University medical centers could transform patient care by adopting the principles of the principles of learning health systems, according to researchers from Weill Cornell Medicine and the University of California in San Diego. In this approach, information from electronic health files, clinical trials and daily hospital operations are analyzed in real time to discover ideas that continuously improve patient care.
THE perspectivePublished on June 17 in NPJ Health Systems, reasons why a more intelligent, more efficient and equitable care model can be created by exploiting existing data to support learning at the system scale. However, the adoption of this model remains limited.
Dr Peter Steel
“The integration of various databases is part of the creation of a dynamic health care system,” said the main author Dr Peter SteelAssociate Professor of Clinical Emergency Medicine at Weill Cornell Medicine and Emergency Medicine Doctor at New York-Presbyterian / Weill Cornell Medical Center. “Practitioners will be able to see more easily and more quickly what works and what does not work; And what results in unnecessary costs. “
Also, contributing to the perspective Dr Robert HarringtonThe dean of Stephen and Suzanne Weiss by Weill Cornell Medicine, and Dr Christopher Longhurst And Dr Gabriel WardiBoth of the University of California, San Diego.
The authors say that the creation of a learning health system is particularly important now, because university establishments face financial tensions caused by the increase in research costs, the drop in margins and the increasing expectations of patients. The perspective is a call for action so that academic health centers make systemic changes by rethinking the way in which they generate and apply knowledge.
Obstacles to implementation
The idea behind this approach is not new – medical researchers first considered learning health systems when hospitals have passed from paper to electronic health files. However, electronic health records were mainly designed for the convenience of clinicians and patients, rather than researchers and quality improvement initiatives. Data Siloes also complicates the establishment of learning health systems. Information – patient stories, laboratory results, imaging or billing records – is stored in distinct and disconnected systems that do not communicate with each other.
Therefore, he can often take years to collect and analyze the data necessary to improve patient care, noted the authors. A functional learning health system could shrink this period to weeks while maintaining ethical research and centered on the patient and using solid safety systems to ensure patient confidentiality. This information can then be used to revise processing directives, improve patient safety and stimulate innovations.
Dr Robert Harrington
Part of the question, according to the authors, is insufficient integration between clinical care, research and education. Ideally, future doctors could learn to use data to ask and respond effectively to clinical questions that will bring together different stakeholders to collaborate.
“A learning health system, powered by AI, has the potential to raise clinical care and results,” said Dr. Harrington. “When we allow future clinicians to learn from each clinical meeting, we can improve quality and efficiency in a way that we could not before.”
Beyond the organization and analysis of data, the cost of implementing a learning health system can reach tens of millions of dollars. But long -term performance can be strategic: years after implementation, health care organizations that successfully use this approach could become much more competitive than those who do not, said the authors.
Artificial intelligence leads to a change
Despite challenges, recent progress in artificial intelligence makes the adoption of learning health systems more critical. Patients are starting to expect the doctors to use AI to offer personalized and proactive care, but AI depends on clean, well -structured and real data. “The AI can only keep its promise if it is built on the basis of learning infrastructure,” said Dr. Steel.
AI tools can quickly analyze medical data volumes, helping doctors identifying early alert signs of illness, rationalizing operations and making faster and more individualized decisions. A learning health system allows essential quality control, ensuring that AI tools are permanently monitored for safety, bias and efficiency.
“Academic medical centers are faced with a rapidly evolving financing landscape, even if the costs of technological transformation and administration in health care continues to increase,” said Dr. Steel. “The implementation of the learning health system is no longer a theoretical objective, but a strategic imperative.”