A new scoring system can predict whether COVID-19 patients will require invasive mechanical ventilation, researchers report.
The score uses three variables to predict future risk: heart rate; the ratio of oxygen saturation (SpO2) to fraction of inspired oxygen (FiO2); and a positive troponin I level.
“What excites us is it’s a really benign tool,” said Muhtadi Alnababteh, MD, from the Medstar Washington (D.C.) Hospital Center. “For the first two variables you only need to look at vital signs, no labs or invasive diagnostics.”
“The third part is a simple lab, which is performed universally and can be done in any hospital,” he told this news organization. “We know that even rural hospitals can do this.”
For their retrospective analysis, Dr. Alnababteh and his colleagues assessed 265 adults with confirmed COVID-19 infection who were admitted to a single tertiary care center in March and April. They looked at demographic characteristics, lab results, and clinical and outcome information.
Ultimately, 54 of these patients required invasive mechanical ventilation.
On multiple-regression analysis, the researchers determined that three variables independently predicted the need for invasive mechanical ventilation.
Calibration of the model was good (Hosmer–Lemeshow score, 6.3; P = .39), as was predictive ability (area under the curve, 0.80).
The risk for invasive mechanical ventilation increased as the number of positive variables increased (P < .001), from 15.4% for those with one positive variable, to 29.0% for those with two, to 60.5% for those with three positive variables.
The team established cutoff points for each variable and developed a points-based scoring system to predict risk.
It was an initial surprise that troponin – a– would be a risk factor. “Originally, we thought COVID-19 only affects the lung,” Dr. Alnababteh explained during his presentation at CHEST 2020. Later studies, however, showed it can cause symptoms.
The case for looking at cardiac markers was made whenof young athletes who recovered from COVID-19 after experiencing mild or no symptoms showed that 15% had signs of myocarditis on .
“If mild COVID disease in young patients caused cardiac injury, you can imagine what it can do to older patients with severe disease,” Alnababteh said.
This tool will help triage patients who are not sick enough for the ICU but are known to be at high risk for ventilation. “It’s one of the biggest decisions you have to make: Where do you send your patient? This score helps determine that,” he said.
The researchers are now working to validate the score and evaluate how it performs, he reported.
Existing scores evaluated for COVID-19 outcome prediction
The MuLBSTA score can also be used to predict outcomes in patients with COVID-19.
A retrospective evaluation of 163 patients was presented at CHEST 2020 by Jurgena Tusha, MD, from Wayne State University in Detroit.
Patients who survived their illness had a mean MuLBSTA score of 8.67, whereas patients who died had a mean score of 13.60.
The score “correlated significantly with mortality, ventilator support, and length of stay, which may be used to provide guidance to screen patients and make further clinical decisions,” Dr. Tusha said in a press release.
“Further studies are required to validate this study in larger patient cohorts,” she added.
The three-variable scoring system is easier to use than the MuLBSTA, and more specific, said Dr. Alnababteh.
“The main difference between our study and the MuLBSTA study is that we came up with a novel score for COVID-19 patients,” he said. “Our study score doesn’t require chest x-rays or blood cultures, and the outcome is need for invasive mechanical ventilation, not mortality.”
A version of this article originally appeared on.