- Having your head in the clouds is more a complement than a criticism these days as more and more technology services shift off site and into the hands of third-party providers.
For the healthcare industry, the cloud seems a natural fit. From EHRs to data storage to software as a service (SaaS) capabilities, cloud-based products offer lower costs, greater capacity for scalability, dedicated service and support, and near-continuous uptime.
But providers are no longer satisfied with merely having terabytes of virtual storage for their clinical and administrative information available to them. They want to be able to access and analyze it in a speedy, secure, and sophisticated manner. Can big data analytics become a viable cloud service offering, too?
Cloud EHR vendors like athenahealth and Practice Fusion, both of which have seen success in the general marketplace, may say that the question has already been answered. Both companies offer real-time analytics capabilities that draw on the huge pools of user data held in centralized repositories, ready for slicing and dicing into actionable insights.
Health information exchange organizations like Maine’s HealthInfoNet are also operating as cloud-based analytics hubs, collecting data from their member organizations and distributing information that can be used for population health management and clinical decision support.
For providers looking inward, however, the options aren’t quite as clear cut. Many organizations with server-based EHR systems or in-house data warehouses suffer from a condition known as being “information rich but knowledge poor,” write researchers from the University of Ontario Institute of Technology in a study published in the Journal of Medical Internet Research.
“Over the past few decades, our society has transitioned to a state where bottlenecks have shifted from a lack of data to limitations in extracting meaningful knowledge from an abundance of data and subsequently using that knowledge to drive decisions,” the authors write.
Huge volumes of clinical data added to EHRs at every moment cannot be quickly and thoroughly translated into concrete, timely clinical decision support (CDS) information due to the limited computing resources of most healthcare organizations.
Cloud-based analytics-as-a-service tools may be able to alleviate those pressure points and provide real-time CDS capabilities that will improve the quality of patient care by “combining the on-demand aspects of cloud computing with the democratization of information enabled by big data analytics.”
The researchers examined the impact of a cloud-based big data analytics framework in the NICU of a children’s hospital, one of the most challenging and data-heavy environments in healthcare. The sheer volume of available data is staggering.
“Heart rate, respiration rate, and blood oxygen are displayed each second resulting in 86,400 readings each day,” the authors write. “A premature newborn infant’s heart beats more than 7000 times an hour, which is approximately 170,000 times a day.”
“A newborn infant’s neurological function could also be monitored resulting in multiple waveforms each generating tens of millions of data points per patient per day. Drug and nutrition infusion data from smart infusion pumps can be more than 60 different fields provided every 10 seconds. Given that these infants can have more than 10 infusions concurrently, infusion can generate more than 1 GB of drug infusion data from a single patient per day.”
Yet very few healthcare organizations have the dedicated resources and tools that can generate meaningful reporting from these enormous data sets. Analytics-as-a-service providers do, and providers are increasingly becoming aware of it.
In 2014, a HIMSS Analytics survey found that 82 percent of hospitals were using cloud-based tools for health information exchange, data storage, and disaster recovery. In May of 2015, Black Book Research noted that the same percentage of small providers view cloud-based EHR technology as one of the most important trends in healthcare. A June report predicted that the cloud-based technology market will be worth $5.7 billion by the end of the decade.
“The focus of healthcare technology vendors needs to be on mobile, cloud, and data integration to successfully meet the future demands of the changing healthcare landscape,” said Doug Brown, Managing Partner of Black Book. “The majority (69 percent) of small practices plan to increase their investment in the advancements made by their current cloud-based vendor.”
Offloading the technical burdens of big data analytics is a popular strategy for cash-strapped providers, who are turning to outsourcing their IT infrastructure at a break-neck pace.
"Most hospital leaders see no choice but to evaluate and leverage next generation information and financial systems as an outsourced service in order to keep their organizations solvent and advancing technologically,” Brown observed in November, when a separate poll found that three-quarters of hospitals are interested in outsourcing their big data analytics, population health management, or revenue cycle analytics tasks.
At the Hospital for Sick Children in Toronto, the cloud-based big data analytics framework offers a complex and detailed look at critical issues, like sepsis and neonatal apnea, using novel data mining techniques.
“Data are sent to the physiological and clinical database via the stream-computing platform,” the researchers explain. “At the same time, the stream-computing platform runs the current deployed medical rule for sepsis detection.”
“Upon patient discharge, their data including physiological and clinical data will be loaded into the big data platform by the relational database management system (ie, bulk move). Temporal abstractions (TAs) are then performed for the specific service of critical care, in this case sepsis detection, which involves reading from the clinical rules and physiological/clinical tables, and writing the patient TA to the TA table.”
Temporal data mining can then be applied to the results, which may update the clinical rule table. The new rule can be run on the stream-computing platform to aid future real-time analytics.
Frameworks like these are important because they deviate from the traditional “black box” bedside medical devices that provide monitoring and alerts for most patients. Each device, such as a heart monitor or respiratory monitor, provides a single stream of data about a very specific attribute, and rarely interfaces with other monitors to deliver any overarching insight or prediction about the course of a condition that may only be identifiable through the interplay between multiple data points.
These devices may also face interoperability and integration challenges that cause vital data to fall by the wayside. Medical device integration problems are a critical patient safety concern, and prevent providers from viewing and using all available data to make split-second decisions about the course of treatment.
By its definition, big data analytics allows providers to synthesize multiple streams of data and generate new information from larger conglomerations of data. But the technical infrastructure required to complete these tasks can be confusing, expensive, and time-consuming to build and maintain.
Cloud-based services that dedicate all their time and effort to crafting the algorithms and data mining tools that allow more intelligent clinical decision support may be a promising option for healthcare providers who cannot develop these capabilities in-house due to time or budget constraints. By delegating complex big data analytics work to the experts, providers may be able to reap the rewards of robust clinical decision support capabilities while freeing up resources to devote to other aspects of high quality patient care.