How Technology Integration Supports Precision Medicine

Tufts Medicine in Burlington, Mass., is one health system deploying the IT infrastructure necessary to personalize care. The organization built a digital health ecosystem in an Amazon Web Services-hosted environment. The AWS cloud platform provides security, reliability and easier access to data across the system, creating a seamless user experience.

Chief of Digital Modernization Jeremy Marut says a public cloud is the foundation of Tufts Medicine’s digital health ecosystem, integrated with 40 applications and with the ability to pull data from diverse segments such as dietary, badging and visitor management apps. Clinicians can use personalized data from many sources, including wearables and chronic disease-specific management apps.

The money saved by deploying a public cloud could be redirected to invest in artificial intelligence (AI) and machine learning (ML) tools to further the development of personalized medical treatments, Marut adds. For instance, the savings will help Tufts Medicine reduce sepsis and ensure that patients receive the right treatments.

READ MORE: Why should healthcare organizations hire a data quality manager?

For ML purposes, Tufts Medicine turned to Amazon HealthLake to democratize access to health data, according to Marut. HealthLake stores, transforms, queries and analyzes health information within minutes.

“We are taking those algorithms that are feeding that lake and opening it up to a greater quantity of users that are trained to care for patients and improve their lives,” Marut says.

In addition, Amazon Lex-powered chatbots are connecting patients to more personalized care.

Marut describes a typical physician’s approach to using ML data: “Now we’re taking all of this information that was really for an organization to get the highest levels of insurance reimbursement and personalizing the care that we give to this person because of the work that was exposed by all of these modern tools.”

Still, the journey toward more personalized healthcare is just getting started. “We’ve got things to stick to your arm, we’ve got AI, Apple Watch devices. We’ve got all these things that send data, but we’re not using it to create a course of action, a journey, for this particular person yet,” Marut says.

Integration and Automation Drive Precision Medicine

Precision medicine has evolved the way clinicians are treating patients with serious conditions compared with what was standard practice 20 years ago.

Through its private-public partnership with Oracle Cerner as part of the Tiger Institute for Health Innovation, University of Missouri Health Care launched a fully automated genomic testing interface within its electronic health record platform in June 2022. MU Health Care accelerated patient testing and improved patient safety and outcomes, says Dr. Richard Hammer, vice chair of clinical affairs in the pathology and anatomical sciences department at the University of Missouri School of Medicine.

Integrating the testing interface in the EHR was an improvement over manually sending genomic testing orders and scanning the results into a folder, according to Hammer.

“The reports usually would come back by fax, which then did not come directly to the clinicians but required uploading to medical records, where they were then scanned into the EHR,” he says. “That took some time.”

WATCH: Find out how Providence enhances patient care with the cloud.

Using automation has sped the process of ordering tests, sending them out and getting results back to clinicians so they can decide on treatments. “We’ve increased the efficiency and the timeliness of the results back to the clinician,” Hammer says.

MU Health Care has more than 400 active interfaces with multiple systems, says interim CIO Deb Dow. The health system used Infor’s Cloverleaf Integration Engine and built standard HL7 interfaces between the genomic testing location, Foundation Medicine (which provides the genomic testing) and MU Health Care, she adds. This allowed the organizations to share orders and results. After a few weeks of training, MU Health Care staff made the new integration part of regular workflow.

By spring 2023, MU Health Care plans to start integrating precision medicine data as structured data rather than in a PDF, making it easier for physicians to find the data. In a structured format, physicians can search the data in a defined field, Hammer says.

Precision Health Can Be Applied to All Disciplines

The University of California, Irvine, has integrated its technology to create targeted health and wellness strategies for patients. In February 2022, the university launched the Institute for Precision Health, an interdisciplinary campuswide initiative to develop solutions in precision medicine. A future brick-and-mortar home for IPH will let data scientists collaborate with clinicians to develop analytics tools.

“Precision health is truly an amalgamation of multiple critical pillars. Understanding the multidisciplinary nature of this work is really how IPH came about,” says Dr. Peter Chang, ­assistant professor in residence at UCI’s Department of Radiological Sciences. He formed IPH along with co-director Leslie Thompson, a professor in the school of medicine and the school of biological science.

IPH’s launch was driven by a need to make sense of health data. It linked EHRs, genomics and other medical data with ML’s analytic power, Chang says.

UCI has made various investments in precision health analytics, including recently replicating its EHR system in Syntropy, a data management and analytics platform hosted in the AWS cloud. “It essentially gives us the flexibility and power to look at all of our data using sophisticated analytics in a very efficient way,” he says.

EXPLORE: How technology helps identify and track social determinants of health data.

Other data computing environments on the UCI campus borrow architecture developed at major technology companies, such as the Kubernetes orchestration engine that originated at Google, Chang adds. Combined with containerization technology, Kubernetes allows Chang’s team to easily orchestrate AI software development on a flexible, hybrid, on-premises, cloud-based computing cluster. The system is able to automatically scale computing resources, including specialized NVIDIA GPUs that his team leverages to train algorithms on high-dimensional data such as images, waveforms and videos. “Taking advantage of a hardware and software stack optimized by major AI companies allows our team to maximize our investment in computing resources and focus on our goal of realizing precision medicine,” Chang says.

As a radiologist, Chang relies on predictive modeling and data analysis for personalized treatment and health maintenance. “Whether in a research or clinical environment, the same key technology, such as NVIDIA hardware or cloud-native services, is driving AI innovation,” he says.

Chang believes healthcare is still in the early stages of precision medicine and making sense of analyzing the data in a sophisticated way. But the future means linking multiple types of data and using ML to make sense of it, he says.