Implications of big data analytics in developing healthcare frameworks – A review

The domain of healthcare acquired its influence by the impact of big data since the data sources involved in the healthcare organizations are well-known for their volume, heterogeneous complexity and high dynamism. Though the role of big data analytical techniques, platforms, tools are realized among various domains, their impact on healthcare organization for implementing and delivering novel use-cases for potential healthcare applications shows promising research directions. In the context of big data, the success of healthcare applications solely depends on the underlying architecture and utilization of appropriate tools as evidenced in pioneering research attempts. Novel research works have been carried out for deriving application specific healthcare frameworks that offer diversified data analytical capabilities for handling sources of data ranging from electronic health records to medical images. In this paper, we have presented various analytical avenues that exist in the patient-centric healthcare system from the perspective of various stakeholders. We have also reviewed various big data frameworks with respect to underlying data sources, analytical capability and application areas. In addition, the implication of big data tools in developing healthcare eco system is also presented.

The full article can be viewed below.

Big Data Analytics Best Practices: Understanding its capabilities and potential benefits for healthcare organizations​

To date, the health care industry has not fully grasped the potential benefits to be gained from big data analytics. While the constantly growing body of academic research on big data analytics is mostly technology oriented, a better understanding of the strategic implications of big data is urgently needed. To address this lack, this study examines the historical development, architectural design and component functionalities of big data analytics. From content analysis of 26 big data implementation cases in healthcare, we were able to identify five big data analytics capabilities: analytical capability for patterns of care, unstructured data analytical capability, decision support capability, predictive capability, and traceability. We also mapped the benefits driven by big data analytics in terms of information technology (IT) infrastructure, operational, organizational, managerial and strategic areas. In addition, we recommend five strategies for healthcare organizations that are considering to adopt big data analytics technologies. Our findings will help healthcare organizations understand the big data analytics capabilities and potential benefits and support them seeking to formulate more effective data-driven analytics strategies.

Best Practices

• Governance - Implementing big data governance successfully can allow for more efficient utilization of data
• Sharing - Developing an information sharing culture improves quality and accuracy 
• Training - Training key personnel to use big data analytics is the key to utilizing outputs effectively
• Cloud computing - Incorporating cloud computing into the organization's big data analytics can help address cost and data storage issues.  However, this must be balanced with patient information protection
• New ideas - Generating new business ideas from big data analytics promotes innovation, productivity, and competitiveness  

The full article can be viewed below.  

Scalable and accurate deep learning with electronic health records

Predictive modeling with electronic health record (EHR) data is anticipated to drive personalized medicine and improve healthcare quality. Constructing predictive statistical models typically requires extraction of curated predictor variables from normalized EHR data, a labor-intensive process that discards the vast majority of information in each patient’s record. We propose a representation of patients’ entire raw EHR records based on the Fast Healthcare Interoperability Resources (FHIR) format. We demonstrate that deep learning methods using this representation are capable of accurately predicting multiple medical events from multiple centers without site-specific data harmonization. We validated our approach using de-identified EHR data from two US academic medical centers with 216,221 adult patients hospitalized for at least 24 h. In the sequential format we propose, this volume of EHR data unrolled into a total of 46,864,534,945 data points, including clinical notes. Deep learning models achieved high accuracy for tasks such as predicting: in-hospital mortality (area under the receiver operator curve [AUROC] across sites 0.93–0.94), 30-day unplanned readmission (AUROC 0.75–0.76), prolonged length of stay (AUROC 0.85–0.86), and all of a patient’s final discharge diagnoses (frequency-weighted AUROC 0.90). These models outperformed traditional, clinically-used predictive models in all cases. We believe that this approach can be used to create accurate and scalable predictions for a variety of clinical scenarios. In a case study of a particular prediction, we demonstrate that neural networks can be used to identify relevant information from the patient’s chart.

The full article can be viewed below.
 

Authors: Wullianallur Raghupathi and Viju Raghupathi

The healthcare industry historically has generated large amounts of data, driven by record keeping, compliance & regulatory requirements, and patient care. While most data is stored in hard copy form, the current trend is toward rapid digitization of these large amounts of data. Driven by mandatory requirements and the potential to improve the quality of healthcare delivery meanwhile reducing the costs, these massive quantities of data (known as ‘big data’) hold the promise of supporting a wide range of medical and healthcare functions, including among others clinical decision support, disease surveillance, and population health management. Reports say data from the U.S. healthcare system alone reached, in 2011, 150 exabytes. At this rate of growth, big data for U.S. healthcare will soon reach the zettabyte (10²¹ gigabytes) scale and, not long after, the yottabyte (10²⁴ gigabytes). Kaiser Permanente, the California-based health network, which has more than 9 million members, is believed to have between 26.5 and 44 petabytes of potentially rich data from EHRs, including images and annotations.

Click below to read more.

PULSE POLL: Physician attention

Each month, the Truven Health Analytics PULSE® Healthcare Survey polls approximately 3,000 Americans to gauge attitudes and opinions on a wide range of healthcare issues. This independent, multi-modal (land line, cell phone, Internet) survey collects information from 80,000 randomly selected US households. The results depicted here represent responses from 3,003 survey participants interviewed from Nov. 3 – 15, 2017, and 3,007 participants interviewed from Oct. 1 – 13, 2013. The margin of error in both polls is +/- 1.8 percentage points. Truven Health is part of the IBM Watson Health™ business.This Truven Health Analytics PULSE® Healthcare Survey asked Americans about their experiences with primary care physicians. Respondents were asked identical questions to a study conducted in 2013, providing a comparison. Some of the topics covered include scheduling, device usagee, and time spent with patient. 

The full results can be viewed below.  

HEALTH POLL: Prescription Drugs

Every other month, the Truven Health Analytics®-NPR Health Poll surveys approximately 3,000 Americans to gauge attitudes and opinions on a wide range of healthcare issues. Poll results are reported by NPR on the health blog Shots (npr.org/sections/health-shots/) and on air. The Truven Health Analytics-NPR Health Poll is powered by the Truven Health Analytics PULSE® Healthcare Survey, an independently funded multi-modal (land line, cell phone, internet) survey that collects information from approximately 80,000 US households annually. The results depicted here represent responses from 3,003 survey participants interviewed from June 1 – 15, 2017. The margin of error is +/-1.8 percentage points

Given the amount of money spent on retail prescription drugs in the US ($324.6 billion in 2015 ), the Truven Health Analytics-NPR Health Poll asked Americans about their experiences with and attitudes toward prescription drugs and drug pricing. While 97% of respondents who received a prescription for medication in the last 90 days filled it, the most cited reason by respondents who did not fill their prescription was cost (67%), and 12% of all respondents said that cost drove them to purchase prescription medication outside the US. 

The full results can be viewed below.  

Reducing Overprescribing Through Peer Comparison Letters

Antipsychotic agents, such as quetiapine fumarate, are frequently overprescribed for indications not supported by clinical evidence, potentially causing harm.  In this randomized clinical trial, a peer comparison letter randomized across the 5055 highest Medicare prescribers of the antipsychotic quetiapine fumarate reduced prescribing for at least 2 years. Effects were larger than those observed in existing large-scale behavioral interventions, potentially because of the content of the peer comparison letter, which mentioned the potential for a review of prescribing activity.  This indicates that behavioral nudge interventions can raise the quality of prescribing, but research is still needed on how to most precisely target unsafe prescribing behavior.

The full article, entitled "Effect of Peer Comparison Letters for High-Volume Primary Care Prescribers of Quetiapine in Older and Disabled Adults: A Randomized Clinical Trial" can be viewed here.  

According to Accenture analysis, when combined, key clinical health AI applications can potentially create $150 billion in annual savings for the United States healthcare economy by 2026.

At hyper-speed, AI is re-wiring our modern conception of healthcare delivery. AI in health represents a collection of multiple technologies enabling machines to sense, comprehend, act and learn1, so they can perform administrative and clinical healthcare functions.

Unlike legacy technologies that are only algorithms / tools that complement a human, health AI today can truly augment human activity—taking over tasks that range from medical imaging to risk analysis to diagnosing health conditions.

Data Brief: Physician Perspectives on Access to Patient Data

What types of actionable patient intelligence are most important to physicians today? In October 2017, Surescripts commissioned a survey of 300 U.S. physicians to gather insights on their access to and need for patient data. We wanted to understand where they get data, which sources they trust and what data they find most valuable.

Key Findings

• Half of Physicians Say Data Access Could Be Much Better
• The Biggest Data Gap: Medication Adherence
• Price Transparency Aids Prescribing Decisions
• Lack of Patient History Data Impedes Care Coordination

The full data brief can be viewed below

Recognizing the investment that the health insurance industry will continue to place in these programs, a multi-stakeholder group representing patients, physicians, hospitals and pharmacists has developed the following principles on utilization management programs to reduce the negative impact they have on patients, providers and the health care system. This group strongly urges health plans, benefit managers and any other party conducting utilization management (“utilization review entities”), as well as accreditation organizations, to apply the following principles to utilization management programs for both medical and pharmacy benefits. They believe adherence to these principles will ensure that patients have timely access to treatment and reduce administrative costs to the health care system.