ORIGINAL ARTICLES   Free access

Minerva Anestesiologica 2014 January;80(1):29-38

Copyright © 2014 EDIZIONI MINERVA MEDICA

lingua: Inglese

Clinical characteristics, hospital outcome and prognostic factors of patients with ventilator-related pneumothorax

Hsu C. W. ^{1, 2}, Sun S. F. ^{2, 3}, Lee D. L. ^{2, 4}, Chu K. A. ⁴, Lin H. S. ⁵ ✉

¹ Intensive Care Unit, Department of Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan; ² School of Medicine, National Yang-Ming University, Shipai, Beitou District, Taipei, Taiwan; ³ Department of Physical Medicine and Rehabilitation, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan; ⁴ Department of Chest Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan; ⁵ School of Nursing, Fooyin University, Ta-liao,Taiwan

Background: Mechanical ventilation is a common cause of iatrogenic pneumothorax in intensive care units (ICU). Most of the patients with ventilator-related pneumothorax (VRP) have underlying lung diseases and is associated with increased morbidity and mortality. The prognostic factors of VRP are not clear. The objective of this study was to find the possible prognostic factors.
Methods: Analysis of retrospectively collected data of patients with pneumothorax induced by mechanical ventilation. Data were obtained concerning demographics, acute physiology and chronic health evaluation (APACHE) II score, organ failure, underlying diseases, interval between the start of mechanical ventilation and pneumothorax, arterial blood gas, respiratory parameters and patient outcomes.
Results: One hundred and twenty-four patients with VRP were included for analysis. The incidence rate of VRP was 0.4% (124/31,660), and the mortality rate was 77.4%. The patients with VRP had higher hospital mortality rate than that of mechanically ventilated patients without pneumothorax (77.4% vs. 13.7%, P<0.001) or patient with procedure-related pneumothorax (77.4% vs. 29.4%, P<0.001). Most cases of VRP occurred in the early phase of mechanical ventilation, and 8.9% of the patients had a later episode of pneumothorax on the opposite lung. The interval between two episodes of VRP was short, at a median time of 2 days. Cox regression analysis showed that tension pneumothorax (P=0.001), PaO2/FiO2<200 (P=0.002), and APACHE II score (P=0.008) were significantly associated with death.
Conclusion: VRP patients with tension pneumothorax or PaO2/FiO2<200 had a higher risk of death. APACHE II scores were associated with mortality in the VRP patients with PaO2/FiO2≥200 mmHg.