Worried about Big Data? Blame the Black Death.
The bubonic plague claimed the lives of exactly 7,165 Londoners during a one week period in the infamous 1665-66 outbreak. During that same seven-day period, another 1,132 locals died from then-common diseases like smallpox and rickets. How do we know this? Because the practice of data collection in healthcare is much older than the actual practice of modern medicine: Hundreds of years before the discovery of antibiotics eliminated the possibility of a similar epidemic, 17th century English publishers began gathering local death records, tallying up the causes and selling the results back to the local community so specific neighborhoods afflicted by disease could be avoided.
In the process, these London publishers also laid the framework for healthcare’s earliest structured datasets by creating a standardized list of 60 conditions that contributed to an individual’s death. In fact, the most recent update of the World Health Organization’s International Statistical Classification of Diseases and Related Health Problems, which includes 16,000 diagnosis codes, can be traced back to this original list of fatal 17th century maladies in London.
Today, the $3.2 trillion healthcare sector faces another inflection point.5 The Affordable Care Act means more individuals than ever before have regular access to healthcare services, while once-untreatable diseases are becoming manageable chronic conditions. The business of healthcare, too, is changing quickly. Our fee-for-service model that incentivized volume is transitioning to a value-based system that reimburses for outcomes. Healthcare also is digitizing rapidly, a migration that’s creating entirely new challenges like IT interoperability and data security.
The ongoing digitization of healthcare also is forcing us to confront issues, not unlike those 350 years ago in London: What healthcare data is valuable—and why?
Roughly 50 percent of all U.S. healthcare costs are generated by about 5 percent of patients. Addressing the lopsided nature of healthcare spending in the United States has long been a priority for stakeholders, but it’s becoming more immediate as bundled payment programs come online and force providers and payers to share the financial risks of treating patients.
The vast amount of clinical data collected as a natural byproduct of treating patients holds tremendous promise, but we’re far from realizing its full potential: The complexities involved in crunching the necessary information simply exceed current industry capabilities. For one, today’s clinical datasets tend to be retrospective, an orientation that, for example, provides helpful insights into what’s happened to these high-cost patients (readmitted to the hospital, infections, etc.). But current capabilities don’t readily shed much light on what’s to come, which is vastly more important.
Still, advances are being made. One area that shows considerable promise is predictive analytics, a field that combines Big Data with artificial intelligence. For example, International Business Machines’ Watson supercomputer accurately prescribed cancer treatments in 99 percent of the cases it reviewed recently at the North Carolina School of Medicine. During the same trial—and perhaps more impressively—the IBM supercomputer “was able to provide additional options missed by its human counterparts in 30 percent of the cases.”
Another priority for healthcare data scientists involves connecting the dots between seemingly unrelated data points across a vast spectrum of information, including clinical, financial, socioeconomic and geographic. For example, a 2014 Health Affairs study examined what factors should be considered when using analytics to predict patients who “are likely to be high risk or high cost.”
“Attributes associated with high-cost patients may include behavioral health problems or socioeconomic factors such as poverty or racial minority status,” the study concluded. “Thus, integrating data about mental health, socioeconomic status, or other issues such as marital and living status from various sources may significantly change the quality of the predictions that can be made.”
Financial information, in particular, must play a larger role in healthcare decision making in the coming years, as more advanced treatment options come online, patients take on more of their own healthcare expenses and healthcare transitions away from volume-based payments. For example, Medicare has set a goal of converting 50 percent of its fee-for-service payments to value-based reimbursement programs by 2018. As part of this cost-containment initiative, the agency is rolling out MACRA, which once fully implemented will incentivize physicians to accept bundled reimbursements in cases like heart attacks, strokes, and total joint replacement.9 In order to successfully treat them, however, participating physicians will need to carefully consider a wide range of not just financial information, but also clinical data in order to manage countless challenges related to patient selection, staffing levels, supply chain and revenue cycle.
According to a recent Healthcare IT survey, 2017 is expected to be a breakthrough year for data and analytics in healthcare.10 More than 80 percent of participants interviewed for the survey said they were adding analytics capabilities to existing systems or building new applications from scratch. One-third also expected to invest in artificial intelligence capabilities, while nearly two-thirds planned to launch prescriptive analytics tools before the end of the year. They also have high hopes for these new applications: More than two-thirds said their investments would enhance delivery or improve the quality of care at their facilities.
Interested in learning more about some of the ways Alveo is harnessing the power of Big Data to enable your healthcare business? Let’s talk. We provide clients with tailored business solutions that simplify workflow, minimize operating costs, and maximize reimbursements. Our services include patient eligibility verification, claims processing, remittance advice, patient statements, patient payment portal, customized reporting and analytics, and a unique electronic prior authorization solution set. We process more than $1 billion claims each month with a 98 percent annual client retention rate. And through our connections with more than 4,000 payers, we possess a 96 percent clean claim rate.