Background: COVID-19-associated pulmonary aspergillosis (CAPA) has emerged as a serious complication of severe SARS-CoV-2 pneumonia in critically ill patients requiring invasive mechanical ventilation (IMV). Several host- and treatment-related risk factors have been identified, but the role of environmental factors-particularly ICU room air pressure-remains underexplored.

Methods: We conducted an ancillary analysis of the COVID-ICU study, a prospective multicenter cohort of COVID-19 patients admitted to ICUs in France, Belgium and Switzerland during the first wave of the pandemic. Our analysis included 1233 mechanically ventilated patients. We compared the cumulative incidence of probable invasive pulmonary aspergillosis (IPA), defined according to the 2024 Invasive-Fungal-Diseases-in-Adult-Patients-in-ICU (FUNDICU) consensus criteria, between patients isolated in negative-pressure rooms and those housed in neutral-pressure rooms. Secondary outcomes included the incidence of putative IPA, the incidence of putative IPA or Aspergillus colonization (both defined according to the AspICU algorithm), and the association between probable IPA occurrence and clinical outcomes.

Results: Probable IPA occurred significantly less frequently in patients isolated in negative-pressure rooms, compared to those in neutral-pressure rooms (2.0% vs. 4.8%; adjusted cause-specific hazard ratio [cHR] 0.44, 95%CI 0.21-0.90, p = 0.024). No significant between-group difference was observed in the incidence of putative IPA alone or in combination with Aspergillus colonization (1.3% vs. 2.1%, adjusted cHR 0.65, 95% CI 0.23-1.77, p = 0.4 and 3.3% vs. 5.8%, adjusted cHR 0.66, 95% CI 0.30-1.42, p = 0.29, respectively). The occurrence of probable IPA was associated with longer IMV duration and ICU length-of-stay (adjusted HR 0.45, 95% CI 0.27-0.76, p = 0.002 and 0.59, 95% CI 0.38-0.91, p = 0.017, respectively).

Conclusion: In this large multicenter cohort of mechanically ventilated COVID-19 patients, isolation in negative-pressure ICU rooms was associated with a lower incidence of probable IPA, but the small event rate and potential confounders indicate this finding needs to be replicated and validated by other studies.