Primary enteric drainage of the pancreas allograft revisited

Historically, primary enteric drainage (ED) of exocrine secretions in pancreas allografts was associated with a poor outcome, mostly as a result of infectious complications. On the other hand, bladder drainage (BD), which is presently used in the majority of institutions, is associated with substantial urologic morbidity. The aim of this study is to reassess the role of primary ED by reviewing our experience with ED versus BD in simultaneous pancreas-kidney transplantations.

The records of all pancreas-kidney transplantations performed between October 1990 and September 1996 were reviewed (n = 42). Enteric drainage was used in the last 16 (38%) and BD in the first 26 (62%). The BD and ED groups were comparable with respect to donor and recipient characteristics.

Length of stay for the transplantation (mean +/- standard deviation) was significantly shorter with ED than with BD (12.9 +/- 5.6 versus 20.4 +/- 9.6 days, p = 0.007). The total number of readmissions (1.7 +/- 1.5 versus 1.2 +/- 1.2 days, p = 0.2) and the length of hospital stay in the first 6 months after discharge (13.7 +/- 16.2 versus 10 +/- 11.3 days, p = 0.4) were similar between BD and ED. Complications requiring admission were distributed as follows in BD and ED recipients: recurrent/persistent urinary complications (46% versus 6%, p = 0.01), dehydration (27% versus 6%, p = 0.05), symptomatic graft pancreatitis (8% versus 6%, p = 0.9), gastrointestinal disturbance (27% versus 12%, p = 0.1), and wound infection (12% versus 19%, p = 0.5). The duration of the operative procedure was shorter in ED than in BD (4.3 +/- 0.9 versus 5.4 +/- 0.8 hours, p = 0.01). Reoperation during the initial transplantation stay was necessary in 23% of the patients having BD, compared with none having ED (p = 0.04). Similarly, fewer ED patients underwent reoperations compared with BD patients in the first 6 months after discharge (38% versus 69%, p = 0.04). Hospital charges for ED were lower than for BD for the initial admission ($73,458 +/- 17,103 versus $107,193 +/- 32,965, p = 0.001). Actuarial patient (96% versus 94%, p = 0.6), kidney (85% versus 87%, p = 0.9), and technically successful pancreas (90% versus 85%, p = 0.6) survival rates at 1 year were similar for BD and ED.

Our results indicate that, compared with BD, ED is associated with less morbidity and shorter hospitalization without compromising outcome. Primary ED is a viable alternative to BD in simultaneous pancreas-kidney transplantation. More clinical experience with careful cost-effectiveness analysis is needed to better assess the implications of primary ED.