Patient outcomes in enhanced ICUs have catalyzed the development of prediction algorithms designed to provide patient-specific alerts at the bedside, thus bringing significant advances to the positive intensive care future. Despite the potential, the journey has encountered obstacles, especially using retrospectively collected data, which often lead to performance metrics that do not reflect their real-world efficacy. The lack of models tested in the real world highlights the gap in efforts to improve patient care. In this context, Visig is a groundbreaking pattern recognition system designed to detect early signs of potential mortality in ICU patients. This innovative approach uses genetic algorithms to detect vital sign changes in decreased vital signs and will redefine patient management and care in an adult intensive care setting.
A landmark study conducted by Andrew Kramer of Prescient Healthcare Consulting, in collaboration with Dr. Marc Lafonte and Dr. Ibrahim El Husseini of Robert Wood Johnson-Barnabas University Hospital, Simon Didcote, Simon Didcote, Paula Maurer of Medical Deciestnection Network Dr. Frantz Hastrup and Dr. James Krinsley of Simon Didcote, Stanford Hospital, shed light on the effectiveness of Visig. Their important work was published in Medical Informatics Unlockhighlighting the role of the system in changing the care of ICU patients.
“Our aim is to leverage the predictive power of machine learning to enable clinicians to make the information they need to make informed decisions,” Dr. Kramer said.
Previous investigation (Intensive Care MedicineOctober 2013) verified the prediction accuracy of the algorithm-based Visig, whose scores are closely related to the increase in mortality. The current study uses a two-stage approach to assess the clinical utility of Visig. Initially, clinicians turned a blind eye to the system’s scores before introducing a stage in which these scores can be accessed through a user-friendly interface. “This study has been carefully designed to measure the impact of the system on clinical outcomes,” Dr. Kramer noted.
Visig’s predictive model relies on continuous monitoring of vital signs and mechanical ventilation status, resulting in a composite score with three levels of mortality risk. This score is easy to explain and is updated every 30 minutes to make it timely. “This approach allows us to clinically demonstrate evidence of patient decline and hopefully reduce unexpected harmful outcomes,” Dr. Kramer explained.
The results of this study were surprising, indicating significant improvements in patient care. “Incorporating Visig into clinical workflows may greatly improve outcomes, especially in reducing ICU residence time and duration of mechanical ventilation,” Dr. Kramer reported. The study also observed a significant decline in ICU readmissions, demonstrating Visig enhancement The ability to care for and promote long-term patient health immediately. The work of Dr. Andrew Kramer and his team demonstrates the benefits of incorporating machine learning tools such as Visig (such as Visig) into an intensive care setting. By providing clinicians with real-time insights into the patient’s condition, Visig supports informed clinical decision-making, which significantly improves patient outcomes.
Journal Reference
AA Kramer et al., “Prospective evaluation of machine learning-based clinical decision support system (VISIG) to reduce adverse consequences in adult critically ill patients,” Informatics in Medicine, 2024.
doi: https://doi.org/10.1016/j.imu.2023.101433.
About the Author
Dr. Andrew Kramer Over the past 22 years, he has been actively involved in intensive care research. He is a co-developer of disease systems with severity of Apache IV, Apache IVA, MPM-III and OASIS, resulting in over 100 predictive models used worldwide. In addition, he is the author of more than 80 manuscripts in high-influence journals, including two that have been cited more than 500 times. Dr. Cramer received his PhD in Human Genetics from the Virginia Medical College. He then received a postdoctoral fellowship in epidemiology. Dr. Kramer joined Cerner Corporation in 2003 and worked there until 2015, leading the company’s intensive care research efforts. In 2015, he left Cerner to establish a prescient healthcare consulting firm dedicated to providing novel analytical solutions in intensive care.
Dr. James Clinsley Graduated from Yale University and Cornell University School of Medicine. He completed his internal medicine training in pulmonary critical medicine training at New York University and Yale School of Medicine. He served as intensive care officer at Stanford Hospital from 1998 to 2020 and served as professor of clinical medicine at Columbia University’s Waglos School of Physicians and Surgeons. Since 2003, he has published extensively on many aspects of glucose control, as well as topics involving mechanical ventilation. A complete list of his publications can be found:
