Abstract
Introduction
Hospital – acquired pneumonia (HAP) and ventilator – associated pneumonia (VAP) represent a significant healthcare burden with increased mortality among critically ill patients. Unfortunately, universal definitions and established clinical characteristics remain debated, especially in patients receiving extracorporeal membrane oxygenation (ECMO) support for cardiopulmonary failure. This study evaluates factors associated with respiratory culture positivity in patients receiving ECMO.
Methods
In this single – center retrospective cohort of adult patients receiving ECMO between January 2022 and June 2024; we evaluated clinical parameters and patient factors associated with suspected HAP/VAP, defined as a new positive respiratory culture with clinical characteristics of infection. Additionally, isolated organisms were assessed for persistent positivity.
Results
Among the 88 samples collected from 35 patients, no clinical or laboratory differences were identified when comparing samples with positive organisms recovered vs negative culture growth. ECMO-associated HAP/VAP incidence rate was 18.7 per 1000 ECMO – days. Recovered pathogens were primarily from the Enterobacteriaceae family or Pseudomonas aeruginosa at 33.3% and 27.8% respectively. Only 27.3% of patients who met criteria for HAP/VAP had evidence of radiographic changes. A minority of HAP/VAPs (27.3%) had microbiologic cure after treatment completion and no patient cleared Burkholderia cepacia, Staphylococcus aureus, or Stenotrophomonas maltophilia in this cohort.
Conclusion
Timely diagnosis and management of HAP/VAP is essential in the care of the critically ill. However, clear incidence rates and diagnostic accuracy has been hindered by heterogenous definitions and poor sensitivity/specificity of clinical indicators. The need for a unified diagnostic criteria are amplified in ECMO where clinical management is confounded by a persistent proinflammatory state and frequent respiratory tract colonization.
Keywords
Get full access to this article
View all access options for this article.
