ESPN 53rd Annual Meeting

ESPN 2021


 
Continuous renal replacement therapy in low weight patients: a retrospective study on newborns and infants treated with Ca.R.Pe.Di.E.M (Cardio-Renal, Pediatric Dialysis Emergency Machine) in Padua University Hospital
LISA CUSINATO 1 DARIO GREGORI 2 ELISA BENETTI 1 MATTIA PAROLIN 1

1- PEDIATRIC NEPHROLOGY DIALYSIS AND TRANSPLANT UNIT, PADUA UNIVERSITY HOSPITAL
2- UNIT OF BIOSTATISTICS, EPIDEMIOLOGY AND PUBLIC HEALTH, DCTVPH, UNIVERSITY OF PADOVA
 
Abstract:

OBJECTIVES: 
CRRT in low weight patients is a clinical and technical challenge: Ca.R.Pe.Di.E.M. is one of the first miniaturized dialysis machine dedicated to these patients.
Aim of this study: Identify demographic, clinical and dialytic factors potentially related to mortality and length of stay in Intensive Care Unit (ICU). 

METHODS:
A retrospective analysis was performed enrolling patients treated with Ca.R.Pe.Di.E.M. between 01.01.2014 ad 30.09.2020 in University Hospital of Padua. A Random Forest model was implemented for outcomes “mortality” and “length of stay”. 
 
RESULTS:
20 patients received 211 CRRT session (75% CVVH, 25% CVVHD) in 3033 hours. Their body weight was between 1610-9100g. The median age at ICU admission was 38 days[IQR 9,75-142], median weight at CRRT start 4015g[IQR 3400-4970] with median fluid overload 23%[IQR 10-28,5]. Most of patients (75%) were treated using a 5Fr CVC (4Fr in 10%, 4.5Fr in 5% and 6.5Fr in 10%). Circuits were primed with 5% albumin in 16 patients; no one received blood priming. Downtime occurred in 2.3% of total time of treatment; the most common cause was clotting. Survival at CRRT discontinuation and hospital discharge were 95% and 50% respectively (4 patients were discharged in peritoneal dialysis and 6 without need of renal replacement therapy). Median ICU length of stay was  7.5 days[IQR 3-25,75]. A Random Forest model was implemented. Predictors best related with mortality were numbers of ICU days before CRRT, need of inotropic support, fluid overload rate, bicarbonates and haemoglobine level, anticoagulation and dialytic dose. Predictors best related with length of stay were loss of weight rate, pH and haemoglobin level. 
 
CONCLUSIONS:
CRRT with Ca.R.Pe.Di.E.M is safe and efficient in a heterogeneous group of low-weight patients, without major complication and significant downtime.