Validity and Accuracy of the Derived Left Ventricular End-Diastolic Pressure in Impella 5.5

Background: Consensus regarding on-support evaluation and weaning concepts from Impella 5.5 support is scarce. The derived left ventricular end-diastolic pressure (dLVEDP), estimated by device algorithms, is a rarely reported tool for monitoring the weaning process. Its validation and clinical accuracy have not been studied in patients. We assess dLVEDP's accuracy in predicting pulmonary capillary wedge pressure (PCWP) and propose a corrective equation.

Methods: We included 29 consecutive patients treated with Impella 5.5: 12 in a generation cohort and 17 in a validation cohort. dLVEDP and PCWP were measured 5-fold every 8 hours during support, totaling 698 series with 3490 measurements. Variables such as Impella 5.5 performance level, heart rhythm, pacemaker settings, sex, mechanical ventilation, and body mass index were recorded. Linear regression was used to correct dLVEDP-PCWP discrepancies. Analysis included Bland-Altman plots, linear regression, histograms, and violin plots.

Results: The raw dLVEDP and PCWP data did not coincide satisfactorily. The Impella 5.5 dLVEDP overestimation was 3.5±1.5 mm Hg (mean±SD), increasing with higher pressures and unaffected by cardiac rhythm, mechanical ventilation, and performance levels. Statistical correction using the formula modified dLVEDP=-0.457+(1-sex[1=male, 0=female])×0.719-0.0496× body mass index+1.015×body surface area+0.811×dLVEDP significantly reduced the overestimation (P<0.01) to 0.0±1.2 mm Hg.

Conclusion: dLVEDP, calculated by the Impella 5.5 Smart Algorithm, is a feasible and effective tool for continuously monitoring PCWP at performance levels 3 to 9. Correction of dLVEDP by using the described equation further enhances its accuracy. Hence, hemodynamic surveillance via dLVEDP may aid in managing and weaning temporary microaxial support, potentially reducing the need for continuous monitoring with a Swan-Ganz catheter.