Last updated: Lesson of the Month - August 2014…
on 04 Aug 2014

Peritoneal Dialysis (PD)


Lisa Crowley and Martin Wilkie

PD is (predominantly) a home based therapy for patients with end-stage renal disease, and is reliant on the patients' own peritoneal membrane acting as the dialysis membrane. It has the advantage of allowing patients more independence in managing their own disease; and has benefits in terms of quality of life (Thong, 2008). It is often an excellent first choice for patients starting dialysis, particularly when they still have some native renal function (Chaudhary, 2011). PD regimes are designed on a much more individualised basis than patients on HD.

While the fluid is in situ, solutes move from the patients blood, across the peritoneal membrane, down the concentration gradient into the dialysate fluid. An osmotic gradient is created by high concentration of glucose (occasionally amino acid or glucose polymer solutions are used) in the dialysate fluid, which removes water from the patient. Acid-base balance is regulated by the absorption of lactate or bicarbonate (or combinations) from the dialysis fluid.

Clearance of urea and creatinine by PD is dependent on the volume of dialysate drained at the end of a session. Clearance can be maximised by increasing the volume of the fluid exchanged and by ensuring the exchange time is optimised to allow equilibration between plasma solute concentrations and the dialysate. Water removal is adjusted by altering the dialysis dwell time and by altering the glucose concentration of the dialysate.

The latest ISPD guidelines can be found here and the Renal Association here.

Types of PD

Peritoneal dialysis is a technique whereby infusion of dialysis solution into the peritoneal cavity is followed by a variable dwell time and subsequent drainage. Ansari (2011) has reviewed the use of PD in AKI, but it is more commonly used to treat advanced CKD. There are several types:

Continuous ambulatory peritoneal dialysis (CAPD)
This is a continuous treatment usually consisting of four to five dialysis exchanges per day. These are usually 2 litres. Exchanges are performed at regular intervals throughout the day, with a long overnight dwell. Diurnal exchanges last 4 to 6 hours, and the nocturnal exchange remains in the peritoneal cavity for 6 to 8 hours.

Automated peritoneal dialysis (APD)
This is a continuous treatment carried out with an automated cycler machine. Multiple short-dwell exchanges are performed at night. 10-12L are usually exchanged, over 8-10 hours. There are lifestyle advantages to this technique, ie it leaves the daytime free. It can also be used to increase the dose of dialysis, or water removal in certain patient groups.

Tidal peritoneal dialysis
The cycler is programmed to leave a percentage of the preceeding dwell in the peritoneal cavity before the next fluid instillation during night-time cycles. So the peritoneal cavity is never free of fluid. It is useful for patients with abdominal pain between cycles, and those with frequent 'low drain' alarms overnight; but generally does not improve clearances, which was the original aim of its development.

Suitability for PD

Most patients will be suitable for PD catheter insertion, however it may be contraindicated in the following patients:

  1. BMI. Morbidly obese patients may be unsuitable, but many overweight patients are managed on PD very successfully
  2. Previous abdominal surgery and diverticular disease. This can cause intra-abdominal adhesions and make insertion of the catheter technically difficult or impossible
  3. Abdominal stoma
  4. Abdominal/inguinal hernias. Will need repairing if the patient would like PD
  5. The patient (or relative) will need to be able to manage the technical aspects of PD

Advantages of PD

PD may be associated with better preservation of residual renal function (RRF). This may account for better early outcomes on PD compared to HD (Jansen, 2002; Moist, 2000).In 2006, Vonesh and colleagues concluded that survival on PD is equal if not better than survival on HD in non-diabetic and younger diabetic patients.

Key point: in PD, residual renal function may be better preserved than in HD.

Initial PD use preserves possible sites for vascular access for later in the patient's 'renal career'. There is some evidence that suggests initial treatment with PD, then subseqent transfer to HD, has a superior outcome to a single dialysis modality. There is little evidence that PD leads to better longterm outcomes when compared directly to HD. Studies that suggest that this true, usually do not take comormidity into account (Miskulin, 2002).

Epidemiology and Outcomes

According to the Renal Registry (Steenkamp, 2010), there were 49,080 adult patients receiving RRT in the UK in 2010, equating to a UK prevalence of 794 pmp (64 pmp on peritoneal dialysis). Transplantation was the most common treatment modality (48%), HD was used in 44% and PD in 8% of RRT patients.

The age adjusted 1 year survival for patients on PD in 2007 was 94.5%, compared to 87.3% for patients on HD, continuing the trend in improvement on preceding years. These figures do not take into account the relative co-morbidities of the two patient groups; and in these patients outcome is determined more by co-morbidity than modality.

Prescription and Principles

Prescribing Peritoneal Dialysis 

Dialysis Dose

The optimal dose of dialysis is not defined in the same way as it is for patients on HD. Indeed:

Key point: the relationship between small solute clearance and outcome is much more closely related to an individual’s residual renal function than the ‘dose’ of PD.

Clearly patients on PD need to receive adequate ‘dialysis’ doses, but many other important factors contribute to patient well-being, including infection rates, nutritional status and ultra-filtration volume. Even though PD is individualised to a much greater extent than it is for patients on HD, there are minimum recommended doses:

Key point: the Renal Association recommends that ‘a combined urinary and peritoneal Kt/Vurea of ≥1.7/week or a creatinine clearance of ≥50L/week/1.73m2 should be considered as minimal treatment doses’

The dose of dialysis should be increased in any patient experiencing uraemic symptoms. Adjusting doses of PD is highly technical and you should always consult a senior PD nurse, if you are considering this issue

Residual Renal Function

The recommended dose considers the total renal and peritoneal contribution to clearance. Theoretically, renal and peritoneal clearances have been considered equivalent. However, this does not seem to be the case, since renal clearance strongly correlates with survival, while peritoneal clearance does not (Bargman, 2001).

This is part of the reason, some nephrologists try hard to preserve RRF in PD patients by various methods: avoiding ACEi and ARBs, NSAIDs and other nephrotoxic drugs; and by using high dose diuretics.

This is not to say that peritoneal clearance is not important. In fact, studies in anuric patients have shown a clear correlation between survival and dose of PD. It is likely that since renal clearance is associated with many other important physiological functions, such as volume regulation and hormone synthesis, the favourable effects from these additional benefits, mask the effects of solute removal.

Key point: the importance of preserving RRF cannot be over-emphasised.

Body Size

Body mass is associated with increased generation of creatinine and urea, and by definition affects normalised clearances. It is logical that patients with more mass will require higher doses of PD.

Peritoneal Transport

One of the predominant determinants of small solute clearance is peritoneal transport. Faster transporters, will have higher clearances, all other prescription factors being the same. However, peritoneal glucose absorption is similarly increased and consequently the osmotic gradient diminishes resulting in reduced ultrafiltration in high transport states.

This may be why earlier studies suggested that high peritoneal transport states are associated with a higher risk of death than low and low-average transport. More recently, it has been suggested that possible adverse effects of a higher transport status (eg lack of ultrafiltration) have been mitigated, possibly by the use of APD and icodextrin
(Davies, 2006).

Peritoneal Equilibration Test (PET)

Twardowski developed the PET (peritoneal equilibration test) in 1987 (Twardowski, 1987). This is a method of measuring the rate at which solute moves across the peritoneal membrane. It was the first standardised method to quantify individual peritoneal membrane characteristics and to compare the individual results with larger populations.

It involves infusion of a 2.27% glucose dialysate for a 4 hour period. Dialysate samples are taken at 2 and 4h and analysed for creatinine and glucose. The ultrafiltration volume is also then measured.

Key Point: the Renal Association recommends regular monitoring of the dialysis membrane function with a PET test, 6 weeks after starting PD, then annually thereafter.

Key Point: in a PET test, a dialysate ultrafiltration volume of <100 ml suggests ultrafiltration failure.

The results of this test will classify the patient into 1 of 4 groups according to the function of the membrane:

  • High transporters (15% of patients) - higher clearance, lower ultrafiltration, more suited to APD
  • High average transporters (35%)
  • Low average transporters (35%)
  • Low transporters (15%) - lower clearance, higher ultrafiltration, more suited to CAPD

Dialysis prescription is initially based on the class of membrane. For example, patients with high transport membranes ('high transporters') reabsorb glucose quickly from the dialysis fluid, so shorter dwell times and higher glucose concentrations are needed to avoid unwanted fluid reabsorbed. One aim of improving dwell times for higher transporters is to reduce glucose use and exposure.These patients may be more suited to APD.

Whereas patients with low transport membranes ('low transporters') require longer dwell times and lower concentrations of glucose in the dialysate as glucose absorption is much slower. These patients may be more suited to CAPD.

The PET becomes increasingly important once residual renal function has declined and the patient has less urine output .

Treating the Patient as a Whole

Even though achieving minimum adequacy standards is part of good PD care, it is important to look at the patient as a whole. For example, if they are 80 years old and have comorbidities that are likely to shorten survival, 'accepting' less than perfect adequacy data, may be appropriate; especially as there is little evidence of higher small solute clearances leading to longer survival. Also, if the patient is 'well', responding to an ESA, has reasonable PTH control, for such a patient, their dialysis is adequate.

Non-adherence in PD

Analyses of PD fluid delivery data, has shown that in upto 30% of patients, one or more PD exchanges are regularly not done. Therefore, 'increasing' the patient's dialysis dose may not be the appropriate response to 'poor' adequacy data. Education may be preferable.

Peritoneal Membrane

The peritoneal membrane is a complex heterogenous, semi-permeable membrane with multiple pores. The early models of peritoneal membrane transport included multiple sites of resistance to the flow of solutes across the membrane. These included the capillary fluid film overlying the capillary endothelium, the capillary endothelium per se, the endothelial basement membrane, the interstitium, the mesothelial cells and the fluid overlying the peritoneal membrane. Newer concepts such as the Three Pore Model suggest that the major resistance to peritoneal transport is in capillary endothelium and its basement membrane.

Water and Solute Removal, and the Three Pore Model

The Three Pore Model is a theoretical model validated by clinical observations. It suggests that the peritoneal capillary is the critical barrier to trans-peritoneal transport. Solute and water transport across the peritoneal capillary is mediated by pores of three different sizes.

  • Large pores (100-200 A) exist in small numbers and constitute < 0.1% of all pores. They are probably clefts within endothelial cells. They allow passage of macromolecules such as protein into the dialysate
  • Small pores (40-60 A) are more numerous, and represent the majority of the total pore area. They are gaps between endothelial cells. They are believed to transport small solutes (removal of urea and creatinine, and absorption of dialysate (eg glucose)); accounting for part of the fluid removal by ultrafiltration
  • Ultra-small or transcellular pores (4-6 A) are water channels or aquaporin-1, and are present in the endothelial cells of the peritoneal capillaries. They are numerous and resemble the water channels present in red blood cells and renal proximal tubules. They transport water only (sieving); accounting for approximately 50% of water removal by the osmotic effect of dialysate glucose

Three pore model

Water Removal - via Ultrafiltration and Osmosis

In PD, ultrafiltration is achieved via osmosis – either crystalloid (with glucose) or colloidal (with icodextrin).

During ultrafiltration in PD, and unlike HD, solutes do not move across the membrane in direct proportion to their concentration in blood. Sodium is held back or sieved at the aquaporin barrier while water moves through. Sieving makes ultrafiltration a less effective form of convective solute transport.

Solute Removal, via Diffusion

The semi-permeable peritoneal membrane allows solutes and water to be transported from the vascular system to the peritoneal cavity and vice versa through diffusion. Diffusion is the process of solutes moving from an area of higher concentration to an area of lower concentration, which is the case when dialysate is instilled into the peritoneal cavity. Actually, solutes move randomly in both directions, but there is simply more solute moving from the high to low concentration side than in the opposite direction. Eventually, the concentrations become equal on both sides of the membrane. This is termed equilibrium.

The movement of solute molecules is random and driven by thermal energy. This energy is proportional to absolute temperature (degrees Centigrade above -273). This thermal energy is transferred to kinetic energy which is the multiple of mass and velocity. Since this energy is the same for different sized molecules at the same temperature, the larger molecules must move more slowly in order to have the same energy as the smaller molecules. Thus, the diffusive rate depends on molecular weight.

Factors Affecting Solute Clearance

Solute clearance is influenced by the membrane permeability and size, characteristics of the solute, the volume of dialysate instilled, and blood flow to the membrane. Solute transport can be increased by maximising the contact of dialysis solution with the membrane, by placing the patient in a supine position or increasing the exchange volume.

Key point: to increase solute clearance, exchange volume needs to be increased.

Factors Affecting Water Removal

Increasing the glucose concentration of dialysis solution enhances fluid removal by increasing the osmotic gradient between the plasma and the peritoneal fluid. The higher the glucose concentration, the higher the fluid removal. The osmotic gradient is always greatest at the beginning of the dialysis exchange. As osmotic equilibration is achieved, the gradient decreases. Some reabsorption of fluid occurs when dialysate dwells beyond the point of equilibration.

Key point: to increase water removal, glucose concentration need to be increased.

Osmotic Agents

The osmotic agent normally used in PD fluid is glucose. It is not an ideal osmotic agent, as it is readily transported across the peritoneum. Osmosis is the movement of water across a semi-permeable membrane from an area of low solute concentration to an area of high solute concentration. The hydrostatic pressure gradient and the osmotic gradient between the blood and the dialysis solution influence osmosis.

Three main types of glucose concentration are available for use:

  1. 1.36 g/dl (%) (sometimes called a 'weak bag'). Patients often call it a 'yellow bag'. This fluid contains 27g of glucose in a 2 litre bag. Glucose concentration = 76 mmol/L. It should be considered the default fluid in PD. Longterm use of stronger bags is thought to contribute to ultrafiltration failure
  2. 2.27 g/dl (%) ('medium bag'). Patients often call it a 'green bag'. This fluid contains 45g of glucose in a 2 litre bag. Glucose concentration = 126 mmol/L
  3. 3.86 g/dl (%) ('strong bag'). Patients often call it an 'orange bag'. This fluid contains 77g of glucose in a 2 litre bag. Glucose concentration = 214 mmol/L. This very hypertonic fluid is usually used in patients with volume overload

Notes: Extraneal (Icodesxtrin) is a 'purple bag' and Nutrineal (amino acid) is a 'blue bag'

Continued use of strong bags could result in removal of 7-10 L per day which can be dangerous since it can induce haemodynamic instability due to massive ultrafiltration; and high glucose content of hypertonic of PD fluid may be deleterious to the peritoneal membrane - and is thought to contribute to ultrafiltration failure. Substances that are absorbed from dialysis solution into the systemic circulation include glucose and calcium. The increased concentration of dextrose in dialysis solution causes glucose to move into the systemic circulation.

For all of these reasons, alternative osmotic agents like Icodextrin have been developed.

Alternative Osmotic Agent: Icodextrin

Key Point: Icodextrin is preferable to using higher glucose concentrations. This is important since improved fluid management has been shown with its use; and better metabolic and technique survival in patients with diabetes (Takatori, 2011):

Icodextrin, a glucose polymer, can be used as an alternative to glucose. It acts at the small intercellular pores and is only slowly lost from the peritoneal cavity; so the fluid removal is sustained, and ultrafiltration is improved. This process is called colloid osmosis. It allows for a higher concentration of osmotic agent. The most commonly used type is 7.5 g/dl (7.5%) Icodextrin, which is a cornstarch derived glucose polymer. It is used for a single long dwell overnight in CAPD, or the day dwell in APD. It is also used in other types of medication such as chemotherapy.

When used in PD fluid, 20-30% of the icodextrin is absorbed to the systemic circulation and is metabolised to oligosaccharides such as maltose, maltotriose and maltotetrose. This leads to a small fall in serum sodium and increase in osmolality or osmolar gap in plasma, which is not of clinical significance. These metabolites are not available to the human cells as a source of glucose. Icodextrin metabolites, especially maltose, can react with some glucometer enzymatic reactions to give falsely high results (with risk of undiagnosed hypoglycaemia).

Key point: glucose polymer can falsely elevate blood glucose levels (risking undiagnosed hypoglycaemia).

The website gives all the relevant information regarding this issue.

Fluid Removal by Different PD Fluids

This has been summarised by Mujais in 2002 (see graph below). As can be seen different fluids will lead to different net UF volumes, depending on how long the fluid is left in:


Electrolytes and Buffers in PD Fluid

Calcium Concentration

The major factor influencing the systemic calcium absorption is the amount of ionised calcium in the plasma and in the dialysis solution. Commercially available PD solutions contain 1–1.75 mmol/1 of calcium. In the UK, 2 calcium levels are commonly used:

  1. 1.75 mmol/L ('traditional dialysate') - ie, hypercalcaemic to the blood
  2. 1.25 mmol/L ('low calcium dialysate') - ie normocalcaemic to the blood

1.75 mmol/L Ca dialysate
Since the normal concentration of diffusible ionised calcium is 1.1-1.3 mmol/1, calcium is absorbed or lost depending on the diffusive gradient direction. A 1.75 mmol/1 calcium-containing solution was traditionally used to provide an additional source of calcium in uraemic patients in whom the vitamin D deficiency usually leads to reduced calcium absorption in the bowel. The clinical use of phosphate binders containing calcium salts and vitamin D exposes patients in dialysis to a risk of hypercalcaemia that in turn could lead to vessel and soft tissue calcification (eg calciphylaxis).

1.25 mmol/L Ca dialysate
With the use of a low calcium-containing solution, an increased amount of phosphate binders containing calcium salts can be used. However, the serum calcium concentration should be carefully monitored, since low serum levels worsen uraemic osteodystrophy. Most UK units now use this fluid.

Sodium Concentration

The sodium concentration in the ultrafiltrate during peritoneal dialysis is usually less than that of extracellular fluid, with a concentration of 132-3 mmol/L. There is a tendency toward water loss and the development of hypernatraemia.The reason for the tendancy to hypernatraemia during rapid cycling is due to the effect of sodium sieving. Dehydration rarely occurs – apart from in low transporters.  Most anuric patients on PD remove approx 100 mmol of sodium per day or less.

Potassium Concentration

There is usually no potassium on PD dialysate. Why? Potassium is cleared by peritoneal dialysis at a rate similar to that of urea. With chronic ambulatory peritoneal dialysis and 10 L of drainage per day, approximately 35 to 46 mmol of potassium is removed per day. Daily potassium intake is usually greater than this, yet significant hyperkalemia is uncommon in these patients. Presumably potassium balance is maintained by increased colonic secretion of potassium and by some residual renal function. Given these considerations, potassium is not routinely added to the dialysate.

Buffer, and Correction of Metabolic Acidosis

The buffer present in most commercially available peritoneal dialysate solutions is lactate. In patients with normal hepatic function, lactate is rapidly converted to bicarbonate, so that each mmol of lactate absorbed generates one mmol of bicarbonate. Even with the most aggressive peritoneal dialysis there is no appreciable accumulation of circulating lactate. The rapid metabolism of lactate to bicarbonate maintains the high dialysate-plasma lactate gradient necessary for continued absorption.

Commercially available PD solutions contain 35–40 mmol/L of lactate. With the 35 mmol/L lactate-buffered solution ('normal lactate' solution), a moderate to mild acidosis is usually recorded in the majority of patients. With the 40 mmol/L lactate-buffered solution ('high lactate'), the percentage of patients with normal acid-base status increases.

Use of bicarbonate as a buffer had been precluded due to instability and reaction with metal ions such as calcium and magnesium during storage. Newer twin-chamber designs have overcome this problem. And there are now bicarbonate-buffered solutions, and combinations of lactate and bicarbonate. Bicarbonate is separated from divalent metal ions during storage, and mixed shortly prior to infusion. A 34 mmol/L bicarbonate-buffered solution improves blood bicarbonate control, when compared to a 35 mmol/L lactate-buffered solution. Bicarbonate-buffered fluid is mainly used to improve peritoneal pain or discomfort experienced by some patients during the infusion of the lactate-buffered solutions.

The pH of commercially available peritoneal dialysis solutions is purposely made acidic by adding hydrochloric acid to prevent glucose from caramelising during the sterilisation procedure. Once instilled, the pH of the solution rises to values greater than 7.0. There is some evidence that the acidic pH of the dialysate, in addition to the high osmolality, may impair the host’s peritoneal defenses.

The catabolic effects of metabolic acidosis on protein and amino acid metabolism can only be reversed by a full correction of this condition. In a randomised study on 200 CAPD patients, normal venous bicarbonate (27.2 mmol/L) was associated with nutritional benefits such as an increase in body weight and midarm circumference, and decreased morbidity as compared with mildly low venous bicarbonate (23.0 mmol/L) (Stein, 1997).

Summary of PD Fluid Content

Substance Concentration
Sodium 132-3 mmol/L
Potassium 0 (-2) mmol/L
Calcium 1-1.75 mmol/L
Magnesium 0.25-0.75 mmol/L
Chloride 95-102 mmol/L
Lactate 35-40 mmol/L
Lactate/bicarbonate 25/10-15 mmol/L
Bicarbonate 34-39 mmol/L
Osmotic Agent  
Glucose 1.36-3.86 g/dl
Amino acids 1.1 g/dl
Icodextrin 7.5 g/dl

Solute concentrations in PD fluid have been reviewed by Dombros (2005)


PD catheter

Cuffed Permanent Catheter

Most PD catheters are of this type, which is often called a Tenckhoff catheter. This was invented by Henry Tenckhoff in Seattle, in 1968. It has a low risk of infection, and a low risk of bowel perforation. In a patient with advanced CKD, it is usually left for 10-14 days before it is used. In AKI, it can be used immediately. The haemodynamic status of the patient may limit the feasibility of this catheter, especially in sick ITU patients with AKI.

The commonest type is a 'straight Tenckhoff'. There is also a 'coiled catheter' design, which provides an increased size of catheter tubing, and may provide better dialysate flow; with less inflow pain, risk of migration and omental obstruction. Though the evidence for these claims is limited.

There are four catheter insertion techniques:

  • Surgical dissection and placement under direct vision (mainly done by a surgeon, under GA)
  • Blind placement by Tenckhoff trocar or Seldinger technique (mainly nephrologist, under LA)
  • Peritoneoscopic placement
  • Laparoscopic placement

Pre-operative Care

Pre-operatively the patient needs to have been prescribed laxatives (several days before insertion) as this will make catheter insertion easier. These should be continued post-operatively. Most units will also prescribe prophylactic antibiotics, but the protocols vary between units - so check the local policy in your hospital. The exit site should be marked by the doctor performing the procedure. Finally, it is important to ensure the patient has emptied their bladder before catheter insertion to avoid bladder perforation.

PD Insertion

The catheter is introduced through the anterior abdominal wall into the peritoneal cavity. Catheters are usually made from silicon or polyurethane. The catheter should be angled down, with the tip located in the patients pelvis. The catheter has one or two Dacron cuffs that are inflated in the subcutaneous tissues that help prevent leakage and infection and hold the catheter in place. Dialysis can be carried out immediately after the catheter is inserted, but if possible it should be left for 10-14 days, before use.

PD catheter position

This is a satisfactory position for a PD catheter. But other positions are acceptable, if it works. A 'good catheter' is one that works (see complications).

Post Insertion Care

Immediately after catheter insertion it is important to watch for the following warning signs:

  • Faecal matter in the PD effluent (indicating bowel perforation). Laxatives should be continued post-operatively
  • Urine draining from the PD catheter (indicating bladder perforation)
  • Excess blood in the PD effluent. Small amounts of blood is normal, but large amounts can indicate intraperitoneal bleeding
  • Post-operative peritonitis - this is noted by the presence of cloudy effluent.  It is important to manage the exit site very carefully to minimise the risk of infection.  Management is carefully laid out in the ISPD catheter placement guideline (the same as the Renal Association catheter guideline)
  • Exit site infection
  • Fluid leak from around the catheter site
  • Post-operative ileus

After successful insertion of the catheter the patient is usually discharged home, either that day, or the following day. They are usually brought back as an outpatient 10-14 days later, to be educated on how to use their PD catheter and dialyse independently at home. Though it is probably safe to use low volume APD (1-1.5L) in a recumbent position (to reduce chance of a leak) immediately after insertion, in selected patients (eg if very 'uraemic' and treatment needs to be started soon).

Dressing changes should be minimised in the initial period post-insertion, with the initial dressing left unchanged for a week, unless infection is suspected. The catheter should also be immobilised with tape or dressings to reduce trauma due to movement.

Infective Complications

Infection is the commonest complication of PD and can take the form of peritonitis, exit site infection and catheter tunnel infection. Figueiredo (2010) has reviewed good catheter and exit site care


Peritonitis typically presents as abdominal pain; and cloudy dialysate is drained from the abdomen. Fluid should be sent for analysis of cell count, gram stain and culture

Key point: the diagnosis is confirmed by a white cell count > 100/μl; with > 50% neutrophils indicative of peritonitis; NB an excess of lymphocytes may suggest TB peritonitis

Key point: the most commonly isolated organism is Staphylococcus epidermidis. But more severe infections can be caused by S. aureus; or Gram negative organisms like Pseudomonas or Enterobacteriacae

If peritonitis is suspected, then empirical antibiotics should be started as soon as fluid samples and bloods (including blood cultures) have been sent

Key point: the minimum recommended duration of antibiotic therapy is two weeks, extended to three weeks for more serious or slowly resolving infections

The Renal Association suggests empirical antibiotic therapy should cover gram positive and gram negative organisms, including Pseudomonas. Individual units will have their own protocols, so always refer to your local guidelines. It is common practice for PD peritonitis to be treated with intraperitoneal antibiotics, with the exchanges containing the antibiotics left to dwell for 6 hours

Unusual Organisms and Other Causes of a Cloudy Dialysate

Occasionally peritonitis will be caused by fungi (which has a high mortality) - Candida for example. 'Culture negative peritonitis' may be due to prior antibiotic use, or fastidious organisms like mycobacterium. There is also a condition known as 'eosinophilic peritonitis', which is form of allergic reaction to the dialysate. This diagnosis should be considered if the Gram stain and cultures are negative and more than 10% of peritoneal leucocytes are eosinophils. Antibiotics are not necessary when treating eosinophilic peritonitis

Chylous ascites produces a creamy white dialysate, with a very high triglyceride content. It arises from a blockage or damage to intra-abdominal lymphatics. It can be associated with malignany involving the lymphatics, including lymphoma. It may occur intermittently, related to fat-containing meals

Malignancy can also present as a cloudy dialysate and may be identified by radiological imaging or presence of malignant cells on cytology of used dialysate

In the past, certain batches of icodextrin caused a sterile peritonitis, attributed to contamination with high peptidoglycan concentrations. This is now very unusual

When investigating PD peritonitis, remember that an abdominal X-ray is largely unhelpful as there may be free air under the diaphragm even in the absence of peritonitis (PIC). This may be normal in a patient on PD, and does not indicate a perforated abdominal viscus

'Surgical' Peritonitis

It is important not to forget 'normal' surgical peritonitis in the differential diagnosis of PD peritonitis. Causes include perforated abdominal viscus, pancreatitis, appendicitis and diverticulitis. Clues will include brown or dark PD dialysate and mutiple enteric/anaerobic organisms. The patients are usually, but not always, sicker than in PD peritonitis. If in doubt, a laparoscopy or laparotomy are indicated

NB: when investigating pancreatitis, serum amylase may be misleadingly low in patients receiving icodextrin, due to the interference of its metabolites with the amylase assay

Catheter Removal

Decisions on whether the catheter needs to be removed or not, should be reviewed on an individual basis. But if the patient has fungal, pseudomonal or TB peritonitis, or if the patient is unwell or if they are not improving after 1 week of treatment (regardless of the organsism), then the catheter should be taken out

Exit Site Infection

Exit site infections present as erythema of the skin surrounding the exit site of the PD catheter, with or without purulent discharge. Cultures should be sent immediately, before considering antibiotic therapy. If the patient is otherwise well they may not need antibiotics immediately, and you can wait until cultures become available. If there is any indication of a severe (or recurrent) infection then empirical antibiotic therapy (refer to local guidelines) should be started straight after cultures have been sent

As with peritonitis, two weeks of therapy is required; with further extension if resolution is slow or incomplete. Antibiotic prescribing will differ between centres, but the Renal Association guidelines suggest covering for S. aureus and P. Aeruginosa, which are important organisms

Catheter Tunnel Infection

Tunnelled line infections may present in a similar way to exit-site infections, but there may also be pain and swelling of the subcutaneous tissues surrounding the catheter. Ultrasound scan will reveal collection(s) of fluid around the catheter. Treatment is similar to that for exit site infections, but often do not respond as well. It is usually necessary to remove the catheter. Remember, both exit-site infections and tunnelled-line infections can progress to peritonitis and should be treated seriously, with thorough clinical examination and investigation



Other Complications

Other complications will be discussed in the following sections:

  • Mechanical problems
  • Long-term changes to the peritoneal membrane
  • Ultrafiltration failure and inadequate dialysis
  • Metabolic and other problems

Mechanical Problems

Drainage Problems

Under normal circumstances a 2 litre bag of dialysate should take about 15 minutes to run in and 20 minutes to drain out (although patients know how long it takes them, and will usually know when something is different). Although the catheter tip is positioned in the pelvis, it can migrate - causing problems with drainage. The catheter can also become obstructed with kinks, fibrin or blood clots, but constipation is also a major cause of poor drainage – particularly in patients new to PD 


There is no such thing as malposition. Even though there is a desired position for a PD catheter (pointing down, and in the pelvis), if it is in any other position, and works without complication, that position is satisfactory

This is a malpositioned catheter, if it does not work


There are three types of fluid leak:

Fluid leaks
Pericatheter leaks may occur early or late after tube insertion, and may be precipitated by coughing, straining or heavy lifting. Patients should be advised against these actions; which is part of the reasons why laxatives are prescribed before and after catheter insertion

Abdominal wall leaks
Abdominal wall oedema may arise by leaking through the catheter site, or previous abdominal incisions. There may be induration or spongy feeling of the abdominal wall skin, with a peau d'orange appearance 

Genital oedema
This is an uncommon but distressing complication of PD. Mechanisms include fluid tracking around the catheter or via an incision site via the subcutaneous tissues, or via a hernia or patent processus vaginalis

Management of fluid leaks
Leaks are treated with a period of rest from PD, to allow the leak to heal. PD should be recommenced initially with lower volumes, or with omission of the ambulant day dwell, if the patient is on APD. Persistent leaks may require catheter removal or surgical repair


Herniae are relatively common complications of PD. Pre-existing hernie should always be repaired before starting PD. This can be at catheter insertion. They may occur after starting PD, as a complication of the increase in intra-abdominal pressure. Sites include: inguinal, para-umbilical/umbilical, catheter insertion site, other incision sites, and ventral herniae. In cases of Gram-negative peritonitis, one should always consider the possibility of bowel strangulation as the hernial orifice will not always be obvious

Management is via surgical repair. This may require a period of discontinuation of PD, and temporary HD. Though, some patients, with adequate residual renal function, may allow a period without dialysis. They should be reviewed at least weekly, with biochemistry measured. It is also possible to resume PD very soon after repair, especially if a mesh repair is performed, with initial use of low volumes of dialysate in the supine position


Fluid can also leak from the peritoneal space through a diaphragmatic hernia, into the pleural space. This will present as a pleural effusion. Any pleural-fluid glucose concentration greater than serum is considered to be highly supportive of PD-related hydrothorax. There is no reliable 'cut-off' figure. An isotope scan can also help with the diagnosis

Management includes therapeutic pleural aspiration, and temporary discontinuation of PD. Few patients will be able to go back to PD. Occasionally a period of rest and recommencement of PD with initially smaller volumes may be successful. If the effusion recurs, pleurodesis may allow ongoing treatment with PD


Long-term Changes to the Peritoneal Membrane

Long term PD is known to change the characteristics of the peritoneal membrane. This is thought to be due to bioincompatible elements in the dialysis fluid, particularly breakdown products of glucose. These changes can alter the properties of the membrane, effecting the efficiency of dialysis.Therefore, the Renal Association recommends regular monitoring of the dialysis membrane function with a PET test (see above) 6 weeks after starting PD, then annually thereafter

Encapsulation peritoneal sclerosis (EPS)

Encapsulation peritoneal sclerosis (EPS) has a overall mortality of approximately 50% - but this is variable. The older term 'sclerosing peritonitis' is now out of favour, mainly because it is not necessarily a 'peritonitis'. In this condition a thick fibrotic layer engulfs the contents of the abdomen, which causes a cocoon surrounding the peritoneum and binding the intestine - causing bowel obstruction.

Key point: EPS also presents in more cryptic ways - eg abdominal pain, ascites, weight loss, malnutrition, haemoperitoneum or diarrhoea. Although typically occuring whilst on PD, it can present after the patient has been transferred to HD, or transplanted. The diagnosis should be considered in any ESRD patient with unusual abdminal symptoms

It is more common with increasing length of time on PD and levels of exposure to glucose in dialysate. It is also thought to occur in some cases in response to peritonitis. There is very little data on how to treat this condition, and various treatment strategies have been tried, including surgical procedures and immunosuppression, but there is currently no agreed best practice

We would recommend referral to a tertiary referral centre in the UK that has experience in the management of this condition – eg Manchester (Titus Augustine) or Cambridge (Chris Watson). There is emerging evidence for the benefit of peritoniectomy and adhesiolysis

Ultrafiltration Failure and Inadequate Dialysis

Failure of ultrafiltration in PD refers to the inability of the technique to remove sufficient volumes of fluid to maintain fluid balance. This is more of a problem in patients who no longer produce urine, ie after 18 months from the start of dialysis

Initially, this can be managed by imposing a fluid restriction (patients tend not to like this!), or prescription of diuretics to boost residual urine output. Care must be taken however to avoid intravascular volume depletion and dehydration

Patients who are ‘high transporters’ may be more at risk of ultrafiltration failure, because they tend to absorb glucose from the dialysis fluid, and therefore reabsorb more water by osmosis. If this is deemed to be the reason for failure of ultrafiltration, then faster exchanges using a cycler machine combined with icodextrin used for the long dwell, may be of benefit. In patients with ultrafiltration difficulties, it is important to avoid absorbtion from any of the exchanges; and this requires careful scheduling of the cycles based on the patients transport characteristics

Finally, long-term changes to the peritoneal membrane (see above) from long-term PD can affect water movement across the membrane. Dialysis adequacy should be regularly reviewed, and is assessed in terms of solute clearance

Metabolic and Other Problems

Glucose Absorption

Weight gain and hyperglycaemia are particular problems associated with long-term PD. This can be particularly problematic for diabetic patients who may require specialist advice to deal with the increased glycaemic load – often requiring insulin therapy


Presence of blood-stained dialysate can be alarming for the patient. It is commonly a transient self-limiting event of unknown cause, possibly due to catheter trauma on the peritoneal membrane. It also occurs in relation to the menstrual cycle, at times of ovulation or menstruation. Occasionally it occurs as a feature of peritonitis. Rare causes include: abdominal malignancy, pancreatitis, hepatobiliary disease, bleeding from a polycystic kidney, liver disease, or EPS

Protein Losses in PD (vs Urinary Losses in Nephrotic Syndrome)

The amount lost varies from patient to patient but averages between 5-15g per day (humans eat approximately 80g in a Western diet). Serum albumin is not low in most PD patients, and this tells us something about nephrotic syndrome. In other words, as patients with nephrotic syndrome have a low serum albumin (by definition) and a urinary loss often less than 5g a day. Whereas PD patients have higher PD losses but run a normal serum albumin. This means that there must be other reasons why the serum albumin is low in nephrotic syndrome. It is known that hepatic albumin synthesis is decreased in nephrotic syndrome, which explains part of the hypoalbuminaemia

Key point: there is a significant amount of protein lost in dialysate

Protein loss stabilises and remains relatively constant unless the patient experiences peritonitis. Then the protein loss increases during the infection. It is very important that protein intake is adequate in the PD patient. Daily protein requirements average 1.2-1.5 g/kg of body weight. Other substances lost in dialysate are amino acids, water-soluble vitamins, hormones and some medications



Top Tip: The Importance of Preserving RRF in PD Patients Cannot be Over-emphasised

  1. All patients approaching the need for dialysis should be considered for PD (including CAPD, APD and assisted APD)
  2. The clinical effectiveness (mortality and morbidity) of PD is equal to conventional thrice weekly haemodialysis
  3. PD has been shown to preserve residual renal function for longer than conventional haemodialysis. The relationship between small solute clearance and outcome is much more closely related to an individual’s residual renal function than the ‘dose’ of PD. Therefore, in PD, the importance of preserving RRF cannot be over-emphasised
  4. A PET will define the patients’ peritoneal membrane characteristics which will show the rate at which glucose is absorbed through the membrane (and thus the ultrafiltration capability) and the rate at which solutes are cleared
  5. Renal Association recommends a minimal treatment dose of ‘a combined urinary and peritoneal Kt/Vurea of ≥1.7/week or a creatinine clearance of ≥50L/week/1.73m2'
  6. Peritonitis is the commonest infective complication. The diagnosis is confirmed by a white cell count > 100/μl (with > 50% neutrophils). Patient education and re-training are essential to ensure PD peritonitis is minimised
  7. Treatment for PD peritonitis is usually by intra-peritoneal and oral antibiotics according to local protocol based on ISPD guidelines
  8. Icodextrin, often used for the long dwell of peritoneal dialysis, can interfere with glucose monitoring equipment. It’s therefore important to inform patients and nurses to use the appropriate glucose testing monitor (see
  9. Permanent PD access is vital for effective technique. PD catheters can be inserted by local or general anaesthetic; no technique or type of catheter being shown to be superior
  10. As a self-care treatment, PD patients should be supported at home by regular communication and education from specialist nurses







Abbott WE, Shea P. The treatment of temporary renal insufficiency by peritoneal lavage. Am J Med Sci 1946; 211: 312

Ansari N. Peritoneal Dialysis in Renal Replacement Therapy for Patients with Acute Kidney Injury. Int J Neph 2011

Balázs J, Rosenak S. Zur Behandlung der Sublimatanurie durch peritoneale Dialyse. Wien Med Wchnschr 1934; 47: 851

Bargman JM et al. Relative Contribution of Residual Renal Function and Peritoneal Clearance to Adequacy of Dialysis: A Reanalysis of the CANUSA Study. JASN 2001; 12(10): 2158-2162

Blackfan KD, Maxcy KF. The intraperitoneal injection of saline solution. Am J Dis Child 1918; 15: 19

Boen ST, Mion CM, Curtis FK, Shilipetar G. Periodic peritoneal dialysis using the repeated puncture technique and an automatic cycling machine. Trans Am Soc Artif Intern Organs 1964; 10: 408-14

Boen ST. Peritoneal Dialysis in Clinical Medicine. Springfield, Illinois, Charles C. Thomas, 1964

Carpenter S et al. An Interlinear Transliteration and English Translation of Portions of Ebers Papyrus. Possibly Having to Do With Diabetes Mellitus (2006)

Chaudhary K, Sangha H, Khanna R. Peritoneal Dialysis First: Rationale. CJASN 2011;  6(2): 447-456

Churchill DN, Taylor DW, Keshaviah PR, and the CANUSA Peritoneal Dialysis Study Group. Adequacy of dialysis and nutrition in continuous peritoneal dialysis: association with clinical outcomes. J Am Soc Nephrol 1996; 7:198–207

Clark AJ. Absorption from the peritoneal cavity. J Pharmacol 1921; 16: 415

Cunningham RS. Studies on absorption from serous cavities, III. Am J Physiol 1920; 53: 488

Davies SJ. Mitigating peritoneal membrane characteristics in modern peritoneal dialysis therapy. Kidney International 2006; 70: S76–S83

Diaz-Buxo JA, Farmer CD, Walker PJ, Chandler JT, Holt KL. Continuous cyclic peritoneal dialysis: a preliminary report. Artif Organs 1981; 5(2): 157-61

Doolan PD et al. An evaluation of intermittent peritoneal lavage. Am J Med. 1959; 26(6): 831-44

Dombros N. Peritoneal dialysis solutions. Nephrol Dial Transplant 2005; 20 [Suppl 9]: ix16–ix20

Engel, D. Uber die ausspulbarkeit von toxin, antikorpern und artfremden eiweis durch des lebende
peritoneum [The rinsability of toxin, antibody and foreign protein through the living peritoneum]. Centralblf Bacteriologie 1927; v103: 423-427

Figueiredo A, Goh BL, Jenkins S, Johnson DW, Mactier R, Ramalakshmi S, et al. Clinical practice guidelines for peritoneal access. Perit Dial Int 2010; 30: 424–9

Ganter G. Ueber die Beseitigung giftiger Stoffe aus dem Blute durch Dialyse. Munch Med Wochschr 1923; 70: 1478

Grollman A, Turner LB, Levitch M, et al. Hemodynamics of bilaterally nephrectomized dogs subjected to intermittent peritoneal lavage. Am J Physiol 1951; 165: 167

Grollman A, Turner LB, McLean JA. Intermittent peritoneal lavage in nephrectomized dogs and its application to the human being. Arch Intern Med 1951; 87: 379

Hales S. A method of conveying liquors into the abdomen during the operation of tapping. Phil Trans Roy Soc 1753; 43: 20-21

Hamburger HJ. Ueber die Regelung der osmotischen Spannkraft von Flüssigkeiten in Bauch-und Pericardialhöhle. Archiv f Phys du Bois-Reymond 281, 1895 

Heusser H, Werder H. Untersuchungen über Peritonealdialyse. Bruns Beiträge z klin Chir 1927; 141: 38

ISPD: The Emergence of Peritoneal Dialysis, 1987

Jacobs C, Broyer M, Brunner FP, Brynger H, Donckerwolcke RA, Kramer P, et al. Combined report on regular dialysis and transplantation in Europe, XI, 1980. Proc Eur Dial Transplant Assoc 1981; 18: 2–58

Jansen MA, Hart AA, Korevaar JC, et al. Predictors of the rate of decline of residual renal function in incident dialysis patients. Kidney Int 2002; 62: 1046-1053

Klapp R. Ueber Bauchfelresorption. Mitt Grenzgeb der Med u Chir 1902; 10: 254

Kop PSM. Peritoneaal dialyse (Peritoneal dialysis). Thesis for obtaining the doctorate in medicine at the University of Groningen University, on 7 July 1948

Krediet RT. 30 years of peritoneal dialysis development: the past and the future. Peritoneal Dialysis International 2007; 27 (Supplement 2); S35-41 A later history of PD

Leathes JB, Starling EH. On the absorption of salt solutions from the pleural cavities. J Physiol 1895; 18: 106

Li PK, Law MC, Chow KM, Chan WK, Szeto CC, Cheng YL, et al. Comparison of clinical outcome and ease of handling in two double-bag systems in continuous ambulatory peritoneal dialysis: a prospective, randomized, controlled, multicenter study. Am J Kidney Dis 2002; 40: 373-80

Maiorca R, Cantaluppi A, Cancarini GC, Scalamogna A, Broccoli R, Graziani G, et al. Prospective controlled trial of a Y-connector and disinfectant to prevent peritonitis in continuous ambulatory peritoneal dialysis. Lancet 1983; 2: 642-4

Mcbride PT. Milestones in the development of peritoneal dialysis, 1980?
Another history of PD

Miskulin DC et al. Comorbidity and other factors associated with modality selection in incident dialysis patients: the CHOICE study. Choices for healthy outcomes in caring for end-stage renal disease. Am J Kid Dis 2002; 39: 324-36

Moist LM, Port FK, Orzol SM, et al. Predictors of loss of residual renal function among new dialysis patients. J Am Soc Nephrol 2000; 11: 556-564

Mujais S, Vonesh E. Advancing Fluid Management in Peritoneal Dialysis: Profiling of peritoneal ultrafiltration. Kidney International 2002; 62: S17–S22

Nolph KD, Sorkin MI, Prowant B, Webb J. National Conference on continuous ambulatory peritoneal dialysis, Kansas City, MO. Perit Dial Bull 1981; 1: 65-6

Odel HM, Ferris DO, Power MH. Peritoneal lavage as an effective means of extrarenal excretion. A clinical appraisal. Am J Med 1950; 9: 63-77

Oreopoulos DG. Chronic peritoneal dialysis. Clin Nephrol 1978; 9: 165-73

Oreopoulos DG, Shahrzad Ossareh S, Thodis E. Peritoneal Dialysis. Past, Present, and Future. IJKD 2008; 2 (4): 171-182

Orlow WN. Einige Versuche über die Resorption in der Bauchhöle. Arch f Phys Pflüger 1895; 59: 170

Palmer RA, Quinton WE, Gray JE. Prolonged peritoneal dialysis fro chronic renal failure. Lancet 1964; 28; 1(7335): 700-2

Palmer RA. As it was in the beginning. History Of Peritoneal Dialysis. Peritoneal Dialysis International. 1980 A good history of PD

Popovich RP, Moncrief JW, Decherd JF, et al. The definition of a novel portable/wearable equilibrium peritoneal dialysis technique. Trans Am Soc Artif Intern Organs 1976; 5: 64 (Abstract)

Popovich RP, Moncrief JW, Nolph KD, Ghods AJ, Twardowski ZJ, Pyle WK. Continuous ambulatory
peritoneal dialysis. Ann Intern Med 1978; 88: 449-56

Putnam TJ. The living peritoneum as a dialyzing membrane. Am J Physiol 1922; 63: 548

Reid R, Penfold JB, Jones RN. Anuria treated by renal decapsulation and peritoneal dialysis. Lancet 1946; 2(6430):749-53

Rhoads JE. Peritoneal lavage in the treatment of renal insufficiency. Am J Med Sci 1938; 196: 642

Rosenberg, M. Uber stickstoffhaltiqe retentiosstoffe im blute und in anderen korpeflussigkeiten bei nephritikern [Nitrogenous retention of substances in blood and in other body fluids in the core of the kidneys]. Berl Klin Wochenschr 1923; v53: 548-565

Seligman AM, Frank HA, Fine J. Treatment of experimental uremia by means of perirtoneal irrigation. J Clin Invest 1946; 25: 211-9

Starling EH, Tubby AH. The influence of mechanical factors on lymph production. J Physiol (London) 1894; 16: 140

Steenkamp R et al. Renal Registry Report (2010). Chapter 2. UK RRT Prevalence in 2009: national and
centre-specific analyses

Stein A, Moorhouse J, Iles-Smith H, Baker F, Johnstone J, James G, Troughton J, Bircher G, Walls J.
Role of an improvement in acid-base status and nutrition in CAPD patients. Kidney Int 1997; 52(4): 1089-95

Takatori Y et al. Icodextrin Increases Technique Survival Rate in Peritoneal Dialysis Patients with Diabetic Nephropathy by Improving Body Fluid Management: A Randomized Controlled Trial. CJASN 2011; 6: 1337–1344

Tanret P. Sem Hop 1946; 22(30):1477-80

Tenckhoff H, Schechter H. A bacteriologically safe peritoneal access device. Trans Am Soc Artif Intern Organs 1968; 14: 181-186

Thong MSY, Kaptein AA. Quality of life in patients on peritoneal dialysis. ‘Nolph and Gokal’s Textbook on Peritoneal Dialysis’, USA: Springer, 2008

Twardowski ZJ, Nolph KD, Khanna R, et al. Peritoneal equilibration test. Perit Dial Bull 7:138-47, 1987

Vonesh EF, Snyder JJ, Foley RN, Collins AJ. Mortality studies comparing peritoneal dialysis and hemodialysis: what do they tell us? Kidney Int Suppl 2006; 103: S3-11

Von Recklinghaussen FT. Die Lymphgefässe und ihre Beziehung zum Bindegewebe. Berlin, Germany: Hirshwald. 1862

Warrick C. An improvement on the practice of tapping; whereby that operation, instead of a relief for symptoms, becomes an absolute cure for an ascites, exemplified in the case of Jane roman; and recommended to the consideration of the royal Society, by Christopher Warrick, of Truro, Surgeon. Philos Trans R Soc Lond. B Biol 1753; 43: 12–19
This article can be read on this hyperlink

Wear, JB, Sisk IR, Trinkle, AJ. Peritoneal levage in the treatment of uremia. J Urol 1938; v 39: 53-62

Wegner G. Chirurgische Bermekungen über die Peritonealhöle, mit besonderer Berucksichtigung der Ovariotomie. Arch Klin Chir 1877; 20: 51-55

Weston RE, Roberts M. Stylet-catheter for peritoneal dialysis. Lancet 1965; 1(7394): 1049


CARI, Australasia. Dialysis Adequacy. Oct 2005

CARI, Australasia. Peritonitis treatment and prophylaxis, 2004

CSN, Canada: Adequacy and Nutrition in PD, 1999

European Best Practice Guidelines for Peritoneal Dialysis. Dec 2005

UK, Renal Association: Peritoneal Dialysis in CKD. Woodrow G, Davies S. July 2010

UK, Renal Association: Peritoneal Access.Wilkie M, Jenkins S, Shrestha B. Sept 2009

US, KDOQI. Clinical Practice Guidelines and Clinical Practice Recommendations Updates. Peritoneal Dialysis Adequacy, 2006


Pictures of early dialysis equipment in the 'Home dialysis central website: museum'