Secondary bacterial infection in COVID-19 patients is associated with increased mortality and disproportionately affects critically ill patients. This single-centre retrospective observational study investigates the comparative efficacy of change in procalcitonin (PCT) and other commonly available biomarkers in revealing or predicting microbiologically proven secondary infection in critical COVID-19 patients. Adult patients admitted to an intensive care unit (ICU) with confirmed SARS-CoV-2 infection between 9 March 2020 and 5 June 2020 were recruited to the study. For daily biomarker and secondary infection, laboratory-confirmed bloodstream infection (LCBI) and ventilator-associated pneumonia/tracheobronchitis (VAP/VAT) data were collected. We observed a PCT rise in 53 (81.5%) of the patients, a C-reactive protein (CRP) rise in 55 (84.6%) and a white blood cell count (WBC) rise in 61 (93.8%). Secondary infection was confirmed in 33 (50.8%) of the patients. A PCT rise was present in 97.0% of patients with at least one confirmed VAP/VAT and/or LCBI event. CRP and WBC rises occurred in 93.9% and 97.0% of patients with confirmed VAP/VAT and/or LCBI, respectively. Logistic regression analysis found that, when including all biomarkers in the same model, there was a significant association between PCT rise and the occurrence of LCBI and/or VAP/VAT (OR = 14.86 95%CI: 2.20, 342.53; p = 0.021). Conversely, no statistically significant relationship was found between either a CRP rise (p = 0.167) or a WBC rise (p = 0.855) and the occurrence of VAP/VAT and/or LCBI. These findings provide a promising insight into the usefulness of PCT measurement in predicting the emergence of secondary bacterial infection in ICU.
Keywords: C-reactive protein; COVID-19; procalcitonin; secondary infection.