Studies look at new and old scores, biomarkers for predicting COVID-19 outcomes

The SOFA score did not perform well for predicting mortality in patients admitted with COVID-19, multiple recent studies found, but levels of D-dimer and C-reactive protein had prognostic value, according to the latest research.


Several recent studies looked at the effectiveness of biomarkers and scores for predicting outcomes in patients with COVID-19.

Two pneumonia severity scores (PSI and CURB-65) outperformed two alternatives (qSOFA and the newer viral pneumonia-specific MuLBSTA) at predicting adverse outcomes after admission for COVID-19, according to a study published by the Journal of General Internal Medicine on Feb. 11. The retrospective Spanish study included 10,238 patients with COVID-19, 20.9% of whom died in the hospital. The area under the receiver-operating characteristic curve (AUROC) was higher for the PSI and CURB-65 (0.835 and 0.825, respectively) than for qSOFA and MuLBSTA (0.728 and 0.715, respectively). The qSOFA score, which the authors noted is the easiest to calculate, was the most specific at 95.7%, but sensitivity was only 26.2%. PSI had the highest sensitivity at 84.1% and a specificity of 72.2%. None of the scores performed well on the secondary outcomes of ICU admission or use of mechanical ventilation. The authors noted that at least 22 scores have been developed for COVID-19, but they will require external validation, so some well-established scores may prove useful.

A second study, published as a research letter by JAMA on Feb. 17, agreed that the SOFA score is not useful as a predictor of mortality in ICU patients with COVID-19. It included data from 675 patients from 18 U.S. ICUs who were intubated four hours or longer after receiving oxygen; 59.3% met the study's primary outcome of death or discharge to hospice. Their worst median SOFA score prior to intubation was 6 (interquartile range, 4 to 8). The AUROC for predicting the primary outcome was 0.59 for the SOFA score, compared to 0.66 for age alone. “The discriminant accuracy of the SOFA score for mortality prediction in patients prior to intubation for COVID-19 pneumonia was poor and significantly inferior to simply using age,” the authors said. “The SOFA score possesses inadequate discriminant accuracy to be used for ventilator triage of COVID-19 patients.”

A hospitalized COVID-19 patient's change in C-reactive protein (CRP) level may predict survival, according to a study published by the Journal of Hospital Medicine on Feb. 17. The retrospective analysis analyzed 324 patients who received corticosteroid treatment for COVID-19 at one U.S. medical center by whether their CRP levels decreased by at least half within 72 hours of steroid treatment. The 131 patients whose levels did respond had a significantly lower mortality risk than the 92 whose didn't (25.2% vs 47.8%; P<0.001). The authors concluded that CRP levels may serve as an early marker of response to corticosteroid therapy in COVID-19. “Future studies are needed to validate these findings in other cohorts and to determine if markers other than CRP levels may be predictors of a therapeutic response or if CRP nonresponders would benefit from other targeted therapies,” they wrote.

Another study, published by Critical Care Medicine on Feb. 12, found that ICU patients' D-dimer levels on day 1 or 2 predicted their 28-day mortality risk. Out of 3,418 patients from 68 U.S. hospitals, 93.6% had a D-dimer level above normal and 34.5% died within 28 days. Patients with a D-dimer at least 8 times normal had significantly higher risk of death than those with a level less than twice normal (unadjusted odds ratio, 3.11 [95% CI, 2.56 to 3.77]; adjusted odds ratio, 1.81 [95% CI, 1.43 to 2.28]). Adjustment for early initiation of therapeutic anticoagulation did not attenuate the risk, leading the authors to conclude that their data should not be used to justify anticoagulation as a means to reduce mortality in ICU patients with COVID-19. “Higher D-dimer levels may be a general marker of disease severity rather than reflective of a unique pathophysiology driving mortality. Alternatively, elevated D-dimer levels could be indicative of a hypercoagulable state, but initiation of therapeutic anticoagulation with an elevated D-dimer may be too late to alter the pathologic process,” they wrote.