IntroductionThe frequency of bile duct injuries associated to cholecystectomy remains constant (0.3–0.6%). A multidisciplinary approach (endoscopical, radiological, and surgical) is necessary to optimize the outcome of the patient. Surgery is indicated when complete section of the duct is identified (Strasberg’s E injuries) requiring a bilioenteric anastomosis as treatment. Nowadays, the most frequent technique used for reconstruction is a Roux-en-Y hepatojejunostomy. Long-term results of reconstruction are related to several technical and anatomic factors, but an ischemic duct (with subsequent scarring) plays a mayor role. In this paper, we report the results of biliary reconstructions comparing the extrahepatic—probably ischemic—to intrahepatic—non ischemic—repairs. MethodsWe reviewed the files of patients referred to our hospital (third-level teaching hospital) for bile duct repair after iatrogenic injury from 1990 to July 2006. Injury classification, time lapse since injury, surgical repair technique, and long-term follow-up were noted. In all cases, a Roux-en-Y hepatojejunostomy was done. Partial resection of segment IV was performed in 136 patients to obtain noninflamed, nonscarred, nonischemic biliary ducts with the purpose of reaching the confluence and achieving a high-quality bilioenteric anastomosis. An anastomosis at the level of the confluence was attempted in 293 patients (in 198 the confluence was preserved and in 95 it was lost). In the remaining 80 patients, a low bilioenteric anastomosis was done at the level of the common hepatic duct. We compared intrahepatic (198) and extrahepatic (80) repairs. ResultsA total of 405 cases (88 males, 317 females) were identified, with a mean age of 42 years (range 17–75). All of the injuries were classified as Strasberg E1, E2, E3, E5 (less frequent); those with E4 classification (separated ducts) were excluded. In all cases, the confluence was preserved (N = 293). Thirty-two cases were repaired minutes to hours after the injury occurred. The remaining 373 patients arrived weeks after the injury. In 198 cases, an intrahepatic repair was done, including the 136 in which resection of segments IV and V was part of the surgery. In the remaining 80 cases (operated between 1990 and 1997), an extrahepatic repair was done at the level of the common hepatic duct where the surgeon found a healthy duct. Twelve (15%) of the 80 cases with extrahepatic anastomosis required a new intervention (surgical or radiological), compared to only 8 of the 198 (3%) that had an intrahepatic anastomosis (P = 0.00062). Good results were obtained in 85% and 97% of the cases with extrahepatic anastomosis and intrahepatic anastomosis, respectively. Both groups had a reintervention rate of 7% (20/278). ConclusionsAn intrahepatic anastomosis requires finding nonscarred, nonischemic ducts, thus allowing a safe and high-quality anastomosis with significantly better results when compared to the low-level anastomosis group.