Incidence of healthcare-associated infections in patients with fever during the first 48 hours after decannulation from veno-venous extracorporeal membrane oxygenation.

Introduction: Fevers following decannulation from veno-venous extracorporeal membrane oxygenation often trigger an infectious workup; however, the yield of this workup is unknown. We investigated the incidence of post-veno-venous extracorporeal membrane oxygenation decannulation fever as well as the incidence and nature of healthcare-associated infections in this population within 48 hours of decannulation.

Methods: All patients treated with veno-venous extracorporeal membrane oxygenation for acute respiratory failure who survived to decannulation between August 2014 and November 2018 were retrospectively reviewed. Trauma patients and bridge to lung transplant patients were excluded. The highest temperature and maximum white blood cell count in the 24 hours preceding and the 48 hours following decannulation were obtained. All culture data obtained in the 48 hours following decannulation were reviewed. Healthcare-associated infections included blood stream infections, ventilator-associated pneumonia, and urinary tract infections.

Results: A total of 143 patients survived to decannulation from veno-venous extracorporeal membrane oxygenation and were included in the study. In total, 73 patients (51%) were febrile in the 48 hours following decannulation. Among this cohort, seven healthcare-associated infections were found, including five urinary tract infections, one blood stream infection, and one ventilator-associated pneumonia. In the afebrile cohort (70 patients), four healthcare-associated infections were found, including one catheter-associated urinary tract infection, two blood stream infections, and one ventilator-associated pneumonia. In all decannulated patients, the majority of healthcare-associated infections were urinary tract infections (55%). No central line-associated blood stream infections were identified in either cohort. When comparing febrile to non-febrile cohorts, there was a significant difference between pre- and post-decannulation highest temperature (p < 0.001) but not maximum white blood cell count (p = 0.66 and p = 0.714) between the two groups. Among all positive culture data, the most commonly isolated organism was (41.7%) followed by (33%). Median hospital length of stay and time on extracorporeal membrane oxygenation were shorter in the afebrile group compared to the febrile group; however, this did not reach a statistical difference.

Conclusion: Fever is common in the 48 hours following decannulation from veno-venous extracorporeal membrane oxygenation. Differentiating infection from non-infectious fever in the post-decannulation veno-venous extracorporeal membrane oxygenation population remains challenging. In our febrile post-decannulation cohort, the incidence of healthcare-associated infections was low. The majority were diagnosed with a urinary tract infection. We believe obtaining cultures in febrile patients in the immediate decannulation period from veno-venous extracorporeal membrane oxygenation has utility, and even in the absence of other clinical suspicion, should be considered. However, based on our data, a urinalysis and urine culture may be sufficient as an initial work up to identify the source of infection.