This study reviewed the 18-year experience of acute dialysis in the pediatric intensive care unit, in order to identify factors that could predict outcome, and to determine whether newer modalities of acute dialysis have influenced this outcome. Sixty-six children (ages 1 day to 19 years) received acute dialysis from May 1980 to April 1998. Factors predicting outcome were analyzed using univariate and Cox regression analysis. Modality of dialysis in the first 15 years was exclusively peritoneal dialysis. with a mortality of 63.9%. However, in the last 3 years, with increasing patient numbers, continuous hemodiafiltration (CHDF) was the modality of choice (56.7%), with a mortality of 73.3%. Univariate analysis showed that age <1 year, coma, acute tubular necrosis, disseminated intravascular coagulopathy, assisted ventilation, and hypotension were associated significantly with poor outcome (P<0.05). Cox regression analysis revealed that mortality was significantly higher in patients on mechanical ventilation (RR 5.96, 95% CI 1.82-19.50), or with age <1 year (RR 2.00, 95% CI 1.08-3.73). In conclusion, despite the increasing use of CHDF over the last 3 years, there was no significant improvement in mortality, probably related to the fact that more critically ill patients were dialyzed.

Chronic peritoneal dialysis (PD) for children in Singapore was instituted in 1988 at the National University Hospital with adult nurses providing dialysis services during the first 10 years. In 1998, a specialist pediatric dialysis nursing team was recruited. This study was conducted to determine the impact of dialysis nursing service on PD-related outcomes during the two nursing periods. Comparing the adult (group 1) and pediatric (group 2) nursing periods, the peritonitis rate was significantly higher in group 1 (RR 1.90; 95%CI 1.27-2.84), and this association did not weaken after adjusting for age, gender, and exit site infections. Exit site infection rate (RR 2.16; 95%CI 1.44-3.23), risk of peritonitis during the first year (RR 3.65; 95%CI 1.68-7.90), and multiple peritonitis attacks (RR 2.45; 95%CI 1.32-4.55) were higher in group 1. The peritonitis rates for adult patients cared for by the same adult nurses declined sharply from 1.05 episodes per patient-year between 1989 and 1992 to 0.41 episodes per patient-year between 1995 and 1997, however the corresponding pediatric rates did not change (1.48 to 1.06 episodes per patient-year, respectively) until the second era when specialized pediatric nurses were available. In conclusion, establishment of a specialist pediatric dialysis nursing team resulted in significant improvement in infection-related PD outcomes.

Computerized kinetic modeling is a valuable automated peritoneal dialysis (APD) prescription tool for optimizing dialysis adequacy. However, non-compliance results in failure to achieve adequacy targets. The aim of this study was to determine if a nomogram could estimate dialysis compensations for shortfalls in simulated non-compliant patients, such that total weekly urea clearance (Kt/V(urea)) targets are met. Individualized nomograms comprising a series of curves were derived from PD Adequest (ver. 2.0)-predicted Kt/V(urea) data (r (2 ) > 0.99) for different APD prescriptions. The nomogram was then used to estimate the (Nomogram-computed) average of the daily Kt/V(urea) in 14 patients. The study comprised three 1-month phases. Patients were compliant to dialysis in phase I, where Adequest-predicted Kt/V(urea) showed good agreement with both measured (r (I) = 0.72), and Nomogram-computed values (r (I) > 0.99) (p < 0.001). Conversely, in non-compliant phase II, Nomogram-computed values were lower than Adequest-predicted values (p< 0.002). In phase III, the nomogram estimated prescription adjustments required to compensate for shortfalls, such that there was significantly less difference between Nomogram-computed and Adequest-predicted Kt/V(urea) than in phase II (p = 0.005). Thus, despite non-compliance, predicted Kt/V(urea) targets were attained using the nomogram to adjust the daily APD prescriptions. This concept is potentially useful for patients desiring to compensate for inadvertent shortfalls, rather than for 'truly non-compliant' patients.

No abstract