Site Maintenance October 29 – November 5 SCCM is currently performing major updates to its digital infrastructure to better serve members. Any website that requires an SCCM Customer ID and password to log in is currently unavailable until November 5, 2024. SCCM Customer Service and staff have limited capabilities, and response times will be longer during this time.
Big Data Leads to the Discovery Data Science Campaign
07/20/2023
The Discovery Data Science Campaign was launched in 2022 to improve the care of critically ill patients by leveraging the use of large-scale data (big data) for research. The campaign’s goal is to apply findings from data analysis in a clinical environment through standardized data models and shared resources, starting with the 2023 Datathon.
More than three years after the start of the COVID-19 pandemic, Karin Reuter-Rice, PhD, NP, FAAN, FCCM, associate professor at Duke University School of Nursing and School of Medicine in Durham, North Carolina, USA, believes that the public has a far greater appreciation and understanding of the role critical care professionals play in healthcare. “The global pandemic directly aligned critical care medicine with population health,” said Dr. Reuter-Rice. “While there’s been a historic appreciation for critical care, I think COVID-19 highlighted the importance of critical care as a public health partner.”
Partnership is a major focus for Dr. Reuter-Rice, who is a cochair of the Data Science Campaign of the Society of Critical Care Medicine’s (SCCM) Discovery, the Critical Care Research Network. The Data Science Campaign was launched in 2022 to improve the care of critically ill patients by leveraging the use of large-scale data (big data) for research. The campaign’s goal is to apply findings from data analysis in a clinical environment through standardized data models and shared resources. The Data Science Campaign is only one of Discovery’s collaborative research efforts to improve outcomes for critically ill and injured patients.
Critical care is the perfect field to benefit from big data analysis, said J. Perren Cobb, MD, FACS, FCCM, a cochair of the Data Science Campaign and chief of surgical critical care services in the Critical Care Institute at Keck Hospital of University of Southern California, in Los Angeles, California, USA. “Critical care medicine is the most data-rich environment in medicine,” said Dr. Cobb. “The amount of data and information generated on each patient each day is beyond the capacity of the human mind to analyze.”
Dr. Cobb explained that one of the challenges critical care professionals have had to navigate is simply aggregating all of that information in one place. “Now that we’re in the era of big data and artificial intelligence, these challenges are finally being solved and the promise of these analytical approaches can be realized.”
The Data Science Campaign is focused on three domains: data harmonization and sharing, establishing a datahub, and hosting an annual datathon. Data harmonization and sharing are designed to establish principles for data standardization and developing common data elements. Dr. Reuter-Rice hopes that this effort will allow data to be collected and analyzed from a consortium of hospitals and medical centers. “The biggest challenge we have is data heterogeneity, yet we have some very homogeneous diagnoses,” Dr. Reuter-Rice said. “We know there are specific criteria that define sepsis, acute respiratory distress syndrome, and other diagnoses. However, the data collected specific to those diagnoses can be very heterogeneous, which makes it difficult for clinicians to aggregate and translate data. These data challenges interfere with timely development of novel interventions that improve patient outcomes.”
Homogenous data collection facilitates analysis. “Data harmonization can also help to level the playing field between small hospitals and larger health systems,” Dr. Cobb said. Until now, hospitals have been on their own to aggregate and analyze their own data, and many smaller hospitals have not had the resources to commit to analysis, much less implementation. The Data Science Campaign could change this. “In the current era of critical care data science, data from small, medium, and larger hospitals and systems can be aggregated—and analytical resources pooled—to not only do better science, but to make sure the lessons learned are implemented everywhere,” he emphasized.
Where those data are housed is the cornerstone of the Data Science Campaign’s second domain of establishing a datahub—low-cost, robust data storage, security, and ready access for Discovery investigators. Jerry J. Zimmerman, MD, PhD, MCCM, chair of Discovery, explained that the datahub would allow investigators to continue using Discovery’s clinician and data coordination capabilities. “Discovery aims to facilitate conduct of high-quality, multicenter clinical research studies designed to improve patient-centered, clinically meaningful outcomes among critically ill patients,” he said. “It is logical that Discovery should provide leadership in this arena.”
Dr. Zimmerman said the potential of the Data Science Campaign is immense. “The last paradigm shift in medical sciences occurred with the AIDS epidemic and subsequent rapid expansion of molecular biology and genetics,” he noted. “Currently we are witnessing an explosion of data science and its obvious potential in medicine.” The Data Science Campaign’s leadership hopes that some of that potential will be in evidence at this year’s annual Datathon, to be held August 5-6, 2023 in New York City, where clinicians and data scientists will collaborate to address real-world problems using existing datasets.
The 2023 Datathon will focus on data science in critical care, with a specific emphasis on COVID-19 and health equity. The goal is to identify projects that have the potential to lead to publication and improve the care of critically ill patients. “The Datathon will provide proof of feasibility for analytical teams to demonstrate the potential of big data and artificial intelligence,” Dr. Cobb said.
While the Datathon is geared toward clinicians interested in data science, a basic understanding of the field is something all critical care professionals will need in the future, Dr. Reuter-Rice said. “I don’t think you can come into the ICU anymore and not be completely overwhelmed and bombarded by data,” she said. “You have to have some basic understanding that you’re in an environment where a lot of data are being collected; some of those data will drive quality improvement for the health system, for the unit, and also for your patients.”
Dr. Cobb agreed. “All clinicians, from trainees to professors, will need to learn about data science and artificial intelligence, including their potential for good as well as caveats, pitfalls, and risks,” he said. “The Society is embracing its responsibility to lead in this education and training of critical care professionals.”