%0 Journal Article
%T Dynamic decline of lymphocytes predicts extracorporeal membrane oxygenation-related infections: a retrospective observational study.
%A Hao T
%A Jin C
%A Hu D
%A Wu C
%A Zhu Y
%A Xie J
%A Huang L
%A Xu J
%A Chang W
%A Liu L
%A Guo F
%A Qiu H
%A Yang Y
%A Liu S
%J J Thorac Dis
%V 16
%N 7
%D 2024 Jul 30
%M 39144308
%F 3.005
%R 10.21037/jtd-23-1912
%X UNASSIGNED: Limited data are available regarding the current microbiological characteristics of extracorporeal membrane oxygenation (ECMO)-related infections in intensive care units (ICUs) in China. This retrospective study aimed to determine the epidemiology, risk factors and impact on the outcome of ECMO-related infections.
UNASSIGNED: A retrospective observational study from January 2014 to December 2019 was performed, and adult patients receiving ECMO support for more than 48 hours were included in this study. The primary outcome was the incidence rate of ECMO-related infection. Clinical data were recorded, and risk factors associated with an increased risk of ECMO-related infection were analyzed.
UNASSIGNED: A total of 174 adult patients who received ECMO and underwent ECMO for 1,670 days were included in this study. Forty-six patients (26.4%) developed ECMO-related infections, corresponding to 27.5 first episodes/1,000 ECMO days. The most common ECMO-related infection observed was ventilator-associated pneumonia (VAP). Infected patients had longer durations of mechanical ventilation {20.2 [interquartile range (IQR), 12.6, 30.7] vs. 9.0 (IQR, 5.8, 14.7) days, P<0.001}, ECMO support [11.6 (IQR, 8.1, 17.3) vs. 7.6 (IQR, 5.6, 9.7) days, P<0.001] and hospital stays (28.2±20.7 vs. 22.0±15.6 days, P<0.001). The factors independently associated with ECMO-related infection were a dynamic decrease in lymphocyte count [adjusted odds ratio (OR) =3.578, 95% confidence interval (CI): 2.175-4.906, P<0.001] and ECMO duration (adjusted OR =1.207, 95% CI: 1.096-1.330, P<0.001). Compared to patients without infection, infected patients had greater hospital mortality (39.1% vs. 78.3%, P<0.001) and 90-day mortality (40.6% vs. 87.0%, P<0.001). ECMO-related infections were associated with worse outcomes (adjusted Kaplan-Meier curve, log rank test P<0.001).
UNASSIGNED: Patients supported by ECMO had a high risk of developing ECMO-related infection. The most common ECMO-related infection observed was VAP. A dynamic decrease in lymphocyte counts was significantly associated with an increased risk of ECMO-related infection.