Big Data in Healthcare
The rising costs of healthcare in the United States could mean big things for Big Data in the healthcare industry.
Centers for Medicare & Medicaid Services reported that healthcare expenses accounted for 17.9% of America’s gross domestic product (GDP) in 2016, topping the benchmark for a nation of this size and wealth by nearly $600 billion. This percentage equates to $3.3 trillion spent annually on direct and indirect healthcare expenses, which include prescription drugs, inpatient and outpatient services, long-term care, and medical equipment. With costs expected to reach $5.7 trillion by 2026, the healthcare industry is in need of change, and Big Data is the driving force to enable it.
Articulated by Doug Laney in the early 2000’s, Big Data is defined by the three V’s — volume, velocity, and variety.
VOLUME: large quantities of data are accessed from multiple sources, including streams from the IT department and Internet of Things (IoT), social media, and publicly available sources.
VELOCITY: Data is collected at exceptional speeds and must be handled promptly. The use of electronic health records (EHR), smart metering, and sensors create a need for data to be accessed in real-time.
VARIETY: all types of data formats are processed, from structured data found in traditional databases to unstructured data like emails and text files.
Big Data was created on the principle that the volume of data points gathered and compared directly correlates to the accuracy of resulting information and predictions. Increasing the size of the data set inherently uncovers relationships and patterns that were previously hidden. These findings can be analyzed to determine the root causes of failure, detect fraudulent behavior, reduce costs, increase product development, optimize offerings, and enable a more informed decision-making process. Specifically relating to the healthcare industry, analytics from Big Data could be utilized to lower treatment costs, avoid preventable diseases, predict epidemic outbreaks, and improve patients’ quality of life. The question remains — where exactly are healthcare providers getting this data?
Over the past decade, there has been an increase in the supply of information available to healthcare providers due to incentives from insurers, payors, pharmaceutical companies, public stakeholders, and the U.S. government. Payors and insurers create a greater need to share data by shifting to risk-sharing arrangements for physicians that prioritize outcomes rather than volume as found in traditional fee-for-service compensation. Pharmaceutical companies offer their own motivator by assessing drugs’ ability to improve patients’ health as a basis for reimbursement. The U.S. government and public stakeholders have opened their records and shared data from clinical trials and patients covered under public insurance programs. The efforts of these healthcare providers have created an increasing need for Big Data.
The recent influx of data is being used more heavily in five major areas across the healthcare industry — real-time monitoring, healthcare IoT, predictive analytics, prescriptive analytics, and telemedicine.
Real-time Monitoring
Real-time monitoring is one of the biggest tools healthcare providers are using to administer proactive care through the use of wearable sensors and devices. Remote and in-home monitors track patients’ vital signs and send real-time results to primary care physicians. Machine learning is then applied to this data, resulting in more effective interventions and lifesaving assessments. Real-time monitoring limits the need for face-to-face visits, saving hospitals and patients unnecessary costs.
Healthcare IoT
Interconnected devices and sensors used for real-time patient monitoring generate a constant stream of data. This large volume of data is shared between devices, making its way to physicians and patients. With the help of devices like glucose monitors, fetal monitors, blood pressure monitors, and electrocardiograms, new understandings of patient behavior are constantly being uncovered. This is even more true with data gathered from smart medication dispensers that monitor whether patients’ doses have been taken or not. The combination of connected smart devices and an understanding of patient patterns helps reduce direct visits to the doctor’s office, lowering costs and improving patient care. According to Carol McDonald of MapR, the limitless possibilities of Healthcare IoT has made it so appealing that an estimated $120 billion will be invested into the technology within the next four years.
Predictive Analytics
Big Data fuels predictive analytics. This is especially true for socioeconomic data. Attributes like zip codes are used to predict whether or not patients own a vehicle, will comply with their medications, and are more likely to follow up with appointments. Predictive analytics applied to socioeconomic data could help hospitals determine if it’s cheaper to send a taxi or rideshare to pick a patient up for their appointment rather than have them miss it and be readmitted.
Prescriptive Analytics
Subsequent to predictive analytics, prescriptive analytics is applied to Big Data to recommend treatments for patients based on their anticipated behavior. For prescriptive analytics to be possible, historical data from patients with similar conditions need to be available via real-time monitoring, IoT, social media, and medical databases. Machine learning is applied to this data, predicting the trajectory of a patient over time. These results help physicians intervene when a patient is expected to veer off the recommended course, increasing the likelihood of a treatment’s success.
Telemedicine
Telemedicine has become increasingly popular over the years because it allows people to receive treatment remotely through various avenues. Websites like WebMD give patients the ability to self-diagnose based on the symptoms they input, while applications like Healthtap offer a one-on-one service between patients and physicians for diagnoses. Unlike Healthcare IoT and real-time monitoring, telemedicine doesn’t utilize Big Data — it supplements it. Interactions from remote patient monitoring, mobile health, and asynchronous telemedicine leave a data trail which is later amassed into the vast pool of Big Data.
As Big Data becomes more relevant in the healthcare industry, more and more non-traditional players are investing their resources. With unmatched data assets like internet search history, purchasing power, location data, communication preferences, and media consumption habits, it’s no surprise that Google, Amazon, and Apple are tossing their hats into the ring.
Google’s current focus is on researching how to apply artificial intelligence (AI) to genomic data for diagnostics and developing innovative medical devices. They currently support healthcare organizations in the cloud and have partnered with the Mayo Clinic to enhance their search engine with trustworthy sources on health information and common concerns.
Amazon
Amazon will likely utilize their 2017 purchase of Whole Foods to provide valuable insights into what people eat, how they shop, and if they’re more likely to make healthy choices. This data could be used to provide almost-instant grocery delivery services for people without cars and an AI-powered consumer recommendation engine to suggest healthier food options. Amazon’s Alexa is the perfect device to collect data on the home environment and provide 24/7 access to patients who need regular health maintenance. Amazon’s recent partnership with Epic Systems will also turn Alexa into a documentation assistant, making EHRs more accessible and user-friendly.
Apple
Apple is leveraging hardware and software to improve the healthcare industry. The Apple Watch tracks users’ fitness while the iPad and iPhone are used as vehicles of communication and data access for patients and physicians. Apple’s new Health Records App allows patients to view aggregated data from participating organizations, solving the patient data access problem by integrating health records onto familiar platforms. This app makes it easier for patients to access and understand their EHRs.
The possibilities of Big Data are limitless for patients and healthcare providers alike. Patients receive higher quality care and have the means to track the progress of their own health. Physicians are able to prevent additional hospital visits, give more insightful diagnoses, and provide tailored care to each of their patients. Hospitals are able to reduce costs by assessing methods and treatments faster and keeping better track of their inventory. Some skeptics will always be concerned about patient confidentiality and the security of the centralization of medical data. However, the development of new treatments and emergent understanding of the human body far outweigh the potential negatives of Big Data.
