A comprehensive search was conducted in PubMed/MEDLINE, Scopus, Web of Science, and Cochrane Library, enhanced by a manual investigation of reference lists from pertinent articles. The search proceeded from January 2010 to June 2025.

Keywords and MeSH terms included the following: peritoneal dialysis, neonate, infant, newborn, acute kidney injury, extremely low birth weight, catheter, complications, outcomes.

French equivalents were also used: dialyse péritonéale, nouveau-né, insuffisance rénale aiguë, prématuré, faible poids de naissance, cathéter, complications, pronostic.

Original research articles, case series, clinical trials, guidelines, or systematic reviews on PD administered in the neonatal period.

• Publications reporting at least one of the following: PD indications, techniques, catheter types, complications, or clinical outcomes.

• Studies that included preterm and/or extremely low birth weight (ELBW) infants.

• Articles published in English or French.

• Studies exclusively addressing other kidney replacement therapies without neonatal PD data.

• Studies exclusively addressing other kidney replacement therapies without neonatal PD data. The initial search yielded 222 records. After removing duplicates, 218 titles and abstracts were screened. Of these, 179 were excluded because they did not meet the inclusion criteria. Ultimately, 33 studies were included in the qualitative synthesis.

The application of standardized criteria, such as the KDIGO Modified Neonatal AKI Classification (Table I), has significantly enhanced the identification and study of neonatal AKI over the past two decades ⁹. With these enhanced techniques, studies have found that AKI affects almost one in three newborns admitted to NICUs, although reported rates vary across geographic regions, from about 8% to 24% overall and between 3.4% and 4.2% in different centers in India [10,11]. The additional rates differ depending on the newborn’s gestational age, with AKI affecting about 48% of extremely preterm infants, 18% of late preterm infants, and 37% of term neonates ³¹². Notably, large international studies, like the AWAKEN cohort, have found that neonatal AKI is strongly linked to a higher risk of death, longer hospital stays, prolonged mechanical ventilation, and a higher risk of developing chronic kidney disease later in life ¹³.

Due to these serious complications, it is crucial to identify high-risk infants promptly. The STARZ (Sethi Tibrewal Agarwal Raina waZir) score is a practical bedside tool for this purpose ¹³. Severe AKI in the STARZ system corresponds to KDIGO Modified Neonatal AKI Classification stages 2–3 (Table I), which usually requires PD; specifically, a STARZ score ≥ 66 points has been proposed as a threshold predicting this severe form of AKI and the possible need for dialysis ¹³. The score has high sensitivity (92.8%), specificity (87.4%), and accuracy (89.4%), and it can be used even in low- and middle-income settings to help with early management ¹.

Such advancements around standardizing definitions and risk scores have improved the consistency of neonatal AKI research and have enabled clinicians to make treatment decisions more rapidly and accurately.

As previously mentioned, neonatal AKI is a relatively common occurrence associated with longer hospital stays, high morbidity and mortality, a greater need for mechanical ventilation, and an increased risk of chronic kidney disease (CKD) ²¹⁴. Neonatal kidneys are particularly vulnerable to hypoperfusion due to their limited glomerular flow, high plasma renin activity, and high renal vascular resistance, leading to reduced sodium reabsorption, impaired filtration, and low intracortical perfusion ¹⁵. These perfusion abnormalities, related to the renin-angiotensin system and prostaglandins, also make neonates more susceptible to nephrotoxic drugs such as ACE inhibitors and NSAIDs ².

Nephron development starts in the fifth week of gestation and ends around 34-36 weeks, with 60% occurring during the third trimester ¹⁶. The final nephron number varies greatly and increases by about 200,000 per additional kilogram of birth weight (BW) ¹⁶¹⁷. Consequently, low birth weight (LBW) and prematurity reduce nephron endowment and raise the risk of AKI and future CKD. After birth, renal maturation continues for about two years, with rising renal blood flow, decreasing vascular resistance, and progressive achievement of an adult-level glomerular filtration rate (GFR) ².

Neonatal AKI is often triggered by low BW, perinatal events (e.g., infections, hypoxia), nephrotoxic drugs (loop diuretics, aminoglycosides), thrombosis, and congenital anomalies ¹⁸ (Table II). Many of these risk factors are specific to the neonatal period and are rarely seen in older pediatric age groups ⁵. Indeed, prerenal factors account for 76%–80% of cases ⁸¹⁸.

As intimated above, etiological patterns vary between countries and even between centers from middle-income countries ¹⁹. In high-income countries, the main causes are sepsis, post-cardiac surgery ischemic/hypoxic injury, and nephrotoxicity, particularly in preterm infants (42.2%) and those with congenital heart disease (CHD) (11.7%) ¹⁹. In a multicenter Turkish study, AKI was obstructive, intrinsic, and pre-renal in 23.9%, 72.9%, and 3.2% of infants, respectively ²⁰. Similarly, Gerçel et al. ⁶ reported inborn errors of metabolism (12.9%), asphyxia (25.9%), prematurity (24.7%), sepsis (8.2%), necrotizing enterocolitis (NEC) (5.9%), dehydration (5.9%), hydrops fetalis (5.9%), CHD (5.9%), and renal anomalies (4.7%) as the main causes. In premature infants, Okan S et al. ²¹ also reported patent ductus arteriosus, NEC, sepsis, hypoxia, and hydrops fetalis as causes of neonatal AKI.

The two primary KRT modalities for newborns are HD/hemofiltration and PD ²². Since its introduction in the late 1950s, PD has remained the most widely used option for neonatal KRT ²², as it is less expensive and more accessible than HD, particularly in resource-limited settings ⁵²³. Because PD uses the peritoneum as a semipermeable membrane, it does not require vascular access, anticoagulation, or large extracorporeal circuits, which are often challenging in neonates ⁸²³²⁴. PD also allows for the gradual removal of fluids and solutes, reducing the risk of hemodynamic instability ⁸²³.

PD is especially suitable for premature and low birth weight (LBW) infants when vascular access is difficult or impossible ²²²⁵²⁶. In some settings, it is the only immediately available option ³¹⁸. While PD is widely favored in low- and middle-income countries due to the limited availability of trained personnel and equipment for extracorporeal dialysis [10,18,27], PD is contraindicated in neonates with pleuroperitoneal connections, diaphragmatic hernias, recent abdominal surgery, intra-abdominal infection, NEC, or any defects affecting the integrity of the abdominal or peritoneal wall ¹⁸²². When PD is not feasible or must be discontinued, miniaturized continuous kidney replacement therapy (CKRT) systems such as Carpediem can be considered ²⁷.

While PD is effective in managing AKI, it is less efficient than HD or CKRT at rapidly reducing plasma ammonia in severe hyperammonemia ²⁸. Nevertheless, it remains useful for AKI linked to metabolic disorders or mild hyperammonemia ²⁹ and for fluid removal in neonates with heart disease (those post-cardiac surgery or on extracorporeal membrane oxygenation (ECMO)), with fewer anticoagulation-related complications ²²²⁶.

PD in neonates has several specific challenges, such as obtaining appropriately sized catheters, ensuring correct exchange volumes, and compensating for immature peritoneal function ⁶²⁹. Standard catheters often need adaptation for LBW infants ⁶. Capillary leak syndrome in septic neonates may hinder ultrafiltration but can improve solute clearance ²²³⁰. Furthermore, both dialysate volume and dwell time influence PD efficiency: High volumes with short dwell times enhance ultrafiltration, whereas longer dwell times improve solute diffusion ³⁰.

Economically, PD is more cost-effective than its alternatives—amounting to about one third the cost of HD and one fourth the cost of continuous arteriovenous hemofiltration ³¹. In one study, 64% of infants receiving PD survived versus 32% with CKRT ³².

AKI remains a serious and life-threatening problem in NICUs. When conservative treatments, such as diuretics, fluid resuscitation, or inotropes, fail to correct fluid overload, persistent oliguria, electrolyte imbalance, or uremia, KRT becomes necessary ¹⁰³³. As mentioned earlier, among KRT modalities, PD is particularly suitable for preterm and LBW infants.

Oliguric AKI is the most common indication: It was reported in 61.3% of cases by Matthews et al. ³¹, 68.8% by Hakan et al. ²⁰, and 84% in a more recent study ⁵. Other major causes include sepsis (identified in 43% of cases in Tian ⁷), hypoxic-ischemic encephalopathy, congenital metabolic disorders, postoperative cardiac surgery, and severe dehydration ²⁴. In a Turkish series ⁵, the main causes were acute tubular necrosis (69.2%), inborn errors of metabolism (19.2%), congenital nephrotic syndrome (3.9%), bilateral polycystic kidneys (3.9%), obstructive uropathy (1.9%), and renal agenesis (1.9%).

Sinha et al. ³⁴ noted that many AKI cases were multifactorial, often involving sepsis (75%), nephrotoxic antibiotics, hypoxic-ischemic encephalopathy (21%), and the use of diuretics or enalapril in congenital heart disease (CHD) (25%). The main clinical triggers for initiating PD in such contexts are fluid overload, refractory metabolic acidosis, and resistant hyperkaliemia. Notably, not all neonates with congenital anomalies of the kidney and urinary tract (CAKUT) require long-term or chronic dialysis; some only need temporary support during acute decompensation ³⁵.

Metabolic indications include inborn errors such as lactic acidosis and hyperammonemia. Matthews et al. ³¹ found these errors accounted for 35.5% of secondary PD indications, versus 19.2% in other series ⁵.

In neonates undergoing cardiac surgery, early PD initiation reduces the time to achieve negative fluid balance, shortens mechanical ventilation, decreases inotropic support, and improves survival ²². Other reported indications include nonimmune hydrops fetalis and severe salt overload (“salt poisoning”) ²³.

Although NEC is often considered a relative contraindication, it is reported that carefully monitored PD with small fill volumes and short dwell times can be safely used in selected NEC cases with AKI, particularly in postoperative settings, allowing improved fluid balance and renal recovery without increasing the risk of bowel perforation or peritonitis ³⁶.

The choice and placement of PD catheters are critical for managing AKI in neonates. In neonates, PD catheters are typically 8-10 cm long for ELBW infants and 12-16 cm for term neonates, with one third to half the length being inserted inside the peritoneum to prevent kinking and ensure flow ³⁸³⁹. Because of their fragile anatomy, especially in ELBW infants, catheter placement can be technically challenging. However, when appropriate devices and techniques are used, PD can be performed safely and effectively.

The three main types of catheters used are rigid catheters (stylet-type), flexible catheters such as a Tenckhoff catheter (straight or swan-neck) and Cook PD catheters, improvised catheters (pigtail catheters, angiocatheters, intercostal drainage tubes, feeding tubes, Foley catheters, chest drains) ^3102223. The International Society for Peritoneal Dialysis (ISPD) strongly recommends flexible catheters because they offer better inflow/outflow rates, lower leakage, and reduced risks of peritonitis or visceral injury ³¹¹²². The Tenckhoff catheter is the most widely used and preferred for PD patients more than 5 days old, due to its double-cuff design, subcutaneous tunnel, and durability.

When Tenckhoff catheters are unavailable or too large, alternatives have been used successfully, including Blake silicone drains (with side channels) ³⁹, Arrow 14-gauge vascular catheters ³⁰, central venous catheters ²⁹, and improvised IV cannulas or suction catheter tips (which are less expensive, easier to insert, and generally safe in fragile neonates) ²³.

In very small infants, cutting down Tenckhoff catheters may be required, but this can roughen the tip and reduce flow by decreasing side holes ²⁹³³.

Catheter insertion can be surgical (mini laparotomy) or percutaneous (bedside, with guidewire) ³⁷³⁸. Surgical placement is preferred for optimal visualization, but percutaneous insertion is feasible under sterile conditions. Lateral exit sites (right/left lower quadrant) are recommended over median sites to reduce leakage risk ³⁷. The exit site should be kept clean and dry with a daily inspection ³⁷. Prophylactic IV antibiotics (usually a first-generation cephalosporin) are recommended at insertion to reduce infection risk, and heparin (500 IU/L) may be added to dialysate during the first 24-48 h to prevent fibrin obstruction ³⁴.

Despite technical advances, catheter-related complications, such as leakage, blockage, displacement, and infection, remain common, especially due to poorly sized devices ⁸²⁹. The short distance between the umbilicus and symphysis pubis (3-6 cm) in neonates further complicates the placement of standard Tenckhoff catheters ²⁹³³. Nonetheless, when appropriate catheters and preventive measures are used, PD is safe and effective: Ao et al. ²⁹ reported successful PD in 10 neonates using central venous catheters with no complications and full renal recovery at discharge, and Tian et al. ⁸ confirmed the importance of preventive strategies to reduce catheter-related complications.

Several factors influence the effectiveness of neonatal PD, including the infant's size and weight, the compliance of the peritoneal cavity, respiratory status, and the severity of uremia ¹⁰.

Dialysate volumes should be tailored to the infant’s size, starting at 10–20 mL/kg per cycle and gradually increasing to a maximum of about 800 mL/m² of body surface area as tolerated while closely monitoring abdominal distension and respiratory function ³⁷³⁸. Because neonates, especially ELBW infants, have relatively hyperpermeable peritoneal membranes, short dwell times (20–30 minutes) are recommended initially to optimize ultrafiltration and prevent rapid reabsorption; longer dwell times can be used once the patient is more stable ³⁸³⁹.

A 2.5% dextrose dialysate (standard commercial PD solution containing 2.5% glucose as the osmotic agent) is commonly used in the early phase to enhance ultrafiltration by increasing osmotic pressure and drawing water from the peritoneal capillaries into the dialysate ²²³⁷.

Because only small fill volumes are used, manual exchange systems are preferred, and a graduated burette system (Buretrol™) is particularly useful to measure small instilled and drained volumes (as low as 5-10 mL), reducing the risk of fluid overload ³⁷³⁸. Closed gravity-based systems are ideal because they minimize contamination, but open systems with a three-way connector and burette can be used in resource-limited settings if commercial systems are unavailable ²².

While PD is a life-saving therapy for AKI in neonates, it is associated with complications that can influence short-and long-term outcomes.

Catheter-related issues are the most frequent complication, occurring in 20%-60% of cases ²⁹. These include dialysate leakage, exit-site infections, obstruction, blockage, and displacement. In Kara et al.'s study ⁵, 20.6% had leakage, 8.8% blockage, and 5.9% exit-site infection. Gerçel ⁶, meanwhile, reported 8.2% catheter complications (namely drainage failure, hernia, wound dehiscence, leakage). Notably, though, a Tenckhoff catheter with double cuffs and a subcutaneous tunnel significantly reduces these risks ²³. Ustyol ²⁴ found 46.7% complications with automated peritoneal dialysis (APD), but all cases were manageable and nonfatal.

Hyperglycemia was the most common metabolic issue (47.1% in Kara et al. ⁵); other issues included hypokalemia and local exit-site infections ⁸¹¹. These complications are usually manageable and rarely lead to treatment interruption.

Peritonitis is one of the most feared complications of PD in newborns. Reported incidence rates vary widely, but studies suggest they are comparable to or even higher than those seen in adults undergoing PD ³⁷. Kara et al. ⁵ reported an 8.8% rate of peritonitis (due to Klebsiella pneumoniae, Staphylococcus epidermidis, Acinetobacter baumannii), all of which were treated successfully with intraperitoneal and systemic antibiotics. Matthews et al. ³¹ found a 12.9% rate, and Elgendy et al. ³² found an 8.4% rate (close to the 7.9% leakage rate). Peritoneal fluid should always be cultured, and empiric therapy should target both Gram-positive and Gram-negative organisms. Catheter removal is reserved for severe or resistant infections (e.g., fungal or Pseudomonas) or when no improvement occurs within 3–5 days ⁵. Improvised catheters such as IV cannulas are associated with a higher risk of peritonitis, likely due to increased leakage ²³.

Several studies have confirmed that early PD improves outcomes in neonatal AKI. Tian et al. ⁸ observed rapid improvements in kidney function, urine output, electrolytes, acid-base status, and infection markers within 1-5 days of starting treatment. Yildiz ¹⁸ similarly reported the reversal of fluid overload, hyperkaliemia, hypernatremia, and acidosis. Gerçel ⁶, meanwhile, showed PD was feasible across a wide range of gestational ages and birthweights (average duration 11.6 ± 13.7 days).

Despite these positive signs, mortality rates remain high: Kata et al. ⁵ reported a mortality rate of 76.9% (mainly due to multiorgan failure, sepsis, inborn errors, and cardiac failure), while mortality rates of 26.7% in Yildiz ¹⁸, 36% in Elgendy ³², 60% in S. Tangirala ¹¹, and 67.1% in Gerçel ⁶ were also reported. Risk factors include delayed PD initiation, hemodynamic instability, and multi-organ failure. Among survivors, 83.3% had full renal recovery, but some developed moderate CKD or hypertension with proteinuria ⁵. Matthews et al. ³¹ found that 5 of 12 survivors required chronic PD and were awaiting kidney transplantation, suggesting that about 15%–20% may progress to end-stage renal disease (ESRD), while the remaining roughly 80% recover normal kidney function.

Some survivors also presented long-term comorbidities, such as developmental delay, growth retardation, or neurological sequelae ⁶¹⁸³². Interestingly, survival may be better with PD than with CKRT: Elgendy et al. ³² found a 64% survival rate with PD versus 32% with CKRT. Early PD after ischemic AKI following congenital heart disease surgery was also associated with improved survival ¹⁸.

AKI is common and severe in premature neonates, especially those with an ELBW (< 1000 g). Because their kidneys and other organs are immature, this population is highly susceptible to sepsis, hypoxemia, nephrotoxic drugs, and hypoperfusion. AKI affects up to 56% of ELBW infants, with renal hypoperfusion and ischemia accounting for about 18% of cases ¹⁹.

HD is rarely feasible in this population due to difficulty obtaining vascular access, hemodynamic instability, and the need for large extracorporeal volumes, making PD the preferred option [2,10]. Indeed, studies have shown that PD can safely correct fluid and electrolyte imbalances in unstable preterm infants, especially when started early ⁸³⁰.

ELBW infants have fragile abdominal walls and a very small peritoneal cavity. In Noh’s study ¹⁰, infants had a mean birth weight of 706.5 g and 27.2 weeks’ gestation. Standard PD catheters are often unsuitable, as straight catheters in thin-walled abdomens are linked to higher risks of intestinal perforation, leakage, and obstruction ⁵²³. Makeshift devices, such as IV cannulas, chest tubes, and vascular catheters, have been used with variable success ¹⁰³⁰. Leakage is commonly caused by poor catheter anchoring and the lack of subcutaneous fat ²³. Furthermore, most APD machines require minimum fill volumes (> 100 mL) that exceed the safe intraperitoneal capacity of many ELBW infants ¹⁰²³.

Despite PD’s potential benefits, mortality remains high in this vulnerable group. Noh et al. ¹⁰ reported mechanical complications in 67% and a mortality rate of 91.7%, while Kara et al. ⁵ reported an 81.3% mortality, mainly from sepsis, multiorgan failure, and intestinal complications. Other studies have shown mortality rates ranging from 69%–80%, and up to 95% with multiorgan failure ⁸⁴⁰. Still, Ustyol et al. ²⁴ reported successful outcomes in neonates weighing as little as 580 g when PD was started early and managed carefully. Tian ⁸ also observed clinical improvement without adverse effects in a small preterm cohort.

Overall, the evidence shows that starting PD in ELBW infants is an ethically complex decision, given the risk of long-term complications, multiple surgeries, and prolonged hospitalization. Ideally, neonatologists, nephrologists, and ethics specialists should collaborate to balance the benefits of aggressive intervention against long-term quality of life ¹⁰.

Samira Tizki: conception of the article, data collection, data analysis and drafting of the manuscript

Abdelatif Daoudi: supervision, methodology, critical revision of the manuscript

N/A

N/A

Authors did not receive any funding for this article.

Authors declares that they do not have conflict of interest with this article.

ORCID iDs

S. Tizki https://orcid.org/0000-0001-8829-571X

A. Daoudi https://orcid.org/0000-0002-3391-4090