Getty Images

New Model Helps Predict COVID-19 Patient Deterioration

A new model was able to predict deterioration among COVID-19 patients, which occurred in 10 to 20 percent of those admitted to the hospital, a new study shows.

Throughout the COVID-19 pandemic, providers realized that predicting clinical deterioration among patients is an important step to take, a new study published in The BMJ shows.

During the pandemic, hospitals have struggled to maintain the capacity and resources needed to care for COVID-19 patients. Predicting clinical deterioration among patients can help hospitals prepare for future surges.

The study included 33,119 adult patients admitted to a hospital with respiratory distress or COVID-19.

Researchers attempted to measure the relationship between clinical deterioration in the first five days of hospitalization using various linear models. They defined clinical deterioration as death within the hospital or treatments such as mechanical ventilation, heated high flow nasal cannula, or intravenous vasopressors.

Researchers based the linear models on nine clinical and personal variables selected from a list of 2,686. The various cohorts constructed included development, internal validation, and external validation. The researchers created multiple groups to obtain insightful data by comparing the results each provided and the relationships between them.

The development cohort consisted of Michigan Medicine residents focusing on hospital admissions rather than patients since one patient could have multiple visits. This cohort focused on respiratory disease. However, cases involving a patient’s recovery before or during supplemental oxygen were excluded.  

The internal validation cohort also consisted of Michigan Medicine residents, but it focused on patients who used supplemental oxygen and a positive COVID-19 test.

The external validation cohort focused on patients from 12 external medical centers who used supplemental oxygen and had COVID-19.

Results were generally consistent among all three cohorts, with between 10.7 and 21.6 percent of patients experiencing clinical deterioration.

Of the 24,419 patients, 3,757 experienced deterioration within the development cohort. In the internal validation cohort, 206 of 887 patients experienced deterioration, and 1,304 of 8,335 hospital admissions experienced the same among the external validation cohort.

From these results, researchers concluded that not only is clinical deterioration dense among COVID-19 patients, but this information also provides clinicians with reliable information as to what conditions warrant hospitalization.

"[The] model to predict clinical deterioration…was applied externally without the sharing of data, and performed well across multiple medical centers, patient subgroups, and time periods, showing its potential as a tool for use in optimizing healthcare resources," researchers concluded.

The model can also help hospitals optimize care and provide all patients with deterioration-preventing resources.

Monitoring patient deterioration has previously been proven useful, especially within hospitals and emergency departments during the COVID-19 pandemic. For example, a previous study from the earlier stages of the pandemic revealed that X-rays and imaging data could be used to predict COVID-19 patients at risk of deterioration.

Next Steps

Dig Deeper on Artificial intelligence in healthcare

xtelligent Health IT and EHR
xtelligent Healthtech Security
xtelligent Healthcare Payers
xtelligent Pharma Life Sciences
Close