OR WAIT null SECS
Ad: Extracting value from healthcare requires a strong data foundation.
Image has been altered. Licensed from WavebreakmediaMicro - Fotolia
In healthcare, conversations about big data sometimes seem ambiguous. What exactly do innovators mean when they say that sprawling data sets can yield insights that improve patient care and greater health system efficiencies? Too often, the lofty ideals associated with big data and analytics outshine the ideas, implementations and results.
But groundbreaking big data projects are making noteworthy strides across healthcare. The results vary, improving everything from interoperable apps and patient engagement to population health management and robotics programs once considered the stuff of science fiction.
Two elements connect these distinct demonstrations of big data done right. First, each effort exists as a means to extract value from healthcare data. Second, they all depend on strong data infrastructure for clean, interoperable data and the speed required to get these effective solutions into the hands of physicians and other clinicians.
Here are four different looks at how data infrastructure breeds healthcare innovation.
California’s largest nonprofit health data network, Manifest MedEx, provides health information that informs the care of more than 10 million patients. To get the full view of the patient, the health information exchange can’t rely on a single source of data. Instead, Manifest MedEx draws information from 11 million claims records and more than 5 million clinical records.
After aggregating and cleansing that vast amount of data, the organization then leverages the information in high-power tools such as real-time alerts, longitudinal patient records and population health management platforms. For instance, customized event notifications identify only the insights that help one accountable care organization understand the details of each hospital discharge within its network. As a result, the client has cut hospital readmissions and costs.
If interoperability is the destination, Fast Healthcare Interoperability Resources (FHIR) is the vehicle. FHIR, of course, is a standard that enables the electronic exchange of healthcare data among different stakeholders without suffering a loss of information integrity. Take this tool a step further and you get SMART on FHIR, which fuels the development of interoperable apps that use structural clinical data to enhance clinical decision support, care coordination and more.
One example: Boston Children’s Hospital developed a pediatric growth chart app using SMART on FHIR. The app visualizes data on pediatric height and weight measurements and superimposes the information over typical ranges, all in the electronic health record. This data visualization better prepares physicians to discuss out-of-range body weights — a factor that affects outcomes — with patients’ parents.
What’s most unique, however, is that the SMART on FHIR specification allows this app to be used in any health system that follows the same data standards.
When one large New York-based health system would cold-call high-risk patients with diabetes, the results were underwhelming: a 15% patient engagement rate. It was a burden on the health system and patients alike, and neither party benefited from better outcomes.
Instead, the health system began leveraging data to notify clinicians and case managers when a high-risk patient with diabetes entered the health system. If the patient were in the emergency department, a case manager could hustle over there to engage the individual in person. All of a sudden, thanks to actionable data and real-time alerts, the health system’s patient engagement rate climbed to 75%.
In healthcare robotics, perhaps the most interesting use of AI is not in artificial limbs or assisted surgery but physiotherapy. Ayanna Howard, Ph.D., founder of Zyrobotics, builds AI robots that help pediatric patients with developmental disabilities achieve behavior change. The tech runs physiotherapy sessions, earning trust and attention, simulating emotions and using gamification to help children improve their movement and motor skills.
To become more effective, the algorithms must feed on a great deal of data. Information is also critical to the validation of outcomes, measurement of movements and comparison of baselines. But the data foundation yields strong results: In a study, Howard found no difference between how kids with cerebral palsy and typically developing children move.
All of these use cases show that innovators are capable of moving clinical practice and the healthcare system to the next level. But each step forward is possible only because of underlying data infrastructure such as InterSystems IRIS for Health, the first data platform engineered to pull value from health data.
Interoperability and data normalization must be standard if healthcare innovators are to develop the apps that will unlock value. Without data platforms such as InterSystems IRIS for Health, talk of better patient outcomes and greater efficiencies is just that: talk.
Learn how InterSystems IRIS for Health empowers healthcare organizations to extract value from data.