Big data has fundamentally changed the way healthcare organizations manage, analyze, and leverage data. The healthcare industry has historically generated huge volumes of highly valuable and sensitive data that are mandated by regulatory requirements and compliance measures. Most of this data has been digitized now by the adoption of EHR. Especially in these times of a pandemic, these data are capable of providing crucial breakthroughs to support a wide range of healthcare operations, disease surveillance, population health management, and most importantly clinical decision support.
The healthcare data in the U.S. alone is predicted to soon reach the Yottabyte (1024 gigabytes), thereby providing an ideal opportunity for the implementation and expansion of Big Data Analytics. Application of big data analytics in healthcare could be overwhelming not just because of the volume of data but also due to the coordination, correlation, and processing of data from so many different sources. The sources include — clinical decision support systems such as physician’s handwritten notes, prescriptions, medical imagery, lab results, pharmacy inventory, and insurance; patient records from electronic patient records (EPRs); machine-generated vital data; emergency data, articles in medical journals and social media feeds.
We believe the implementation of big data analytics in healthcare is a great challenge and takes strenuous effort, but the takeaway in terms of the scope of growth (potential revenues/market share in the future), the betterment of healthcare quality, and reduction in human effort and cost overweighs any other difficulty and hardship in implementation.
Big Data Analytics — Healthcare Implementation Avenues:
Drug Discovery & Development: Our analysis of various white papers and medico-tech journals shows us that Big data analytics can help reduce the time and costs involved in discovering and developing new drugs and other health-related products. Predictive modeling using Big Data Analytics (BDA) would provide a leaner, faster, and more targeted R&D roadmap for the development of path-breaking drugs and healthcare monitoring devices.
Better Clinical Trials: Based on the current crisis for a COVID-19 vaccine and the fast track processing of clinical trials, we thought of analyzing if BDA could be used in improving the method behind the selection of patients for the trials and the evidence we found suggests that BDA based statistical analysis does help a great deal in improving clinical trial design and patient recruitment thus reducing trial failures and speeding new medicines to the market. The quick turnaround time and the exhaustive collation of healthcare data facilitated by BDA provide for an accurate clinical trial analysis to identify follow on indications and establish the adverse effects before the product reaches the market.
Improves Patient Outcomes: We believe the BDA has tremendous potential in improving patient outcomes because it helps doctors and other medical professionals be more efficient and accurate with their diagnoses and treatments. Medication and decision making based on BDA predictive models has been proven to be more efficient in providing healthcare. Combination and analysis of the structure and non-structured data from EMRs and EPRs, clinical data, financial and operational data, lifestyle data, and genomic data to match treatments with outcomes, predict accurately those patients who are at-risk for disease, thereby improving patient outcomes by forecasting the medical assistance required.
Genome Analytics: Through our research, we have observed that cutting edge BDA has been applied to execute gene sequencing more efficiently in a time and cost-effective manner. This enables genomic analysis to be adopted as a regular part of medical care decisions, thereby providing better patient outcomes.
Financial Aspects of Big Data Analytics Implementation in Healthcare:
Recent studies show us that Big Data Analytics not just increases the efficiency of healthcare processes but also finds and reduces inefficiencies that lie undetected in the system, thereby successfully cutting down costs.
- Research conducted by McKinsey & Company shows that big data applications in Healthcare could save Americans between $300 billion to $450 billion a year.
- By using BDA, Parkland Hospital in Dallas has reduced 30-day readmissions to their hospital and all area hospitals for Medicare patients with heart failure by 31%, for a savings of $500,000 a year.
- By using BDA, Beaufort Memorial Hospital in South Carolina determined that it could save an estimated $435,000 each year by discharging patients half a day early.
- The Minnesota Department of Health recently found that by using BDA, it can prevent unnecessary ER visits, effectively reducing the costs by about $ 2 billion. In the last year alone, the department, 50000 instances where primary care and community health services could have prevented the patient from an ER visit.
- BDA helps prevent unnecessary testing in hospitals. One such instance we noticed was reported by the St. Louis Children’s Hospital where the number of Dravet Syndrome tests conducted in newborns was drastically reduced after using analytical data. Each test costs $6000. Based on BDA, since there were no Dravet Syndrome cases in the last 5 years, the hospital modified its strategy and cut down on running these costly tests.
BDA to Reduce Fraudulent claims and Enhanced Security Measures:
The general fear of implementing Big data is that it has inherent security issues and that using it will make the healthcare organizations more vulnerable than they already are. But advances in Big data security and privacy measures have largely caused the advantages of implementing Big Data to overtake the risks.
Authentication: Most cryptographic protocols employs some method in order to intercept a variety of attacks especially ones such as the Man in the Middle attack (MITM). The usage of endpoint authentication as TLS and SSL to authenticate the server using a mutually trusted certificate authority. Additionally, hashing techniques such as SHA — 256, Kerberos mechanism, Bull’s eye algorithm is also implemented for authentication.
Encryption of Data: This aims to protect privacy and to maintain the ownership of the data throughout the data lifecycle. Although various encryption algorithms have been developed and deployed in BDA such as RSA, Rijndael, AES, and RC6 [24, 26, 27], DES, 3DES, RC4, Blowfish, IDEA, etc., the proper selection of suitable encryption algorithms to enforce secure storage remains a difficult problem.
Data Masking: It is one of the most popular approaches to achieve live data anonymization. It uses the strategy of de-identified data sets or masking those data that act as personal identifiers such as name, SSN, phone numbers, etc. A significant benefit of this technique is that the cost of securing a big data deployment is reduced. As secure data is migrated from a secure source into the platform, masking reduces the need for applying additional security controls on that data while it resides in the platform.
Access Controls: Role-based access control (RBAC) and attribute-based access control ABAC are the most popular models for Big data in EHR. RBAC and ABAC have shown some drawbacks when they are used singularly. Hence in order to attain fine-grained access control, it is important to use RBAC or ABAC in synchronization with encryption or data masking policies.
Monitoring and Auditing: Big data network security systems should be able to identify abnormalities at the earliest and rectify the problems concerning heterogeneous data. Therefore, in order to build an ideal Big Data Analytics system, there has been a model proposed which splits the Big Data Analytics system into four modules: data collection, integration, analysis, and interpretation. Data collection includes security and network devices logs and event information. The data integration process is performed by data filtering and classifying. In the data analysis module, correlations and association rules are determined to catch events. Finally, data interpretation provides visual and statistical outputs to a knowledge database that makes decisions, predicts network behavior, and responses events.
Big Data analytics has a major role in preventing and detecting fraudulent claims. In the US alone, the National Healthcare Anti-Fraud Association estimates the loss to health care fraud to be about $80 billion annually, a figure that’s actually on the more conservative side as other industry sources peg the losses to be over $200 billion. Only 5 percent of these losses are recovered annually.
For instance, we found a report that the Centers for Medicare and Medicaid Services in the U.S. had saved over $210.7 million in frauds in just a year after implementing a big data analytics platform to evaluate claim requests and flag those that differ from the norm.
Improved Staff Management:
Having too many employees working on any given day runs the risk of spending too much on labor. On the other hand, not enough staff can result in poor customer service. Any improvement in knowing the right number of people and the right number of skills to have at a hospital at any given time is important as it improves operational efficiency and effectiveness and improves the bottom line for hospitals. We found that U.S. hospitals have started using Big data predictive analytics to overcome these problems by leveraging a variety of sources to predict how many patients are expected to be at each hospital. The Bergen New Bridge Medical Center was struggling to cope with the frequent midday surges in patient traffic. Making use of recent historical data and running predictive algorithms, the hospital was able to better measure the patient traffic, forecast patient volume, and add an additional shift at 11 A.M.
Better Patient Engagement & Preventive Care:
Patient engagement is a key element of quality healthcare delivery. In order for patients to maintain their well-being, they must take an active role in their care and understand what they must do to stay healthy. Smart devices and mobile apps enable patients to track their own medical information. For example, they can track the number of steps they take over the course of the day or keep an eye on their heart rate during a workout. This health data is then stored in the cloud where doctors can continuously monitor it and supervise their patients. This means that patients don’t need to visit the medical practice for unnecessary checkups.
Patients with chronic diseases must engage in management strategies and adhere to treatment plans in order to maintain their health and keep care costs low. To ensure patients are actively participating in chronic disease management plans, organizations can use patient data to develop predictive risk scores powered by big data analytics and design individualized treatment strategies.
Based on our research we noted that the cost of delivering healthcare in the US is now more than $3 trillion annually. So big data, when combined with other health technologies, can help track and identify diseases long before they happen — and boost preventive care. Creating risk scores based on lab testing, biometric data, claims data, patient-generated health data, and the social determinants of health can give healthcare providers insight into which individuals might benefit from enhanced services or wellness activities. This increases the efficacy of preventive care.
According to a recent report by McKinsey & Company, healthcare costs now account for 17.6 percent of the GDP of the USA. We believe there is a prodigious scope of growth in terms of applying and maintaining Big Data Analytics based predictive data models for a variety of healthcare industry processes and innovations. A recent study shows that more than 60% of the healthcare organizations that have implemented predictive analytics models believe that the concept will save 25% or more of their annual costs in the next five years. It has been reported that effective usage of Big Data could add more than $300 million to the healthcare industry per year. We can convey with a conviction that Big Data in healthcare does not just have the potential for growth but it works and produces proven results for healthcare organizations that have already undertaken it.