PURPOSE: We aimed to clarify the acid-base abnormalities of patients with acute kidney injury (AKI) requiring peritoneal dialysis (PD) in pediatric cardiac care units. METHODS: A retrospective observational study was conducted in a pediatric cardiac care unit in a tertiary care university hospital. The subjects were 40 patients with AKI requiring PD between 2003 and 2005, and controls matched by type of surgery and body weight. Acid-base variables, including blood gas data and electrolytes, were assessed. The Stewart-Figge variables, including strong ion difference apparent (SIDa), strong ion difference effective (SIDe), and strong ion gap (SIG), were calculated. RESULTS: Blood gas analyses showed that the PD group was more acidemic, with a lower mean bicarbonate and a lower mean base excess, typical features of metabolic acidosis. The strong ion analyses revealed that the PD group had lower mean sodium and albumin concentrations. Based on the Stewart-Figge methodology, SIDa was smaller in the PD group than in the control group, but SIG was similar in the two groups. Receiver-operating characteristic curve analyses showed that serum albumin was the only prognostic factor associated with PCCU mortality, even after adjustment for PD treatment. CONCLUSION: Patients with AKI requiring PD in a pediatric cardiac care unit had significant metabolic acidosis compared to controls matched by the type of surgery and body weight. Hyponatremia and hypoalbuminemia were characteristics of these patients. The calculated SIDa was smaller in the PD than in the control group. Only the serum albumin had a significant prognostic value.
PURPOSE: We aimed to clarify the acid-base abnormalities of patients with acute kidney injury (AKI) requiring peritoneal dialysis (PD) in pediatric cardiac care units. METHODS: A retrospective observational study was conducted in a pediatric cardiac care unit in a tertiary care university hospital. The subjects were 40 patients with AKI requiring PD between 2003 and 2005, and controls matched by type of surgery and body weight. Acid-base variables, including blood gas data and electrolytes, were assessed. The Stewart-Figge variables, including strong ion difference apparent (SIDa), strong ion difference effective (SIDe), and strong ion gap (SIG), were calculated. RESULTS: Blood gas analyses showed that the PD group was more acidemic, with a lower mean bicarbonate and a lower mean base excess, typical features of metabolic acidosis. The strong ion analyses revealed that the PD group had lower mean sodium and albumin concentrations. Based on the Stewart-Figge methodology, SIDa was smaller in the PD group than in the control group, but SIG was similar in the two groups. Receiver-operating characteristic curve analyses showed that serum albumin was the only prognostic factor associated with PCCU mortality, even after adjustment for PD treatment. CONCLUSION:Patients with AKI requiring PD in a pediatric cardiac care unit had significant metabolic acidosis compared to controls matched by the type of surgery and body weight. Hyponatremia and hypoalbuminemia were characteristics of these patients. The calculated SIDa was smaller in the PD than in the control group. Only the serum albumin had a significant prognostic value.