suPAR cut-offs for stratification of low, medium, and high-risk acute medical patients in the emergency department

Abstract

Background: Soluble urokinase plasminogen activator receptor (suPAR) levels have previously been associated with readmission and mortality in acute medical patients in the ED. However, no specific cut-offs for suPAR have been tested in this population. Methods: Prospective observational study of consecutively included acute medical patients. Follow-up of mortality and readmission was carried out for 30- and 90 days stratified into baseline suPAR < 4, 4–6 and > 6 ng/ml. suPAR levels were measured using suPARnostic® Turbilatex assay on a Cobas c501 (Roche Diagnostics Ltd) analyser. Results: A total of 1747 acute medical patients in the ED were included. Median age was 70 (IQR: 57–79) and 51.4% were men. Adjusted linear regression analysis showed that suPAR, independently of age, sex and C-reactive protein levels, predicted 30- and 90-day mortality (Odds ratio for doubling in suPAR 1.96 (95% confidence intervals: 1.42–2.70) Among patients with suPAR below 4 ng/ml (N = 804, 46.0%), 8 (1.0%) died within 90-day follow-up, resulting in a negative predictive value of 99.0% and a sensitivity of 94.6%. Altogether 514 (29.4%) patients had suPAR of 4–6 ng/ml, of whom 43 (8.4%) died during 90-day follow-up. Among patients with suPAR above 6 ng/ml (N = 429, 24.6%), 87 patients (20.3%) died within 90-day follow-up, resulting in a positive predictive value of 20.1% and a specificity of 78.7%. Conclusions: suPAR cut-offs of below 4, between 4 and 6 and above 6 ng/ml can identify acute medical patients who have low, medium or high risk of 30- and 90-day mortality. The turbidimetric assay provides suPAR results within 30 min that may aid in the decision of discharge or admission of acute medical patients.