With the proliferation of electronic health records follows a surge in the quantity of big data. The healthcare industry is still figuring out how to most effectively harness this data and use it to progress.
Thought leaders from institutions including Brigham and Women's Hospital in Boston, Johns Hopkins University in Baltimore, University of California in San Diego and Oakland, Calif.-based Kaiser Permanente offer the following six scenarios in which big data could help reduce costs and improve care. Their report is published in Health Affairs.
High-cost patients. The authors suggest using predictive analytics to more closely manage high-cost patients. While this is not a new strategy, the researchers mention analytics need to identify low-risk patients as well to stratify interventions and reallocate resources for both high- and low-risk patients. Predictive analytics should also renew focus on behavioral health problems as well as other issues, using an algorithm that may be more practical instead of accurate since it takes into consideration barriers to care not necessarily included in the EHR.
Readmissions. Nearly one-third of readmissions can be prevented, and predictive analytics and algorithms may help reduce that number, the authors propose. They suggest wearables and allowing providers access to smartphone data to track communication habits could provide important data to help inform healthcare decisions.
Triage. Big data analytics can also be useful in situations regarding patients who aren't continuously monitored, such as those coming to the emergency room. The authors suggest integrating a triage algorithm into clinical work flow to produce real-time risk estimates, allowing providers to anticipate any potential needs or complications.
Decompensation. Analytics can use multiple data sources to help detect decompensation, when a patient's condition worsens. Physiological data — such as electrocardiographic and oxygen monitoring — can be indicative of a patients' movement, activity and, therefore, decompensation. Monitors collecting this information submit the data to a server where real-time analytics can help determine the likelihood of decompensation.
Adverse events. The authors suggest big data analytics can help predict which patients are at risk of a number of adverse events, including renal failure, infection and adverse drug events. In each instance, monitors are collecting real-time data to monitor the onset and/or development of such events.
Diseases affecting multiple organ systems. For patients with diseases affecting more than one organ system, accurately predicting the trajectory of the condition could help providers tailor therapies to patients to optimize treatment.
More Articles on Big Data:
Ready for Pickup: Physician-Prescribed mHealth Apps
How to Succeed in Big Data Without Really Trying: 4 Steps From Booz Allen
Survey: In Analytics, Data Variety More Challenging Than Volume