Video 2 – Big Data Analytics in Health Care



welcome to the second video in a series of presentations selected for topics in biomedical informatics my name is michael byrne our a biomedical informatics research fellow at the University of New Mexico this video is produced by the University of New Mexico's Health Sciences Library and informatics Center and is licensed under a creative commons attribution-noncommercial-sharealike 4.0 international license this means that you're welcome to share and use this video for any non-commercial purpose as long as you give credit to the original authors we hope this and future videos will be informative to those interested in learning more about informatics in healthcare the title of this presentation is big data analytics in healthcare and is intended to give a brief overview of big data as well as some of its potential applications in healthcare as many of you may be aware we're generating more data today than ever before what you may not know however is that it's been estimated that 90% of the world's data was actually generated within the past two years alone to put this into context Americans collectively consume around 18 terabytes of wireless data 86,000 hours of Netflix video and send 3.5 million text messages every minute of every day one word that attempts to describe the scale and scope of this phenomenon is the term big data but what exactly is big data there have been several definitions proposed over the years but in general it refers to data that exceeds the processing capacity of traditional hardware and data storage structures in other words the data are either too big are being generated too quickly for standard hardware to keep up this results in a number of challenges such as issues related to data capture storage analysis and visualization however size is not everything in fact the data meeting one or more of the following criteria can be considered big data these four criteria are often referred to collectively as the four V's of big data the first criterion volume refers to the total amount of disk space required to store the data and is usually the first thing that comes to mind when talking about Big Data the volume of big data is generally much larger than what we are typically used to and can range from terabytes to zettabytes the second criterion is variety which refers to the various forms the data can take in particular it describes whether or not the data is structured or unstructured structured data consists of data that can be easily stored queried and analyzed by machine within the context of healthcare structured data often includes electronic claims data some clinical data such as diagnosis and medications as well as laboratory data however most of the data being generated today about 80% is actually unstructured and is accumulating at a rate 15 times greater than that of structured data other sources of data making their way into the clinic include those produced by wearable fitness devices remote blood pressure monitors continuous glucose monitors and even social media the third criterion is velocity and refers to the rate at which the data are being produced or collected most traditional healthcare data has been static for example paper files and x-ray films but newer technologists medical devices allow real-time data collection and monitoring examples include real-time monitoring of patient blood pressure and heart rate in the ICU which may produce multiple readings a second these high fidelity data require more sophisticated signal processing techniques that can operate on a streaming data allowing for detection of clinical events in real-time the fourth and final criterion is veracity which refers to the overall quality of the data since data may be sourced from various systems there's the possibility that data relating to the same construct may contradict one another or be missing altogether furthermore most analytical methods assume clean and precise data such as that typically found in information Mart's or data warehouses as a result new methods of analysis are required that are robust to these sorts of issues since many applications built on top of these data are used to inform clinical decision making it is imperative that the underlying information that be as accurate as possible general most applications of big data analytics in healthcare can be organized into one of three categories the use of analytics to improve quality and patient experience of care analytics to improve population health and finally analytics to reduce the cost of health care which accounts for approximately 17% of the u.s. GDP additional use cases include using analytics to support research for example G&S health care uses cause and effect models derived from large observational data sets to determine drivers of patient outcomes analytics can also be used in service of population health as mentioned on the previous slide for example Camden coalition of health care providers uses predictive analytics to identify the sickest members of the community and attempt to reduce costs and admissions another example is the use of analytics to support surveillance efforts for example sick weather LLC is a company that uses social media to track outbreaks and deliver real-time illness forecasts while the use of big data in healthcare is promising there are several major challenges that must be addressed for one much of the healthcare data that exists is protected meaning access is restricted primarily to those directly involved in the patient's care this presents several challenges to third-party companies hoping to use the data to develop their products there are also issues concerning potential misuse of patient data by third parties for example insurance companies denying or shifting coverage based on genetic risk factors additionally most institutional policies require audit trails so that a list of disclosures can be provided to the patients if data is relinquished to the cloud it may be difficult or impossible to know who has access to the data despite these challenges it is clear that the use of big data in analytics in healthcare provides tremendous opportunity as we strive to find a ways to improve the quality and efficiency of care while managing rising costs I hope this video has provided some clarity to the term big data and its potential applications in healthcare as well as an appreciation for some of the challenges we are likely to face moving forward this concludes the presentation on big data analytics in healthcare for more information on this and related topics or to find out about informatics fellowships offered the University of New Mexico please visit our website or contact the biomedical informatics research training and scholarship program

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *