Computational fluid dynamics assessment of a triple-lumen hemodialysis catheter

BACKGROUND: Central venous hemodialysis catheters are widely used in critically ill patients, but conventional double-lumen designs have limited functionality and are associated with high recirculation rates and thrombotic risk. Triple-lumen catheters can potentially meet the needs of simultaneous hemodialysis and intravenous therapy, yet their hemodynamic performance has not been systematically evaluated. METHODS: A simplified superior vena cava (SVC)-catheter model was constructed. The Carreau non-Newtonian blood model was employed, and the Navier-Stokes equations were solved under mixed boundary conditions, including pulsatile SVC inflow and constant dialysis flow. A catheter flow rate of 400 mL/min was selected to represent a high-flow intermittent hemodialysis condition and to evaluate the catheter under a relatively demanding hemodynamic setting. Recirculation rate (RR) was calculated using Lagrangian particle tracking with 32,000 particles. Blood damage index (BDI) was calculated using the Grigioni power-law damage accumulation model with 1,400 platelet-sized particles. Wall shear stress (WSS) distribution was analyzed, and standard versus reversed connection configurations were compared. RESULTS: , and high WSS areas accounted for 1.71%. High shear stress was predominantly localized at side-hole edges and the distal tip. CONCLUSION: A computational fluid dynamics (CFD) assessment framework was established for an acute, short-term, non-tunneled triple-lumen hemodialysis catheter for critically ill patients requiring temporary renal replacement therapy and an additional venous route. Under the single high-flow intermittent hemodialysis condition studied, the catheter demonstrated low recirculation, low BDI, and localized high WSS at side-hole edges and the distal tip. As comparator catheters were not simulated within the same framework, comparisons with prior literature are indirect and exploratory only. Direct comparator simulations and experimental validation are needed.