PICU length of stay for paediatric liver transplantation: a single-centre experience from 2000-2021
Jack R Fennessy1, Greer K Waters1, Elena Cavazzoni2.
1School of Medicine, University of Sydney, Sydney, Australia; 2Paediatric Intensive Care Unit, Children's Hospital at Westmead, Sydney, Australia
Introduction: With modern paediatric liver transplantation becoming increasingly resource-intensive, we endeavored to identify the factors that influence both PICU and Hospital length of stay (LOS) and evaluate the effect of a new customised anticoagulation protocol on LOS.
Methods: Data were obtained from the Liver Transplant and ICU databases for 336 transplants performed between 2000-2021 at the Children’s Hospital Westmead. Transplants were analyzed into two epochs, before and after June 2012, representing a change in post-operative anticoagulation management. Significant factors from a univariate regression analysis were included in a multivariate analysis. Significant factors were identified in the final regression models for both PICU and Hospital LOS. Missing variables were imputed using predictive mean matching.
Results: For PICU length of stay, a significant regression equation was found (F(8,301)=20.616, p<0.001), with an R2 of 0.360. Factors that predicted PICU LOS were: Time from Listing to Transplant (β=-0.139, p=0.004), Cold Ischaemic Time (β=0.104, p=0.041), Male Gender (β=0.088, p=0.064), Operation Time (β=0.108, p=0.024), CVVH (β=0.257, p<0.001), Delayed Abdominal Closure (β=0.323, p<0.001), Living Related Graft (β=0.090, p=0.077), and Graft vs. Recipient Weight Ratio (β=0.193, p<0.001). For Hospital length of stay, a significant regression equation was found (F(6,286)=17.470, p<0.001), with an R2 of 0.275. Factors that predicted Hospital LOS were: PELD Score at Listing (β=0.136, p=0.010), Fluid Balance (β=0.124, p=0.020), CVVH (β=0.318, p<0.001), Cold Ischaemic Time (β=0.171, p=0.001), Whole Graft (β=-0.097, p=0.064), and Biliary Atresia (β=-0.216, p<0.001). There was a difference between Epoch one (Mdn=32.4 days) and Epoch two (Mdn=26.3 days) in Hospital LOS (U=9532, p<0.001), but not in PICU LOS (U=12756, p=0.969).
Conclusion: The introduction of the new anticoagulation protocol reduces Hospital LOS, but has no effect on PICU LOS. Patient, surgical, and management factors have been identified to PICU and Hospital LOS. These findings may provide impetus for considering changes in management and further research.
