healthcare have yet to be scrutinized regarding more efficient performance, and efficiency will be a major factor in reigning in the spiraling cost of healthcare.
More data is being created every day in an effort to improve operational efficiency in hospitals. As a simple example, most hospitals can’t tell you the average time it takes for their nurses to receive an IV pump after making the request. That’s a small item with huge implications not only for running a hospital efficiently, but even more importantly, for delivering timely care to their patients. Without knowing the lag time, how can you improve it to make sure a patient’s condition doesn’t deteriorate?
Operational efficiency demands not only the constant analysis of currently measured tasks, but the discovery of new measurements which can help to improve daily operations.
Simply reporting the implications of existing data isn’t enough when you’re searching for the best ways to improve performance. You need to be constantly watching for new opportunities to improve, and this means creating additional data streams based upon the processes you are reviewing.
Complicating matters is the fact that the marketplace now demands this information be collected and made consumable in real time. That requires a business analytics engine which is integrated with a system capable of moment-by-moment data collection. In fact, it is this very integration which permits the creation of data that never existed before. In the near future, data creation will be the key to predicting a typical day at the hospital before it happens.
According to recent statistics from the Office of the National Coordinator (ONC), in just over a year [electronic health record] adoption has doubled, leading to a significant increase in the volume of healthcare data. With increased use of smart medical equipment, implantable devices, patient portals, mobile health, etc., the amount of data entering the healthcare landscape will only continue to grow.
Managing this influx of data is one challenge; however, healthcare leadership also needs to be able to analyze and predict risk, assess opportunities to improve outcomes, reduce unnecessary cost and present actionable information to those engaged at the myriad channels and modalities of care.
Like in other market sectors, those that use data to improve their quality and price competitiveness will not only grow their market share, they will also likely be on the good side of mergers; especially as systems consolidate to improve efficiency. This will become even more critical as healthcare systems develop clinically integrated networks (CINs), expand their direct-to-employer efforts and offer risk-bearing plans to the market.
In order to leverage data in a meaningful way, healthcare organizations must also implement a culture of data. This will require support and buy-in from frontline physicians, as well as those involved in the coordination of care.
Finally, those that lead healthcare innovation will not only master descriptive, comparative and prescriptive analytics, they will also develop prescriptive approaches that turn data into directives and ultimately into action that prevents unnecessary utilization, improves outcomes and drives high value to patients, payers and employers.
As the amount of data available for analysis and the power and sophistication to conduct such analysis continues to increase at an exponential rate, we sit on the precipice of great advancement in healthcare analytics designed to glean previously inaccessible intelligence that will lead to improved population health management and community-based care, health and wellness. We must, however, remain vigilant and think deeply about the motivation and impact of the inferences that are inherent in all analytical exercises. It is not only natural but also an integral part of the exercise to draw conclusions from the analysis of data.
As relatively new consumers of and participants in such analysis, we are obligated to understand the difference, for example, between correlation and causation. As we all have biases, we also have a responsibility to avoid drawing specific conclusions that may align with our world-view but are not yet clearly identified as truly useful signals amidst all the noise. Acting on intelligence gleaned from analysis rings a bell that can’t be un-rung. As our level of excitement and designs on future innovation increase alongside the expanding mountains of data, we must never forget that the true source and usefulness of analysis always rests in human hands bound by moral, ethical and legal constraints that govern the quality of care those that place trust in the healthcare industry expect and deserve. In short, advancement in healthcare analytics is not just about technology. It’s about real, powerful thinking.
As healthcare institutions across the country are trying to utilize Big Data to improve patient outcomes, many of their efforts are encountering considerable roadblocks. Health systems, hospitals and provider groups are recognizing the value of collecting vast amounts of data and analyzing the information to improve the coordination of patient care. However, they are finding that clinical data integration can often be a complex, costly and time-consuming undertaking, and many health systems, hospitals and provider groups have been reluctant to pull the trigger on this first step toward value-based care. Fortunately, there are new approaches available that cut both costs and time constraints.
Through a process called interfacing, health systems, hospitals and provider groups can now extract data from hundreds of different EMR and practice management systems without disrupting practice operations or patient care. While many organizations are focused on pushing insights back into EMRs and other legacy applications (what we call workflow integration), the higher payback comes when organizations extract the data through interfacing, consolidate it and present information to all users (physicians all the way to hospital and health system CEOs), via a patient-centric view. From the doctor’s perspective, they gain the ability to see a comprehensive patient profile, helping them close gaps in care and true-up patient information. If data is consolidated and presented in a timely and meaningful way to a user (i.e., physician), workflows will work themselves out. If the data is not consolidated and presented in a timely and meaningful way, the most seamless workflows will have little impact on achieving the goals of the Triple Aim Plus One: better health outcomes, lower costs and improved patient plus physician satisfaction.
Interoperability of systems and integration of claims and encounter data from across the care continuum is a logical first step in the shift toward accountable care. However, even as organizations gain additional access to volumes of patient data, the challenge lies in bringing context to this information with the goal of guiding care interventions to improve outcomes.
Sophisticated analytic processes are being developed to address this issue. These processes have the ability to drill down vast amounts of data in order to stratify patients by risk, identify gaps in care and uncover opportunities to improve outcomes. For example, advanced analytics can identify trends in patient behavior. If a chronically ill patient normally sees their physician every six months, but hasn’t been in the office for nearly a year, an ACO might be notified to contact the patient to discuss with their PCP or go so far as to schedule an in-home visit by a nurse practitioner.
By applying predictive modeling to this data, organizations can also prioritize opportunities according to their potential to impact outcomes and quality metrics. This allows for strategic allocation of costly resources, such as care coordination efforts and disease management programs, where they will be most effective.
If healthcare organizations aim to achieve cost-saving goals while improving the overall health of their membership, they need much more than data alone. These organizations need actionable insight. Armed with insight as it relates to gaps in care and opportunities to improve quality, value-based care models will be better prepared to provide targeted care to patients in the future.
Analytics is empowering frontline clinicians to distill actionable insights from abundant clinical data and to improve our healthcare system. With innovations in technology made over the past decade, we have unprecedented opportunities to use analytics to improve patient care – and even to broaden our thinking beyond prevailing medical opinion.
In fact, analytics are making inroads in the perinatal specialty. Prevailing medical opinion has paired excessive uterine contractions with high rates of neonatal depression in newborn babies. Using analytics, medical professionals at MedStar Franklin Square Medical Center in Baltimore have examined a large data set and demonstrated that this association is not so simple. A four-year study published in The Journal of Maternal-Fetal & Neonatal Medicine revealed that a better identifier of the risk of serious neonatal depression at birth is the presence of fetal heart rate decelerations in conjunction with excessive contractions, or uterine tachysystole (UT). Although UT occurred in approximately 20 percent of deliveries, only 1 percent of babies with UT developed neonatal depression.
How did these professionals reach this conclusion? By analyzing data gathered from thousands of electronic fetal strip tracings, equivalent to 27 miles of paper documentation. Research of this scale would be almost impossible without technologies for capturing and analyzing the data.
This is just one case where analytics is helping provide better, safer care to women and babies. What is even better? Innovative perinatal tracing technology can help obstetrics physicians and nurses distinguish babies at serious risk of neonatal depression – and get them the attention they need fast.
When we think about analytics, it’s often within the context of technology, computers and spreadsheets. However, advanced analytics actually have the power to predict the future and identify problems before they happen. For example, analytics can predict how a patient will respond to a given treatment. As a result, adoption of advanced analytics will benefit all stakeholders within our industry.
Key to these new capabilities is the growing volume, velocity and variety of data (including new EMR data) and value created by analytics. New systems and algorithms are being created to leverage data. For example, one system currently in production can query administrative data, generate dynamic care plans and empower providers with clinical decision support (CDS). Some systems can also apply “advanced” clinical decision support that includes the latest medical findings from clinical trials and scientific journals. This information can be used to alert care managers about gaps in care or opportunities to align treatment with best practices.
This analytics-based approach helps organizations:
- Predict which patients to contact, and why.
- Close the loop after a health event to help prevent avoidable readmissions.
- Eliminate patient safety issues and duplicate testing.
- Streamline provider workflow.
- Improve communication and collaboration among a patient’s care team.
- Enable patient engagement.
As a result, advanced analytics gives organizations unique ways to achieve the triple aim of improved care, reduced costs and a better patient experience.
The digital age is the age of Big Data where every piece of technology captures data available for later use. The McKinsey Global Institute (MGI) describes data generated in this way as digital “exhaust data,” data that are created as a by-product of other activities (Manyika, 2011).
The rapid expansion in the use of EMRs and digitally driven and connected technology such as MRI scanners, body sensors and automated lab tests, brings the era of Big Data to healthcare. As these technologies evolve and become more widely utilized, the data collected becomes more expansive and granular, yet insufficiently utilized.
The five broad areas to deliver that value are clinical operations, payment/pricing, R&D, new business models and public health. Sub-areas include comparative effectiveness research (CER), clinical decision support, remote patient monitoring, health economics and personalized medicine. The four large data sources for healthcare include clinical (e.g., EMR, images), pharmaceutical (e.g., clinical trials), administrative (e.g., utilization, claims), and consumer (e.g., home monitoring, retail purchases).
New analytic tools such as Semantic Web 3.0 – linked data – offer ways for machines to analyze these disparate data sources leveraging approaches impossible using standard relational databases and statistical methodologies.
The uses of Big Data are numerous and far-reaching. Only through innovative analytical techniques will we be able to truly leverage the healthcare data collected and improve the way we deliver care. Organizations that properly collect, analyze and utilize Big Data will achieve a significant competitive advantage over those organizations that fail to recognize the opportunity Big Data presents.
