Peritonitis

Bacterial peritonitis is the most common serious complication of CAPD and other forms of peritoneal dialysis. Clinical presentations include abdominal pain of varying severity (guarding and rebound tenderness are unusual), and a cloudy peritoneal effluent - without which the diagnosis cannot be made.

Microscopy reveals a neutrophil count of above 100 cells per mL. Nausea, vomiting, fever and paralytic ileus may be seen if peritonitis is severe. The incidence of CAPD-associated peritonitis has been much reduced (to about one episode every three patient years) by use of a Y-disconnect system in preference to previous methods.

CAPD peritonitis must be investigated with culture of peritoneal effluent. Empirical antibiotic treatment is started, with a spectrum which covers both Gram-negative and Gram-positive organisms. Antibiotics may be given by the oral, intravenous or intraperitoneal route; most centres rely on intraperitoneal antibiotics.

Staph. aureus peritonitis should lead to a search for nasal carriage of this organism and Staph. epidermidis peritonitis may indicate contamination from the patient's (or helper's) skin. Relapsing Staph. epidermidis peritonitis with an organism with the same antibiotic sensitivity pattern on each occasion may indicate that the Tenckhoff catheter has become colonized: often this is difficult to eradicate without replacement of the catheter under antibiotic cover.

Gram-negative peritonitis may complicate septicaemia from urinary or bowel infection. A mixed growth of Gram-negative and anaerobic organisms strongly suggests bowel perforation, and is an indication for laparotomy.

Fungal peritonitis often follows antibacterial treatment but may occur de novo. Clinical presentation is very variable. It is rare to be able to cure fungal peritonitis without catheter removal as well as antifungal treatment. Intraperitoneal amphotericin has been associated with the formation of peritoneal adhesions.

Adequacy of peritoneal dialysis

No consensus yet exists on how the adequacy of peritoneal dialysis should be measured and the optimum degree of removal of urea and other waste products to be obtained in unit time. Urea kinetic modelling may be employed, as with haemodialysis.

However, it is increasingly appreciated that patient prognosis improves with greater degrees of waste product removal than were hitherto considered acceptable, and that peritoneal dialysis inadequacy is common when residual renal function declines to zero. With increasing time on treatment, adequacy may become impaired owing to alterations in the efficiency of the peritoneal membrane in transporting waste products, fluid and electrolytes. Under these circumstances conversion to haemodialysis is necessary.