Low prevalence of bacterial coinfection at hospital admission in COVID-19 and the role of procalcitonin as a rule-out biomarker.

BACKGROUND

Empirical antibiotic use at hospital admission is common in patients with COVID-19 despite uncertainty regarding the prevalence of bacterial coinfection. Procalcitonin (PCT) has been proposed to support early decision-making, but its diagnostic value at presentation remains unclear.

METHODS

We performed a post hoc analysis of a prospective cohort of adults hospitalized with SARS-CoV-2 infection between March 2020 and July 2022 who had PCT measured within 48 hours of admission. Bacterial coinfection at admission was investigated according to predefined institutional criteria. The diagnostic performance of PCT and other inflammatory biomarkers was assessed using microbiological confirmation as the reference standard. Clinical impact was evaluated using a composite outcome of intensive care unit admission, invasive mechanical ventilation, or 28-day mortality.

RESULTS

Among 2187 patients, bacterial coinfection was identified in 50 (2.3%). The respiratory tract was the most frequent source, and Streptococcus pneumoniae was the predominant pathogen. Patients with coinfection had higher inflammatory biomarker levels and worse outcomes, and coinfection was independently associated with the composite endpoint (adjusted odds ratio 4.31; 95% CI: 2.35-7.84). PCT showed the highest discriminative performance among evaluated biomarkers (AUC 0.79; 95% CI: 0.71-0.87). For PCT values <0.5 ng/mL, the negative predictive value was 99.1% (95% CI: 98.5-99.4), whereas other biomarkers showed lower discrimination despite similarly high negative predictive values.

CONCLUSION

Bacterial coinfection at admission in COVID-19 is uncommon but associated with worse outcomes. PCT demonstrates good discrimination and a high rule-out value, supporting its use to guide early antibiotic decision-making in this low-prevalence setting.