COVID-19 Acuity Score Identifies Outpatients at High or Low Risk of Adverse Outcomes
- Triage decisions are difficult for outpatients with mild symptoms of COVID-19, as many of them return later with critical illness
- To help clinicians plan follow-up care of such patients, Massachusetts General Hospital researchers developed the COVID-19 Acuity Score (CoVA), which predicts hospital admission, critical illness or death within seven days
- CoVA was prospectively validated on 2,205 patients at Mass General
- The scoring system is automatable and can be incorporated into electronic health records
Subscribe to the latest updates from Neuroscience Advances in Motion
For outpatients with COVID-19 who don't require immediate hospitalization for COVID-19, triage decisions are complicated by the disease's biphasic course. Some patients who initially present with mild symptoms must be hospitalized later and many become critically ill.
To help clinicians plan appropriate follow-up care for outpatients with COVID-19, researchers at Massachusetts General Hospital developed the COVID-19 Acuity Score (CoVA), which can be incorporated into an electronic health record system.
Shibani Mukerji, MD, PhD, associate director of the Neuro-Infectious Diseases Unit, physician Gregory K. Robbins, MD, of the Infectious Disease Division, and M. Brandon Westover, MD, PhD, a neurologist who directs the Clinical Data Animation Center at Mass General, and colleagues describe the scoring system in a paper posted on medRxiv, a preprint server.
Development of CoVA
Researchers studied 9,381 patients who sought urgent care for suspected or confirmed COVID-19 at the Mass General Emergency Department or one of the newly formed Mass General Respiratory Infection Clinics between March 7 and May 2, 2020.
98 variables were analyzed, including demographics, tobacco use history, body mass index (BMI), vital signs, SARS-CoV-2 test result, COVID-19 symptoms, pre-existing diagnoses and radiology findings. At the P = 0.05 significance level, 65 variables were carried forward into logistic regression modeling.
The final model provides acuity scores between 0 to 100. It showed excellent ability to predict hospitalization, critical illness (ICU admission and/or mechanical ventilation) or death within seven days in the development cohort.
Prospective Validation of CoVA
CoVA was tested on 2,205 adults seen at Mass General for suspected or confirmed COVID-19 between May 3 and May 14, 2020. It had excellent performance in predicting adverse outcomes generalized to that cohort.
Predictors of Adverse Outcomes
The model includes 26 variables that increase the probability of adverse outcomes. Besides a positive test for SARS-CoV-2, the top five variables are:
- Age (coefficient 0.7353)
- Respiratory rate (0.2746)
- History of acute ischemic stroke (0.1746)
- Multifocal signs on chest X-ray (diffuse opacities or "ground glass," 0.1293)
- Heart rate (0.1215)
Four variables are inversely correlated with the probability of adverse outcomes:
- Diastolic blood pressure (−0.4724)
- Blood oxygen saturation (−0.3776)
- Systolic blood pressure (−0.1151)
- BMI <18 kg/m2 (−0.0001)
The Role of Comorbidities
Besides ischemic stroke, multiple other neurological conditions were among the variables that robustly predicted adverse outcomes: a history of intracranial hemorrhage, subarachnoid hemorrhage, epilepsy, amyotrophic lateral sclerosis, myasthenia gravis or spinal muscular atrophy. It's unknown whether these diseases are simply markers of poor health or whether COVID-19 amplifies neurological pathology.
Hypertension and diabetes mellitus did not emerge as predictors in the CoVA model, but they were correlated with outcomes in univariate analysis. Both are strongly associated with variables that did become predictors: older age, higher Charlson Comorbidity Index and BMI >35 kg/m2.
view original journal article Subscription may be required
Visit the Department of Neurology
Learn more about neurology research at Mass General