Clinical assessment of cannula performance during adult minimally invasive valve surgery using predictive mathematical models.

OBJECTIVES

Peripheral cannula selection in minimally invasive cardiac surgery (MICS) is crucial, as venous drainage limitations during cardiopulmonary bypass (CPB) can impair end-organ perfusion and overall outcomes. This study evaluates the in vivo venous drainage efficiency of different peripheral venous cannulas for adult MICS by applying dynamic similarity (DS) prediction formulas.

METHODS

This observational study included 75 adult MICS patients with use of single peripheral venous cannulation from January 2017 to June 2023. The analysis included cannulas Bio-Medicus Multi XL Life Support, Nextgen Bicaval and Smart cannulas. Cannula performance was assessed by comparing predicted pressures with in vivo measurements.

RESULTS

A total of 278, 314 and 264 measurements were recorded for the Multi, Next and Smart cannulas, respectively. No significant demographic differences were found between groups. Lin's concordance correlation coefficients for agreement between predicted and measured pressures were >0.95 for the Multi and Next cannulas but substantially lower for the Smart cannula (0.80). Bland-Altman analysis showed a mean bias of -1.09 mmHg for the Multi, -0.15 mmHg for the Next and 3.68 mmHg for the Smart cannula, with the latter exceeding the ±12.5 mmHg limits of agreement.

CONCLUSIONS

DS-based predictions revealed significant performance variability among peripheral venous cannula types. The in vivo performance was adequately predicted for the Multi and Next cannulas but not for the Smart cannula, underscoring the need for real-life evaluations and performance monitoring during CPB. Incorporating dynamic performance assessments into clinical decision-making might optimize venous drainage and patient outcomes.