Bloodstream infections during Impella support in refractory cardiogenic shock: incidence, microbiology and clinical implications
G Chatzis, S Syntila, B Markus, K Sassani, B Schieffer, H Lapp, F KanoAbstract
Abstract
Patients with refractory cardiogenic shock (CS) supported with percutaneous mechanical circulatory support (MCS) are exposed to prolonged intensive care unit (ICU) care, invasive central venous lines and immunological dysregulation. Bloodstream infections (BSI) may worsen outcomes, yet data on BSI burden and microbiology during Impella support are limited.
Purpose
To quantify confirmed BSI in Impella-treated refractory CS, describe microbiology, and explore etiology-specific differences.
Methods
This was a retrospective single-center cohort (Germany, 2013–2021) including adults with refractory CS treated with Impella 2.5/CP. Confirmed BSI was defined as positive blood cultures with clinical correlation; isolates were categorized into Gram-positive, Gram-negative and fungal pathogens and repeated cultures were considered per patient episode. Analyses were descriptive and stratified by CS etiology (acute coronary syndrome (ACS), dilated cardiomyopathy (DCM) / myocarditis, aortic stenosis, sepsis-associated CS, other), survival comparisons were exploratory given confounding by duration of support/ICU exposure (confounding by indication and by duration).
Results
Among 409 Impella-treated patients, confirmed BSI occurred in 23.7% (97/409). Culture positivity was highest in sepsis-associated CS (75.0% (27/36). The microbiology spectrum included frequent staphylococcal species (including skin commensals and S. aureus) and clinically relevant Gram-negative organisms (Enterobacteria and non-fermenters such as Pseudomonas), with fungal isolates less common. This pattern is compatible with a mixture of catheter/device-associated and critical-illness-associated infections. Patients with sepsis-associated CS also demonstrated the highest early mortality in the overall cohort, highlighting a high-risk phenotype in which infection and shock mechanisms overlap. Interestingly enough, the incidence of BSI under Impella support was not higher compared to the CS group of patients treated for the same period without Impella or other MCS device.
Conclusion
Confirmed BSI is common during Impella support in refractory CS, particularly in sepsis-associated CS. These findings support stringent infection surveillance (early blood cultures, daily line/device review), prompt source control, and structured ICU prevention bundles in MCS patients. Key limitations include retrospective ascertainment and potential under-/over-detection depending on sampling intensity. Prospective studies with standardized sampling and time-dependent modeling are needed to disentangle BSI as a driver of deterioration versus a marker of prolonged critical illness and exposure.