In the News: May 21–28
Population Health Benefits From Robust Analytics
A recent Healthcare Finance article discusses how the University of Texas Medical Branch (UTMB) leverages robust analytics and real-time, dynamic reporting capabilities to better patient care and increase reimbursement dollars. Health care organizations involved in at-risk contracting and pay-for-performance arrangements need accurate and timely data that allows providers to look at patient health in a more holistic manner and drill down into specific preventive services.
UTMB established a proprietary application that generates on-demand reports where staff can review potential action steps, for example, planning for long-term and short-term solutions to diabetes care. In 2018, UTMB realized 85.4% of its performance dollars; after implementing the application, that number grew to 91.7%.
Organizations involved in population health must also be concerned with social determinants of health. Investing in partnerships and programs that address the social, environmental and behavioral needs of patients outside the traditional medical setting improves their outcomes. Yet determining the extent to which provider systems can or should contribute to this arena is a complex undertaking. To learn more about ways in which organizations can participate in social determinants of health initiatives, please see the Sg2 report Social Determinants of Health: Stitching Together Solutions, which aims to help organizations weigh their options by outlining a spectrum of system involvement in this space and providing guidance on keeping sustainability at the core of any work—no matter how your organization joins in.
Tumor Detection Improves Through the Use of AI
A recent Wall Street Journal article discusses work being done across 30 health care organizations to jointly train an artificial intelligence (AI) system to aid radiologists in spotting brain tumor boundaries, using a method known as federated learning. Federated learning enables AI algorithms to train on data where they reside, rather than transporting data to a central location. Each participating collaborator trains an identical AI model on their localized data, after which the individual learnings from each partner location are combined, and a master model is developed. The information exchange for this process was facilitated by Intel Corp and Penn Medicine’s software platforms.
According to the American Brain Tumor Association, an estimated 87,000 individuals across the US will be diagnosed with a tumor originating in the brain in 2020. If successful, this project will aid in identifying the boundaries of developing brain tumors, which can then aid in accurate diagnosis and determining appropriate course of treatment. Typical AI models trained on a single health system platform had 70% accuracy, while a model trained with the federated technique had an accuracy of 85.7%. An additional benefit to federated learning is the ability to bypass regulatory and privacy concerns that are typically associated with data sharing.
AI-based solutions are rapidly integrating into the care regimen for complex health care challenges. Oncology has been ahead of the curve when it comes to leveraging AI to enable early and precise cancer detection. To learn how to integrate AI into your service line strategies at a time when AI has become one of the most widely discussed technologies in the industry, please read Sg2’s Expert Insight AI Is Here to Stay: Adapting AI for Service Line Strategy.
Half of Americans Delayed Care Due to COVID-19
A recent Kaiser Health News article called out a Kaiser Family Foundation poll finding that 48% of Americans did not seek medical care because of the COVID-19 pandemic, with 11% of those also responding that the delayed care resulted in a worsened health condition.
In March, as the pandemic ramped up, many hospitals, health systems and providers postponed or reduced nonessential services to prepare for COVID-19 patients. Medical groups also noted a nationwide drop-off in emergency patients, perhaps due to concerns of coronavirus exposure. However, of those polled, 7 out of 10 plan to seek care in the next few months.
Moving forward, COVID-19 has the potential to transform health care consumer purchasing patterns, and Sg2 recommends viewing traditional consumer engagement through a new lens. Health systems must rely on but also adapt the 4 consumer strategy mainstays, while amplifying patient safety. Read the Sg2 Expert Insight Engaging the (Post–COVID-19) Health Care Consumer for more on each of the 4 mainstays.
Tags: artificial intelligence, cancer detection, consumer strategy, coronavirus, COVID-19, data analytics, health outcomes, oncology, population health, reactivation, social determinants of health, tumor detection