Most centers use a two-cuff, silastic intraperitoneal catheter. Many different shapes are available and no shape has proven more efficacious than another.

Laparoscopic placement of the catheter is generally preferred to ensure proper positioning and allow repair of subclinical hernias, defects, adhesions, and redundant omentum. Percutaneous placement under ultrasound and fluoroscopy has also been effective. Open surgical placement of PD catheters is no longer routinely performed.

The intra-abdominal portion contains many side-port perforations to maximize fluid flow.

Although catheters can be used immediately after surgical placement, a preferred healing period of 10 to 28 days prior to initiation of dialysis allows healing of the exit-site and reduces the incidence of early subcutaneous leaks or infection.

The deep cuff is placed in the abdominal wall musculature and after healing secures the catheter in its position.

The subcutaneous superficial cuff allows for granulation tissue to form, creating an additional barrier to infection; however, externalization of this cuff may occur.

The exit-site may be placed anywhere and must be visible to the patient for hygiene and not where the catheter is subject to irritation (e.g., belt line).

Tunneled presternal catheters are associated with reduced infectious complications in obese patients.

The Y-set is the standard setup in manual PD (Fig. 21-1).

Infection risk is reduced as there is only one connection point (at the stem emerging from the patient) where sterile technique may accidentally be broken.

The flush-before-fill technique is where the patient first flushes air out of the tubing by allowing a small amount of dialysis solution to pass into the drain bag; the drain bag is then clamped and the rest of the dialysis solution is infused into the peritoneal cavity via gravity (10 to 15 minutes).

Between exchanges, the bags are disconnected and the stem is capped.

When the specified dwell time is complete, a new transfer set is attached and the peritoneal cavity is drained via gravity (20 to 25 minutes) before a new infusion is started, again with the flush-before-fill technique.

It is generally recommended that the catheter should not be submerged in water, although many physicians allow swimming in salt water.

Standard PD solutions contain sodium, chloride, lactate, magnesium, calcium, and varying concentrations of dextrose (Table 21-1).

Most commercial solutions contain a sodium concentration of 132 mEq/L to allow a net sodium diffusion; this protects against hypernatremia that might otherwise occur, particularly with the more hypertonic solutions.

Potassium is not present in standard solutions and hypokalemia is usually corrected with oral supplementation.

Phosphorus is also absent but PD is not able to remove the daily dietary load and most patients require both dietary restriction and oral phosphorus binders.

In the present era of wide-spread calcium-containing phosphorus binders, a PD solution with a lower calcium concentration (2.5 mEq/L) is preferred to prevent placing a patient in positive calcium balance.

Lactate has been the primary alkaline equivalent used in PD because of technical difficulty with bicarbonate in the manufacturing process; the resultant acidic pH of 5.5 can cause inflow pain and may be damaging to the mesothelium.

Newer dual-chamber bags with bicarbonate are available, allowing for mixing at the time of use, although they are not yet available in the United States.

The hypertonic dextrose concentration can be 1.5%, 2.5%, or 4.25%, with the higher concentrations providing a stronger osmotic force leading to more ultrafiltration.

Most commercially available solutions have color-coded tabs with which patients may be more familiar rather than the actual percentage of dextrose (Table 21-2).

A drawback of dextrose-containing solutions is that 60% to 80% can diffuse inwardly during a dwell, thereby dissipating the osmotic gradient, limiting the achieved ultrafiltration and giving the patient a glucose load.

Newer solutions with long-chain glucose polymers (icodextrin) allow for a more sustained colloid oncotic effect, roughly equivalent to the ultrafiltration of a 2.5% dextrose solution over 12- to 18-hour period. The metabolic products of icodextrin can lead to falsely elevated glucose measurements on some diabetic test strips and blood sugar results in patients on icodextrin must be interpreted cautiously.

Amino acids can also be used as an osmotic agent, particularly in malnourished patients, although frequent or long-term use may lead to a metabolic acidosis and elevations in the serum urea concentration; ultrafiltration capacity is limited, roughly equivalent to a 1.5% dextrose solution.

The membrane separating the blood in the capillaries from the solution in the peritoneal cavity consists of many layers.

Most resistance to solute movement occurs at the capillary endothelium (via filtration pores) and interstitium (which may be of variable thickness).

The total surface area in adults is about 2 m², consisting of both the parietal and visceral layers although most exchange is through the parietal peritoneum. Not all capillaries are readily available to participate in solute exchange as some may be too far from the cavity under baseline conditions.

Larger dwell volumes can stretch the membrane, recruiting more capillaries to participate in solute exchange. Peritonitis increases the vascularity of the membrane thus increasing peritoneal transport. Patients will often ultrafiltrate less during an episode of peritonitis.

The “three-pore model” describes large pores (>25 nm) through which proteins and other macromolecules pass, small pores (4 to 6 nm) for electrolytes and solutes like urea and creatinine, and ultra-small pores (0.3 to 0.5 nm), aquaporins, allowing only solute-free water to pass, acting like a sieve.

The peritoneal cavity can typically accommodate 2 to 3 L of fluid without discomfort or respiratory compromise.

Not all membranes are alike, and there is considerable amount of patient-to-patient variability in the character of the peritoneal membrane. Patients’ peritoneal membrane transport characteristics also change with time and events, usually increasing over time.

Four classes of membrane transport characteristics are defined based on the rate of creatinine diffusion and glucose absorption (Table 21-3).

The Canada-USA (CANUSA) Study Group found a greater risk of technique failure or death in patients with high transport membranes undergoing long, manual exchanges; the underlying mechanism for this increased risk is thought to involve poor ultrafiltration (with resultant hypertension and left ventricular hypertrophy) and increased protein losses.³ Most patients with high transport characteristics are placed on short, automated exchanges with no long dwells to counteract this physiology.

Continuous ambulatory peritoneal dialysis (CAPD) involves patient-operated manual exchanges performed throughout the day.

Fluid volumes of 2 to 3 L are typical, with dwell times each ranging from 6 to 8 hours.

Most patients are first educated and trained in CAPD prior to learning other modalities, as this can be used as a backup or emergency modality in the event of a power outage or machine malfunction.

Patients admitted to the hospital overnight can resort to CAPD if nurse staffing or machine availability is limited.

A sample prescription would be 2 L of 2.5% dextrose solution with four exchanges of approximately 6 hours each (or can be unevenly spaced at more convenient times of the day such as awakening, noon, dinner, and bedtime, with the longest dwell overnight).

Patients with significant residual renal function can be started on 2 or 3 exchanges instead.

In continuous cycling peritoneal dialysis (CCPD), also known as automated PD, the patient undergoes automated exchanges overnight, with three or more relatively short cycles.

The final exchange remains in the peritoneal cavity on awakening, and the patient disconnects from the machine and is free to go about doing daily activities.

The “continuous” label in the name of this modality refers to the retained daytime dwell that allows for solute transfer to occur around the clock.

An extra manual exchange is sometimes added during the day if clearance or ultrafiltration targets are not reached. In general, anuric patients have day dwells.

A sample prescription would have four dwells of 2 hours each, with 2.5 L of 2.5% dextrose solution, and a final fill (daytime dwell) of 2 L of icodextrin prior to disconnecting.

In choosing a PD modality, the patient’s membrane type should be known, as determined from the results of the PET.

Those with high transport membranes dissipate their osmotic gradients more rapidly, and short repeated dwells may be required to achieve adequate ultrafiltration (bringing in fresh hypertonic solution); these patients fare better on CCPD.

Those with low transport membranes have difficulty with solute diffusion and would benefit from the long, evenly spaced dwells of CAPD.

Patients with either high-average or low-average membranes can usually achieve adequate solute removal and ultrafiltration with either modality; thus, selection can generally depend on patient preference.

The 2006 Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines, as do multiple other medical society guidelines, recommend a weekly clearance of urea (Kt/V_urea) of at least 1.7, reflecting combined contribution from PD and residual renal function in patients producing >100 mL of urine per day.⁴ The Centers for Medicare and Medicaid Services (CMS) also deduct payment to dialysis centers if a percentage of PD patients do not reach Kt/V_urea of 1.7. However, no rigorous trial has established an “adequate” target for PD clearance and clinical parameters are paramount.

Clearance adequacy should be measured within the first month after initiating therapy, then at 4-month intervals, unless there has been a change in the prescription or in the clinical status of the patient.

The clearance is calculated from the 24-hour Kt/V_urea shown below, where V_D is the total volume of dialysate used (in L), D_urea is the dialysate urea concentration, P_urea is the plasma urea concentration, and V_urea is the estimated volume (in L) of distribution of urea (total body water or from Watson or Hume formulae).