Elizabeth Wei
Rady Children's Hospital - San Diego
San Diego, California
Disclosure information not submitted.
Kim Vuong, BS
Student
1The Research Associate Program in Emergency Medicine and Critical Care, Department of Emergency Medicine, University of Maryland School of Medicine, Baltimore, Maryland, USA, United States
Disclosure information not submitted.
Elizabeth Ingulli, MD
Attending Physician, Clinical Professor
University of California - San Diego, Rady Children's Hospital, United States
Disclosure information not submitted.
Nicole Coufal, MD, PhD
Attending Physician, Assistant Professor
University of California - San Diego, Rady Children's Hospital, United States
Disclosure information not submitted.
Title: Predictors of successful discontinuation of continuous renal replacement therapy in children
INTRODUCTION/HYPOTHESIS: Continuous renal replacement therapy (CRRT) is needed in 1.5% of pediatric patients with acute kidney injury. Adult data suggests that urine output is the most robust parameter for predicting renal recovery and liberation from CRRT. The goal of this study is to determine factors associated with successful discontinuation of CRRT in pediatric patients.
Methods: A retrospective single center cohort study of all patients < 21 years of age who received CRRT from January 2011 to March 2021 was performed. Patients for whom CRRT was suspended for greater than 12 hours or intentionally discontinued were analyzed. Those who remained off CRRT for at least 7 days (success) were compared with those requiring re-initiation of CRRT (failure). Biochemical and physiologic parameters, including urine output, diuretic administration, vasoactive infusion scores (VIS), and serum creatinine, were collected. The predictive ability of urine output was further analyzed using the area under the receiver operative characteristic (ROC) curve method.
Results: 150 patients were treated with CRRT. Of the 99 patients trialed off CRRT, 76 patients remained off CRRT (success); 23 were restarted (failure). There were no significant differences between the two groups in terms of VIS score, serum creatinine, diuretic administration, or urine output at CRRT initiation. The success group demonstrated significantly higher urine output in the 6-hour (0.833 vs 0.06 ml/kg/hr, p=0.0007) and 24-hour periods (0.83 vs 0.18 ml/kg/hr, p=0.0059) prior to CRRT discontinuation, as well as in the 6 hours (1.8 vs 0.1 ml/kg/hr, p=0.001) and 12 hours (2.4 vs 0.09 ml/kg/hr, p< 0.0001) after CRRT discontinuation. The ROC curve for urine output in the 6 hours prior to CRRT discontinuation was 0.72 (95% CI 0.60-0.84, p=0.0009). Urine output of 0.51 ml/kg/hr in the 6 hours prior to discontinuation had the highest combined sensitivity and specificity (77% and 62%, respectively). Urine output in the 6 hours after CRRT discontinuation demonstrated similar area under ROC curve, sensitivity and specificity.
Conclusions: Patients who successfully liberated from CRRT demonstrated significantly higher urine output prior to CRRT discontinuation. Urine output of 0.5 ml/kg/hr in the 6 hours prior to CRRT discontinuation was predictive of renal recovery.