Maria Martin Delgado, MD
Head of Intensive Medicine
SMI Hospital Universitario Torrejón; Universidad Francisco De Vitoria
Torrejon De Ardoz Madrid, ES, Spain
Disclosure information not submitted.
Elena Gonzalez
MD
Hospital Universitario Torrejon
Madrid, Madrid, Spain
Disclosure information not submitted.
Isabel Rodado, n/a
MD
Hospital Universitario Torrejon, United States
Disclosure information not submitted.
Maria Angeles Estévez
MD
Hospital Universitario Torrejón, United States
Disclosure information not submitted.
Dennis Stanescu
MD
Hospital Universitario Torrejón, United States
Disclosure information not submitted.
Gonzalo Navarro
MD PhD
Hospital Universitario Torrejón, United States
Disclosure information not submitted.
Amelia Pavalascu
MD
Hospital Universitario Torrejón, United States
Disclosure information not submitted.
Noelia Moliner
MD
Hospital Universitario Torrejón, United States
Disclosure information not submitted.
Ángela Algaba
MD PhD
Hospital Universitario Torrejón, United States
Disclosure information not submitted.
Title: Real World Diagnostic Accuracy of PSP for Infection, Sepsis and Mortality in Severe COVID19 Patients
Introduction: Early identification of bacterial sepsis in severe COVID-19 patients admitted to ICU is important to timely initiate antimicrobial therapy and appropriate care. Pancreatic stone protein (PSP), a biomarker that is well-documented for the early identification of sepsis, is evaluated here for the early recognition of bacterial superinfection, sepsis, and prognostication of mortality in the context ICU-admitted COVID-19 patients.
Methods: A single center, prospective, observational study of SARS-CoV-2 positive adult patients admitted to ICU with severe COVID-19 was performed, from February to March 2021. PSP was measured daily by ICU staff until day 10 of ICU stay, ICU discharge or death, whichever came first, using a point-of-care platform (abioSCOPETM, Abionic SA, Epalinges, Switzerland). Demographics, comorbidities, and secondary diagnostics were recorded on admission. Procalcitonin (PCT) and c-reactive protein (CRP) were recorded according to hospital protocols. Patients were retrospectively evaluated for the presence of infection, sepsis, and hospital mortality.
Results: Fifteen patients were included, of which 8 (53.3%) were males with a mean age of 59 years. During the study duration, seven (46.7%) subjects presented with bacterial infection, six (40%) had sepsis, and 3 (20%) subjects died. Area under the receiving operating characteristic (AUROC) curves show that PSP diagnostic accuracy reached during the study for hospital mortality, sepsis and infection were 0.88 (95% CI: 0.69 to 1.00), 0.73 (95% CI: 0.47 to 0.99) and 0.69 (95% CI: 0.41 to 0.97). Average maximum PSP concentration in survivors versus non survivors was 259.5 ng/ml and 538.3 ng/ml (r = 0.053) respectively, and average maximum PSP in sepsis versus non sepsis was 408.5 ng/ml and 253.1 ng/ml respectively (r = 0.151). Spearman’s rank correlation coefficients showed that PSP positively correlated with CRP (r = 0.041) and PCT (r = 0.007), two canonical markers of infection.
Conclusions: In severe COVID-19 patients, high PSP concentration is associated with increased mortality and presence of sepsis and infection. A significant correlation between PSP and canonical sepsis markers such as CRP and PCT was observed, with the added benefit that PSP testing can be done at the patient’s bedside in the ICU.