Filter by input type
Select all
News
Pages
Events
Filter by category
Select all
AI ANALYTICS
Mobile Apps and Internet of Things
Advancement of science
big data
Connected communities
Coronavirus
Courses and training
DIAGNOSIS
Initial Editorial
Editorials
A world in the cloud
Events
Infographics
Artificial Intelligence and Science
IoT Apps
News
Digital platforms
Social networks
Review of scientific publications
Course Summary
Synopsis of essay
Overview of reference frames
Synopsis of recent publications
Use of Digital Platforms
Researchers use machine learning to optimize workflow in hospital admissions by. COVID-19.

Researchers at the Johns Hopkins University School of Medicine implemented a machine learning system to optimize hospital admission decisions for COVID-19.

The high demand for medical care during the pandemic affected emergency departments in the United States, which continued their work of distinguishing patients who need hospital care and those who can be discharged safely. In this way they had to incorporate new variables to identify patients with COVID-19. To this end, Johns Hopkins researchers presented a study, where they show the development, implementation and evaluation of an integrated clinical decision support system through electronic health records (EHR, in English).

This system uses data from EHR and machine learning to estimate the short-term risk of clinical deterioration in patients with confirmed or suspected cases of COVID-19. In this way, the system can determine the risk and need for critical care within 24 hours, as well as the specific care needs of hospitalized patients within 72 hours.

“Model performance and numerous patient-oriented outcomes, including in-hospital mortality, were measured across the study periods. The incidence of critical care needs within 24 h and hospital care needs within 72 h were 10.7 % and 22.5 %, respectively, and were similar in all study periods”, explains the study.

For its part, the machine learning model had a remarkable performance in all conditions since it obtained an AUC range from 0.85 to 0.91 in the prediction of critical care needs and from 0.80 to 0.90 for critical care needs. of hospital care.

However, total mortality remained unchanged during the study period, but achieved a reduction among high-risk patients after implementing the clinical decision support model.

The models were trained with 39 different predictor variables, such as shortness of breath, lactate levels, history of kidney disease, hypotension, among others.

“Our models also provide information that may inform clinical practice and the interpretation of similar models in the future. Diabetes, chronic lung disease, cardiovascular disease, and hypertension have been identified as important risk factors for severe illness and mortality from COVID-19, although these comorbidities were not among the most important predictors for any of our models." explains the study.

In these variables and other demographic variables, age and sex, had little weight in the models. However, despite seeming contradictory, in acute care settings the physiological variables and manifestations of the disease were more important in the clinical course than epidemiological risk factors.

Learn more about this study and its results at the following link:

https://www.nature.com/articles/s41746-022-00646-1

Outstanding news

News by country

Related Content

Secured By miniOrange