Impact of pre-dilution and flushing on continuous renal replacement therapy

Koh KH
Correspondence: Dr Koh Keng Hee, kohkenghee@yahoo.com

ABSTRACT
Introduction Infusing the replacement solution before the filter (pre-dilution) and regular flushing have not been accounted for in conventional mathematical equations. Their effects on various continuous renal replacement therapy (CRRT) parameters, such as ultrafiltration fraction and urea clearance, have not been well studied. We incorporated these parameters into mathematical equations to help in understanding and prescribing CRRT.
Methods We built a mathematical model to evaluate the plasma flow rate, filter fluid inflow rate with pre- and post-dilution and ultrafiltration rate, plasma clearance of urea, and ultrafiltration fraction.
Results In pre-dilutional therapy, the volume of replacement needed to be increased in order to achieve the target plasma clearance rate. The extra volume needed increased exponentially with higher target plasma clearance rate. The higher the targeted plasma clearance in relation to blood flow rate (Qb), the greater the increase. Increasing blood flow rate reduced the ultrafiltration fraction for both pre- and post-dilution therapy. It had no effect on plasma clearance with post-dilution therapy but significantly improved the clearance in pre-dilution therapy. Higher haematocrit resulted in higher ultrafiltration fraction in both pre- and post-dilution therapy. Higher haematocrit had no effect on plasma clearance with post-dilution therapy but slightly reduced clearance in pre-dilution therapy. For a given total volume of ultrafiltration, flushing reduced plasma clearance with both pre- and post-dilution therapy. Flushing slightly reduced ultrafiltration fraction of post-dilution therapy but minimally increased the ultrafiltration fraction of pre-dilution therapy. This mathematical model could be utilised in prescribing Qb and replacement rate based on targeted plasma clearance, ultrafiltration fraction, fluid removal rate and haematocrit. There was close approximation of predicted and measured urea plasma clearance.
Conclusion Pre-dilution therapy reduced urea clearance significantly and this needed to be compensated for by increasing the volume of ultrafiltration or Qb. Flushing was of limited benefit and may reduce urea clearance. In prescribing haemofiltration, Qb and replacement rate could be determined with this model.

Keywords: continuous renal replacement therapy, flushing, haemofiltration, mathematical modelling, pre-dilution haemofiltration, ultrafiltration fraction
Singapore Med J 2006; 47(9): 785-795