Peritoneal leucocyte response to bacterial peritonitis in patients receiving peritoneal dialysis.

We evaluated the quantitative peritoneal leucocyte response to antibiotic therapy in 25 CAPD patients with 57 episodes of bacterial peritonitis. Eighty-eight percent of the peritonitis episodes were initially treated with a first generation cephalosporin, but results of microbial sensitivity studies led to a change in the initial antibiotic regimen in 23 episodes. Overall, 47/57 (82%) episodes were cured by antibiotic therapy alone (responders), while 10/57 (18%) required removal of the peritoneal catheter as a curative procedure (nonresponder). Neither the duration of symptoms on initial presentation nor the status of being a nonresponder could be related to the baseline peritoneal leucocyte values, either the total (PLC) or polymorphonuclear counts (PMN). Since the baseline PLC and PMN showed a 500-fold variation, subsequent changes were expressed as a percent [PLC (%) and PMN-PLC (%)] of the baseline value. On day 3 of peritonitis, PLC (%) and PMN-PLC (%) were less in responders (26% and 10%) than nonresponders (251% and 254%) (p less than 0.001). Differentiation between responders and nonresponders based on PLC (%) and PMN-PLC (%) was associated with a high degree of sensitivity (90%) and specificity (90%). Similar results were obtained for day 4. These data suggest that the temporal pattern of PLC and PMN, when expressed as a percentage of the baseline value, may be useful in predicting those episodes of peritonitis which require removal of the peritoneal catheter.