Digitizing, collecting, and storing large quantities of health-related data (often referred to as “big data”) has created new and exciting opportunities for care management professionals. Extracting knowledge from such large quantities of health-related data (often referred to as “data science”) that are captured in medical records, administrative claims, and many other disparate sources, may significantly enhance a care manager's ability to understand the biological, genetic, and behavioral underpinnings of a person's health.
Specifically, care management professionals (care managers and care coordinators) of the future may be able to use new strategies for collecting and analyzing big data to advance our understanding of who is most likely to experience adverse symptoms (i.e., pain and fatigue) and injury (i.e., inflammation); examine behaviors for maintaining wellness and preventing chronic illness (i.e., obesity, diabetes); and develop personalized health strategies that account for individual variation in genes, environment, and lifestyles.
Most care managers and coordinators however have been traditionally trained as clinicians, and have limited statistical knowledge and ability to interpret findings from big data.1 Thus, there is a need to discuss how care managers can increase their knowledge about data science and its implications for discovering new nursing knowledge.
This article is the first of a multi-part series on data and care management. In this article we define big data and data science, and identify two emerging and priority areas of care management using big data and data science - symptom science and precision medicine. In future posts we will discuss the importance of big data and data science on care management, and present a roadmap (and discuss current roadblocks) for advancing the integration of data insights into the care management workflow.
By definition, big data is large, disparate sources of data which may be unstructured and too complex to store, manage, and/or analyze using standard computer systems and conventional database techniques. Big data is often characterized by five dimensions: volume, velocity, variety, veracity, and value.
The ability to extract knowledge from big data to solve health-related problems has been commonly referred to asdata science– many techniques and technologies are used in this process to discover patterns and determine meaning from large, diverse, and complex datasets2. Data scientists, who are inherently are interdisciplinary, possess a firm understanding of statistics/mathematics, computer science/informatics, and biomedical science.
Symptom science – the systematic investigation of patients’ experience, sequelae, and management of symptoms – is a key theme at Health and Human Services (HHS). One of the central questions facing care managers today is how can we improve both the management of adverse symptoms for those with chronic illness, and as well as our patients’ quality of life. To support research focused on the biological and behavioral underpinnings of symptoms, the U.S. National Institute of Nursing Research (NINR) has funded P30 Centers of Excellence nationwide.12
In a series of ambitious studies of fatigue, researchers at the Virginia Commonwealth School of Nursing Center for BioBehavioral Clinical Research used data elements such as fatigue, stress, depressive symptoms, and cytokine measures, which were common across patient conditions, to enable merging of multiple datasets, thus facilitating a cross-study analysis of biobehavioral phenomena related to fatigue in multiple populations.13 This dataset includes a variety of common data elements, not available elsewhere, in a large representative sample of women with relatively rare medical conditions (e.g., fibromyalgia and/or sickle cell disease). The intent that was fully integrated by design was that primary data might yield substantially more clinical information than a series of post-hoc studies focused on a single condition. This project now provides cost-effective opportunities for researchers to test hypotheses that would otherwise be too expensive, lengthy, and difficult to carry out.
Presently, the vast majority of treatment strategies utilize a “one-size-fits-all” approach. In other words, treatment outcomes are quite successful in many cases while many others are not. Precision medicine, a “targeted approach”, attempts to improve diagnostic precision with which patients are categorized and treated by taking into account their genes, environments, family history, and lifestyles. Thus, precision medicine is based on targeting genetic alterations associated with particular diseases; even for diseases that have been found to be incurable at the present time, precision medicine might help to accelerate discoveries and inform the development of personalized approaches to treatment11.
With the proliferation of individual patient data at the molecular, tissue, and behavioral levels, precise diagnosis and targeted therapies are achievable. The US Federal Drug Administration defines precision medicine as “the tailoring of medical treatment to the individual characteristics, needs, and preferences of a patient during all stages of care, including prevention, diagnosis, treatment, and follow-up.” Analytic strategies are being leveraged to mine patient “omics” data for discovery of new biological associations, with goal of enabling early detection and improving diagnostic clarification to inform treatment.14 Specifically, an individual patient’s genome provides a molecular fingerprint which can be used to assess disease risk, diagnose disease, and monitor treatment effectiveness.
Wearable technologies and other mobile devices generate large amounts of data on behavior, biological processes, and the environment. Currently, patients are able to generate and transmit their own data to health systems using digital health tools such as Apple’s HealthKit and Google Fit. These mobile devices, which are linked to EHRs, allow for real-time monitoring of clinical parameters such glucose, blood pressure, and cardiac rhythm. Health tracking via mobile devices will allow nurse scientists to better understand disease mechanisms, which will inform the development of better risk assessment tools, and more accurately predict clinical outcomes.
Further, care managers and care coordinators can integrate the content and timing of clinical and behavioral interventions in ways that anticipate the changing needs of individual patients, making healthcare a more personalized, integrated, and accessible to those from different demographic and socioeconomic backgrounds.23 Analysis of associations between daily behavior, genetic profiles, and clinical outcomes are only recently beginning11, but with the advent of wearable technologies and other mobile devices, will continue the growth of precision medicine and integrated care.