Temporary Peritoneal Dialysis in Newborns and Children: A Single-Center Experience over Five Years

Abstract

Aim: To evaluate the indications, complications, and outcomes of temporary peritoneal dialysis (TPD) in children with acute renal failure (ARF). Patients and methods: All patients undergoing TPD between February 2006 and January 2011 in a children’s hospital were included in the study. Patient characteristics, indications, complications, and duration of TPD (DPD), requirement of re-operation, length of stay, presence of sepsis, and outcome were recorded. Results: There were 21 newborns (14 prematures), 9 infants, and 9 children. The main nephrotoxic agents were gentamicin (n = 7), netilmisin (n = 5), vancomycin (n = 3), and ibuprophen (n = 3). Patients with multiorgan failure (n = 9) had significantly higher blood urea nitrogen (BUN) and creatinine levels than those without multiorgan failure (n = 30) [BUN: 94 ± 27.3 vs. 34.3 ± 4.9) and creatinine: 4.1 ± 0.8 vs. 1.9 ± 0.2)]. The mean DPD was longer in mature patients than in prematures (newborn: 3.7; children: 7.1). Nine complications were observed (23%) (leakage in three and poor drainage in six patients). Twenty-five patients (64.1%) responded to TPD treatment and were discharged, and 14 patients (10 newborns and 7 of them were premature) died (35.9%). Mortality rate was higher in prematures (n = 7) and patients with a history of nephrotoxic agent (n = 10). Conclusion: TPD is effective especially in neonates with ARF and it is a reliable alternative to the hemodialysis or other continuous renal replacement therapies but it is not free of complications. It has limited effects, particularly in patients with multiorgan failure.

• acute renal failure
• peritoneal dialysis
• newborn
• child
• peritonitis

INTRODUCTION

Acute renal failure (ARF) is a complex condition resulting from abrupt, reversible reduction in glomerular filtration rate (GFR) and needs the arrangement of fluid–electrolyte and acid–base balance. The treatment of ARF involves medical, nutritional, and dialysis methods.^1,2 Although both peritoneal dialysis and hemodialysis (HD) have been used in children for decades, recent continuous renal replacement therapies (CRRTs) are being increasingly used in children as well. However, these modalities are expensive and require highly trained staff and complex apparatus. Peritoneal dialysis has been used as an important modality in children with ARF for nearly 50 years. It has been a fundamental treatment for the newborns and infants who are hemodynamically unstable or at risk of bleeding and in patients with failure of vascular access.³ Since temporary peritoneal dialysis (TPD) has advantages as having ease of access and technical simplicity in comparison to HD and CRTTs, it is still the modality of choice in renal replacement treatment in children in developing countries.⁴

The aim of this study is to assess the demographic profile, complications, and outcome of newborns/children that were managed with TPD.

PATIENTS AND METHODS

From February 2006 to January 2011, 39 infants and children underwent TPD therapy (20 males, 19 females). We retrospectively reviewed age, gender, indications of TPD, complications of TPD, presence of sepsis, use of nephrotoxic agent, duration of TPD (DPD), length of stay (LOS), presence of peritonitis, and outcomes. We also recorded GFR, blood urea nitrogen (BUN), and creatinine values. The main indication for TPD was ARF, which could not be treated in a conservative manner and the primary diseases of neonates as well as children. The indications for TPD are depicted in Tables 1 and 2.

Table 1. Primary diseases of children.

	Inborn metabolic disorders	Sepsis	Sepsis and RDS	Asphyxia	CS	HUS	CKD
Newborn (n = 21) (53.8%)	7 (33.3%)	5 (23.8%)	3 (14.3%)	2 (9.5%)	3 (14.3%)		1(4.8%)
Infant (n = 9) (23.1%)	3 (33.3%)	4 (44.5%)			1 (11.1%)	1(11.1%)
Children (n = 9) (23.1%)	4 (44.5%)	3 (33.3%)			1 (11.1%)	1(11.1%)

Note: RDS, respiratory distress; CS, cardiac surgery; HUS, hemolytic uremic syndrome; CKD, cystic kidney disease.

Table 2. Main indications of PD.

	Fluid overload, hyperkalemia	Fluid overload, hypernatremia	Refractory acidosis	Hyperkalemia
Newborn	13		3 (1 with MD)	5
Infant	3	3	2 (1 with MD)	1
Children	5	1	3 (2 with MD)

Note: MD, metabolic disorder.

GFR was calculated by Schwartz formula. Teflon^® non-cuffed rigid catheters (B. Braun, Bethlehem, USA) were placed at bedside in all patients with ARF. Incision (∼1 cm) was located in lateral infraumbilical abdominal area. Catheter was inserted through this incision and preceded to the pelvic area.

RESULTS

Thirty-nine patients aged between 2 days and 8 years (median age, 25 days) were included in the study. There were 21 newborns, and 14 of them (66.7%) were preterms. Nine patients were infants (1–24 months), and the remaining nine were older children (>2 years old).

The patients were grouped according to their age, gender, maturity, use of nephrotoxic agent, organ failure status, and outcome (Tables 3 and 4). There was no difference in GFR, BUN, and creatinine measurements in terms of the age, gender, and maturity (p > 0.05). The LOS was higher in mature patients than in prematures. DPD was also higher in mature patients than in prematures (p < 0.05). Since there were higher number of deaths (n = 7) in premature neonates, LOS and DPD were longer in mature patients.

Table 3. The mean values of GFR, DPD, BUN, creatinine, and LOS according to gender, age, presence of nephrotoxic agent, maturity, organ failure, and outcome (mean ± standard deviation).

		GFR (mL/min/1.73 m^2)	Duration of PD (day)	BUN (mg/dL)	Creatinine (mg/dL)	LOS (day)
Gender	Male	20.8 ± 3.9	4.64 ± 0.8	55.5 ± 13.2	2.64 ± 0.5	21.5 ± 6.7
	Female	29.7 ± 7.6	4.68 ± 0.8	43.4 ± 11.9	2.35 ± 0.9	20.3 ± 3.7
Age	Newborn	19.9 ± 5.5	3.7 ± 0.7	41.5 ± 10.6	2.23 ± 0.3	18.1 ± 5.9
	Infant	41.2 ± 11.5	4.7 ± 1.2	51.1 ± 25.4	2.58 ± 0.9	26.9 ± 6
	Children	21.5 ± 3.8	7.1 ± 1.4	68.1 ± 10.5	3.11 ± 0.6	21.1 ± 3.9
Nephrotoxic agent	+	23.2 ± 3.7	4.3 ± 0.7	39.1 ± 6.4	1.94 ± 0.3	17.1 ± 3.2
	–	27.8 ± 8.1	5 ± 0.9	59.2 ± 16.3	3.0 ± 0.6	24.7 ± 6.6
Maturity	Premature	10.8 ± 1.3	2.85 ± 0.9	40.5 ± 7.9	1.87 ± 9.1	12.8 ± 4
	Mature	29 ± 5.3	5.1 ± 0.7	51.1 ± 10.7	2.64 ± 0.4	22.8 ± 4.4
Organ failure	Kidney	25.1 ± 3.8	4.1 ± 0.9	34.3 ± 4.9	1.9 ± 0.2	22.1 ± 3.2
	Multiorgan	21.8 ± 3.2	3.3 ± 0.7	94 ± 27.3	4.1 ± 0.8	19.6 ± 2.4
Outcome	Died	22.3 ± 3.1	2.63 ± 0.4	72 ± 13.6	3.9 ± 1.1	9.45 ± 2.7
	Survived	32.7 ± 4	6 ± 0.8	48.5 ± 11.3	2.4 ± 0.7	26.6 ± 5.4

Note: GFR, glomerular filtration rate; DPD, duration of TPD, BUN, blood urea nitrogen; LOS, length of stay.

Table 4. Complications related to TPD and their characteristics.

			Nephrotoxic drug		Age
		Prematurity	+	−	Newborn	Infant	Children	Total
Organ failure	Kidney	8	14	16	14	8	8	30
	Multiorgan	6	4	5	7	1	1	9
Sepsis	+	8	10	8	10	5	3	18
	−	6	8	13	11	4	6	21
Complication	+	3	5	4	5	1	3	9
	−	11	13	17	16	8	6	30
Repositioning	+	3	5	1	4	1	1	6
	−	11	13	20	17	8	8	33
Outcome	Died	7	10	4	10	2	2	14
	Survived	7	8	17	11	7	7	25

In terms of the exposure to nephrotoxic agents, there was no difference in GFR, BUN, and creatinine values. However, patients who were exposed to nephrotoxic agent had a shorter DPD.

There were nine patients with multiorgan failure. Seven of them were newborns, and six of them were premature (85%). Four patients with multiorgan failure had a history of exposure to nephrotoxic agent. In patients with multiorgan failure, BUN (94 ± 27.3) and creatinine levels (4.1 ± 0.8) were found significantly higher than those who had kidney failure only.

A total of 18 patients received nephrotoxic agent and 10 of them died (55.5%).

The nephrotoxic agents used in this study were gentamicin (n = 7), netilmicin (n = 5), vancomycin (n = 3), and ibuprophen (n = 3).

There were a total of nine catheter-related complications (23%). In three patients (7.6%), there was leakage and six patients (15.3%) had poor drainage. All leaking catheters and three of six catheters with poor drainage were revised. The remaining catheters were washed out only to reopen the plug.

None of the patients developed peritonitis and/or catheter infection. Eighteen (46.1%) patients had sepsis. Ten patients with sepsis were newborns, and eight of them (80%) were premature.

Fourteen patients (35.9%) died [10 newborns (71.4%), 2 infants (14.3%), and 2 older children (14.3%)]. Seven (50%) of all deaths were in preterms (p < 0.05).

DISCUSSION

ARF is defined as a sudden decrease in renal function resulting from retention of nitrogenous or non-nitrogenous waste products and concomitant reduction in urine output.^3,5,6 It is a common medical condition, especially in neonates and children in intensive care units.^7,8 Incidence of ARF in critically ill neonates was reported as high as 8–24%,⁹ whereas it was reported between 30.6% and 82% in critically ill children.^10–12

ARF is classified as prerenal (75%), intrinsic (15%), and obstructive (5%) subtypes.¹³ In children, the main causes of ARF are hypovolemia (hemorrhage, dehydration), asphyxia/anoxia, sepsis, nephrotoxic agents (especially aminoglycosides), and low cardiac output due to congenital heart disease and cardiac operations.^4,14 Infants’ kidneys, especially those of prematures, are also vulnerable to these conditions due to the incomplete maturation of nephrogenesis and immature vasoregulation.^15,16 Despite appropriate fluid and drug management, some cases progress and require renal replacement therapy. A recent study from Turkey revealed that 32.6% of all patients with ARF were newborns, and the most common associated conditions were prematurity (42.2%) and congenital heart disease (11.7%) in newborns, malignancy (12.9%) and congenital heart disease (12.3%) in children aged more than 1 month. In the same study, hypoxic/ischemic injury and sepsis were found as the leading causes of acute kidney injury in both age groups.⁷ In accordance with the literature, more than half of the patients were infants and 66.7% of them were preterm in our study. Sepsis was the leading cause of ARF.

The choice of the RRT treatment generally depends on the clinical practice, technical resources, and well-trained staffs of the center. RRT modalities such as peritoneal dialysis, HD, and CRRTs differ in the efficiency and their clinical tolerability. Peritoneal dialysis has many advantages when compared with HD and CRRTS. TPD does not require specialized equipment or vascular access and it is quickly placed at bedside. In addition, it does not need systemic anticoagulation. TPD provides continuously to correct volume and metabolic disturbance without causing hemodynamic instability and it has better cardiovascular tolerance.^17–19 Therefore, TPD is a feasible option for the treatment of selected patients with ARF, especially newborns and infants, and those who are hemodynamically compromised and have severe coagulation abnormalities, difficulty in obtaining vascular access, and inborn metabolic disorder. Our reason to choose TPD in our patients was that most of them were newborns and infants who had acute kidney injury and/or inborn metabolic disorders and were hemodynamically unstable.

However, TPD has also some limitations. Inadequate peritoneal cavity and clearance potential in certain conditions such as intra-abdominal shunts, intra-abdominal adhesions, and fibrosis limit the utility of TPD. Intraabdominal elevation of pressure due to PD may cause diaphragmatic restriction and worsening of respiratory function of the patients. Complications of TPD such as leak, obstruction with non-effective dialysis, and catheter-related infections are observed, and revision of catheter may be required. Similarly in our study, we encountered nine complications (poor drainage in six and leakage in three patients) and we had to revise six (15.3%) of the catheters. Fortunately, all these complications were managed successfully.

Although mortality is high among preterms and infants with ARF and/or inborn errors of metabolism, TPD is an effective modality for the emergent treatment of ARF when the HD and CRRT modalities are not available. Infants, particularly prematures, are primary candidates for TPD in developing countries.

Declaration of interest: There is no financial or commercial interest for the article. The work in this article has not already been published and has not been submitted simultaneously to any other journal.