Viral pneumonia severity score predicts outcomes in COVID-19 pneumonia
Presenter: Jurgena Tusha, MD
The recently developed MuLBSTA score, which incorporates six clinical parameters, predicts disease severity and overall outcomes in patients with COVID-19 pneumonia.
In a retrospective study conducted at a community hospital in Michigan,1 a significant positive correlation was found between the MuLBSTA score and death in hospitalized patients with COVID-19 pneumonia, according to data presented by Jurgena Tusha, MD, from Wayne State University, Detroit.
An influx of SARS-CoV-2 infection has led to several unanswered questions. “One question raised was how to risk-stratify these patients in order to direct further management,” she said. “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.”
The MuLBSTA score was developed in China and is designed to predict 90-day mortality in patients with viral pneumonia. It is based on six factors: multilobe infiltrate, which is weighted the most (5 points); absolute lymphocyte count ≤ 0.8 x 109/L (4 points); bacterial coinfection as detected by sputum or blood culture (4 points); smoking history (2 points for prior smoker, 3 points for active smoker); history of hypertension (2 points); and age ≥ 60 (2 points). The maximum score is 20.
For the study, 163 charts from hospitalized patients with COVID-19 pneumonia from March 15 to April 10, 2020 were reviewed manually. Length of stay outside the intensive care unit (ICU) was 6.46 days and ICU length of stay was 15.5 days. Median age of the 55 patients in the ICU was 68 years. Mechanical ventilation was employed in 78.2% of ICU patients. The overall mortality rate of the cohort was 29.4%, the ICU mortality rate was 50.9%, and the ventilator-associated mortality was 62.8%.
The MuLBSTA score was applied to each patient at the time of hospitalization. The mean MuLBSTA score was 8.67 for patients who survived and 13.6 for those who died. The correlation between MuLBSTA score and mortality was significant (odds ratio [OR] 1.37; 95% confidence interval [CI] 1.23–1.53; P = .0001). The area under the receiver operating characteristic curve of MuLBSTA for predicting in-hospital death at the time of admission was 0.813 (95% CI, 0.74-0.885). The investigators also found a positive correlation between the MuLBSTA score and ventilator support (OR 1.30; 95% CI 1.17–1.44; P = .0001) and length of stay (r [161] = .35; P = .0001). Kaplan-Meier survival analysis showed that patients with a MuLBSTA score greater than 12 had a higher risk of death compared with those with a score 12 or lower (P = .001).
“I thought this was very interesting because this score takes into account some of the factors that make COVID-19 a unique disease, and age and hypertension were included in the score,” said Marc Feinstein, MD, a critical care medicine specialist at Memorial Sloan Kettering Cancer Center in New York City, who moderated the session. Modifying the score based on risk factors unique to COVID-19, such as obesity, other forms of lung disease, and history of coronary heart disease, may improve the ability of the score to predict COVID-19 outcomes, he and Dr. Tusha agreed.
“So far, we haven’t developed any other scores or expanded on this specific score, but we’re trying to see if it does apply to perhaps stratify which patients get which types of treatment,” said Dr. Tusha. “That would be something for the future.”
Reference
- Hoeper M, Ghofrani H, Al-Hiti H, et al. Switching to riociguat in patients with pulmonary arterial hypertension not at treatment goal with phosphodiesterase type-5 inhibitors: subgroup analysis results of the REPLACE study. CHEST 2020; 158(4 Suppl):A2156–A2159. doi: https://doi.org/10.1016/j.chest.2020.08.1857