Kidney Week

Abstract: TH-PO355

Bias Between In Vitro and In Vivo Mass-Transfer Coefficient of Urea (KoA) Is Larger in Pediatric Than in Adult Patients on Chronic Hemodialysis

Session Information

• Dialysis: Dialysate and Clearance
  October 25, 2018 | Location: Exhibit Hall, San Diego Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Dialysis

• 701 Dialysis: Hemodialysis and Frequent Dialysis

Authors

• Gotta, Verena, University of Basel Children''s Hospital, Basel, Switzerland

Verena Gotta,

• Marsenic Couloures, Olivera, Yale University , New Haven, Connecticut, United States

Olivera Marsenic Couloures,

• Pfister, Marc, University of Basel Children's Hospital, Basel, Switzerland, Hilterfingen, Switzerland

Marc Pfister,

Background

Hemodialysis (HD) prescription with respect to blood flow (Q_B) dialysate flow (Q_D) and filter KoA significantly differs between pediatric and adult patients. This may explain why urea dialytic clearance (K_D) has been observed to be significantly underpredicted from Q_B, Q_D and KoA in pediatric HD patients using a mechanistic equation. The objective of this analysis was to evaluate factors that could explain this bias, assuming that it results from a bias between reported in vitro determined KoA and actual in vivo KoA.

Methods

An urea kinetic model previously developed in adult patients was scaled to 923 paediatric and young adult patients aged 1-29 years based on a priori physiologic knowledge (inter-compartmental clearance and volumes of distribution; “base model”). Utilizing data from 2676 HD sessions of those patients with pre- and post-HD urea concentration measurements, a mixed effect modelling approach was applied to evaluate the relationship between individual estimates of KoA correction factors (f_KoA) required for unbiased prediction of K_D (as indicated by unbiased post-HD urea predictions), and prescription related parameters (Q_B, Q_D, Q_D/Q_B ratio, filter reuse, low/high-flux, ultra-filtration rate, duration of treatment).

Results

Q_D/Q_B ratio was the parameter most strongly associated with individual estimates of f_KoA (≈ -10%, 10%, 100% and 200% increased in vivo KoA estimated at Q_D/Q_B ratios of 1.5, 2, 5, and 10, respectively; p<0.001), explaining 18% of inter-individual variability (decrease from 42.8 to 35.1%). Inclusion of Q_D/Q_B ratio as a covariate removed bias observed with Q_B and Q_D. A proposed correction equation for Q_B (true Q_B lower than nominal Q_B at rates >200 mL/min) also corrected bias observed at Q_B > 300 mL/min, but did not reduce inter-individual variability. Inclusion of session duration, low-flux filters, and ultra-filtration rate did further improve the model fit, however with small reductions in inter-individual variability.

Conclusion

While in vivo KoA has been reported to be lower in adult patients than established in vitro, the opposite was observed for pediatric patients in the present study. The observed bias could best be explained by high Q_D/Q_B ratios used in pediatric patients indicating that dialyzers require specific characterization under pediatric conditions.

Funding

• Private Foundation Support