This white paper offers a detailed perspective on how big data is impacting the healthcare industry and its underlying implication on the industry as a whole. It outlines the role of big data in healthcare, its benefits, core components and challenges faced by the healthcare sector towards full-fledged adoption & implementation.
Over the last decade, tremendous progress has been witnessed in the volumes of data that is generated each day through various activities. The capacity of people in harnessing technology to collect, evaluate and understand such data has also increased substantially. The propensity of these trends to intersect with each other has come to be termed as ‘Big Data’. Needless to say, big data has emerged to be critically important to businesses across diverse industrial sectors by enabling them to enhance their efficiency and augment their productivity. The idea that big data plays a vital role in making the world a better place may seem farfetched but, it is true. Evidence can be found in the way big data is being utilized in the healthcare sector.
Healthcare is one of the sectors that has immensely benefitted from big data. Other than enhancing efficiency and increasing productivity, big data in healthcare is being effectively utilized to anticipate potential epidemics, alleviate disease, enhance life expectancy and evade preventable deaths. While population continues to grow and with increased life spans, changes have been witnessed in treatment methodologies. However, it is worthwhile to note that the changes have been largely data driven. The challenge here is to get a better understanding about patients’ health as early as possible. This is necessary to pinpoint any serious illness a patient may have, early on. Identifying a life threatening illness at an early stage enables healthcare professionals to treat and manage the ailment with relative ease as compared to the condition being diagnosed at a much later stage. This is made possible by assessing and analyzing big data.
However, the diverse nature of the data collection community renders the process of extracting and integrating data thoroughly challenging.Every individual stakeholder such as healthcare providers, employers, disease management organizations, payers, wellness institutions, genetic testing organizations and patients, collects data. This has led to path breaking efforts that are spurred by strategic partnerships amongst medical and data professionals. This presents an immense opportunity to look into the future and pinpoint potential challenges before they actually present themselves. Big data from diverse sources viz., genetic records, medical and insurance records, data from social media and wearable sensors are effectively harnessed to outline a detailed picture of the patient and offer a customized healthcare solution.
Predictive analytics has been tactically deployed within big data to derive innovative insights. Apart from the routine clinical and administrative data, amalgamating new data derived from patients pertaining to their health records helps to anticipate ailments and offer timely intervention to patients. Anticipating underlying health issues enables healthcare providers to offer preventive remedies to counteract the effects of such issues. All this is possible by collecting, analyzing and understanding big data.
Big Data in Healthcare
So what constitutes big data? A 2012 report outlines big data as huge quantities of intricate, high velocity and variable data that warrants innovative techniques and technologies to facilitate the acquisition, storage, distribution, management and analysis of information in a cohesive manner. Big data displays characteristics such as variety, velocity, volume and a feature that is particularly visible in the healthcare sector – veracity. Together they are known as the 4V’s of big data in healthcare. Prevailing analytical methods are applicable to the large amount of unanalyzed data pertaining to health and medical records of patients to arrive at resourceful insights that can be intrinsically applied to treatment methodologies. Preferably, available medical data can be analyzed to provide appropriate information to care givers which can then be utilized by them to streamline treatment procedures for a specific patient.
According to (Raghupathi & Viju, 2014) the volume of data in healthcare is projected to grow sizably in the coming years. Moreover, a shift has been observed in models of healthcare refund wherein, significant use and payment for performance occupy crucial importance in the current day healthcare environment. Though the motivating factor here is not profit, it is vital that healthcare organizations obtain necessary tools, techniques and infrastructure to effectively leverage big data or they stand to lose millions in revenue and profits.
At present the healthcare sector is facing a digital revolution. Volume of data is now shifting from primary science to genomics that are clinically based and personalized medicine. Big data is evolving non-stop in the healthcare sector both at a personal and large scale levels (Issa, Byers, & Dakshanamurthy, 2014). As a matter of fact, clinical phenotypes are now being biochemically and quantitatively expressed by utilizing proteomics, metabolomics and transcriptomics. Accumulating and analyzing big data is set to emerge as a core factor that drives innovation in healthcare. New developments in big data analytics has major implications not just on healthcare delivery from a patient and care provider’s perspective but it will also prove to be critical in restructuring biomedical discovery. For instance, a single human genome was decoded in a decade initially, however, with the advent of big data, a human genome can now be decoded within a week by utilizing modern DNA sequencing and informatics approaches. Big data is now being applied in healthcare in an all inclusive manner such that it encompasses:-
- Clinical Discovery
- Systems Medicine & Pharmacology
- Toxicity Prediction
- Electronic Medical Records
Role of Big Data in Healthcare
Loosely structured patient related data is being generated by healthcare monitoring systems from various sensors over a period of time. Such healthcare monitoring systems are intricate and demand the need for effective algorithms and computational prowess to process and analyze the raw data. Big data in healthcare can be termed as data that has been derived from various sensors that encompasses medical, traffic and social data (Augustine, 2014).
Big data analysis infrastructure in healthcare can be developed by putting in place a robust mechanism that is required to accumulate, organize and process data with an objective to derive useful information. This facet can be attributed as data acquisition, data organization and data processing. Acquiring the data can prove to be a key challenge in a big data environment. Since healthcare monitoring systems deal with large quantities of data, a low latency is needed in data capturing while simple query methods can be implemented to process a huge quantity of data.
Objective of applying Big Data Analytics in Healthcare
As data in healthcare is available in large quantities, it is necessary for the healthcare monitoring system to assimilate and process existing data from the primary location where it is stored. This is easily facilitated by ‘Apache Hadoop’ software that offers a unique technology to process huge quantities of data and also enables the data to be kept in original data clusters. Moreover, it is essential that the data in big data be processed in a distributed environment. Analysis of medical information merits the need for deploying a statistical and mining approach. Delivery of analyzed data within a comparatively faster turnaround time assumes high priority in such a scenario. The amalgamation of data pertaining to patients, data pertaining to effects of drugs, research and development data, financial data and medical data by healthcare organizations, can be instrumental in determining existing patterns that lead to providing proactive and improved healthcare solutions. Further, in the chance that healthcare organizations integrate patient-centric data in tandem with data derived from social media within their data management systems, they stand to gain an inclusive insight into the correlation between such data. The primary idea of infusing big data in healthcare is toenable the industry to acquire data from any relevant source, support the data thus collected and analyze it to arrive at conclusions that enable them to:-
- Reduce cost
- Reduce time
- Develop new research
- Optimize decision making
Four V’s of Big Data in Healthcare
One billion smart phones and three billion IP enabled devices are expected to enter service in the coming three years. Remote health monitoring devices would be used by approximately around 4.9 million patients. In addition, around three million patients would be using a remote monitoring device by means of a smart phone hub. Healthcare and medical app downloads are projected at 142 million. In 2012 alone, healthcare had generated around 500 petabytes of data. This figure in 2020 is projected to grow to 25000 petabytes. Summing it in short, it can be conclusively said that big data is poised to emerge as the buzzword for the next generation. Data that supersedes the processing capability of traditional data management systems is termed as big data. Big data displays tendencies where the data is rather large, fast paced and doesn’t conform to the existing database architecture. Big data in healthcare has four prominent dimensions which are generally referred as four V’s viz., volume, variety, velocity and veracity. Volume in this context refers to data in terabytes and zettabytes, variety would refer to unstructured, semi-structured and structured data, velocity would refer to batch processing to real-time streaming of data and lastly veracity would relate to quality and relevance of data. Applicability of each of these dimensions to healthcare data is discussed below in detail (Sreekanth, RR, & Arvind Kumar, 2014).
Data volumes globally, are witnessing an exponential growth. However, growth of data in healthcare is augmented by digitizing existing data and through new data generation. Existing healthcare data that is overwhelming to say the least encompasses individual medical records, clinical trial data, radiology images, population data, human genetics, genomic sequences and FDA submissions. The growth of data in healthcare is also being spurred by new forms of big byte data that relates to 3D imaging, biometric sensor readings and genomics.
An aspect that renders healthcare data as interesting and challenging is the existence of a large variety of data, either in unstructured, semi-structured or structured formats. Since ages, point of care has a tendency to generate data that is largely unstructured. Data in this case would refer to medical records, handwritten notes by doctors and nurses, doctor’s prescriptions on paper, admission and discharge records in hospitals, radiograph films, CT scan, MRI and other relevant images etc. Structured data is comparatively easy to store, query, recall, analyze and can be manipulated by machines. Structured and semi-structured data encapsulates data pertaining to electronic health records, e-accounting and billing, actuarial and clinical data to some extent and readings of laboratory instruments.
The rate at which new data is being generated presents a unique challenge to healthcare organizations. While volume and variety of data that has been accumulated and stored has changed considerably, likewise,the velocity and speed necessary to recover, compare, compute and analyze healthcare data to make strategic decisions has also changed. The situation has witnessed a remarkable transition from sluggish batch processed data handling to real-time data processing. Nowadays, data velocity can be crucial in some medical scenarios where real-time data can be instrumental in life and death situations.
Issues in data quality are of critical concern in the healthcare sector owing to two causes; 1. Decisions in life and death situations hinge largely on access to accurate data. 2. Data in healthcare, especially related to unstructured data is largely variable and presents a scope for error, for instance, handwritten prescriptions that are unreadable. Data veracity in healthcare is confronted with the same challenges as financial data, more so when it concerns payers; veracity in terms of whether it is the right patient, payer, hospital, reimbursement code etc. Veracity issues could further relate to questions such as; whether prescriptions, diagnoses,procedures and outcomes are accurately captured.
Big Benefits of Big Data in Healthcare
Diverse healthcare organizations be it small single doctor clinics, large hospital networks, multi provider groups, and organizations offering advanced and inclusive care, stand to derive maximum benefits by digitizing, combining and adopting big data analytics within their operational systems. Some of the benefits to healthcare organizations would relate to an enhanced ability to detect complex diseases during their initial stages which is imperative for effective and successful treatment. It also facilitates healthcare organizations to provide customized treatment plans to specific individuals based on their preferences. Offering treatment for lifestyle related diseases by intrinsically analyzing data pertaining to lifestyle patterns of patients. In addition, big data also facilitates healthcare organizations to detect frauds and malpractices with relative ease and efficiency (Raghupathi & Viju, 2014).
Big data in healthcare is particularly beneficial in answering several critical questions. It can also be effectively utilized to forecast certain developments or outcomes or a future scenario can be anticipated on the basis of existing historical data. For instance; length of stay of patients, propensity of a patient to opt for elective surgery, patients who are unlikely to gain from surgery, risk to patients arising from medical complications, sepsis or other relevant illness acquired during hospitalization etc. A report by leading research firm McKinsey estimates that big data in healthcare would facilitate savings to the tune of more than $300 billion per year in the United States alone. Two core areas of healthcare where huge savings can be affected through big data are in clinical operations and research and development. The McKinsey report also emphasizes that big data could be effectively utilized to reduce waste and inefficiency in three key areas of healthcare viz., clinical operations, research and development and public health.
Challenges for Big Data in Healthcare
Big data in healthcare is confronted with several challenges that complicate the usage of healthcare data to its maximum potential. To start with, data available with several healthcare organizations (hospitals in general) are more often than not fragmented and exists in silos. As a matter of fact, data related to administrative activities that involves reimbursement, information related to cost and claims process are amassed to be utilized largely by the financial department and by the operations management personnel. While financial data is utilized to execute the business aspects of healthcare, such data is not utilized to notify treatment protocols or patient care. At the same time, clinical data that includes medical history of patients, vital statistics,rate of progress and diagnostic outcomes are stored within electronic health records. This data is exclusively maintained and periodically accessed by care givers, nurses and other core clinical personnel. Also such data is utilized by care givers to monitor patient care and convey appropriate treatment plans to patients. Other data pertaining to quality and outcomes that encompasses infection on surgical sites, rate at which patients return to surgery etc., are handled and accessed by the quality or risk management departments. Results of a clinical informatics survey revealed that 43 per cent of respondents were of the opinion that data maintained in silos within an organization emerged as a major challenge when it came to analyzing clinical data (White, 2014).
Another challenge faced by the healthcare sector while leveraging big data pertains to privacy of patients’ data. Data sharing amongst healthcare organizations is critical and the formation of regional health information companies hinged largely on assimilating data from various stakeholders like payers, providers and public health organizations. As per the Health Insurance Portability and Accountability Act [HIPAA] organizationsare required to protect data pertaining to patients. In this scenario, data sharing can only be facilitated after de-identification however, the challenge lies in the organization’s ability to protect patients’ identity, either directly or indirectly while sustaining the usefulness of existing data. Thus, the major challenges for big data in healthcare relates to creating an appropriate balance amongst patient privacy and maintaining data integrity.
Consumer markets have successfully adopted and implemented big data within their business processes however, implementing big data analytics within healthcare presents unique challenges which need to be overcome. One amongst the several challenges that needs to be overcome relates to fragmented data and data integrity. Big data in healthcare holds tremendous promise and potential to substantially alter the way healthcare is delivered. It can also make an impact on the way care providers adopt and utilize cutting edge technologies to derive strategic perceptions from clinical and other data and arrive at informed decisions(Raghupathi & Viju, 2014).
The coming days will bear testimony to a rampant and enhanced use and implementation of big data analytics within the healthcare sector. However, to completely embrace big data and adopting it in total will most certainly depend on tactically eliminating the challenges posed by big data in healthcare. Since the importance of big data is gradually dawning on stakeholders, concerns related to maintaining privacy, ensuring security, setting standards, appropriate governance and continuous enhancement of the tools and technology will gain prominence. Though big data in the healthcare sector is still in the early phase of development, brisk progress in existing tools and platforms can aggravate the rate at which it matures.
The nuances of medical writing and the upward trends
The regulatory stipulation of Bio-Medical devices to achieve a CE
Get in touch to find out how we can help your organisation.
Our team are ready and would like to speak to you, to understand more about your organisation and how we can help you achieve your goals.