Duodenal atresia: a comparison of techniques of repair.

Although duodenal atresia is a common cause of congenital intestinal obstruction, the optimal technique of repair remains controversial. In a 10-year period, 41 newborns (20 male, 25 premature) underwent either side-to-side duodenoduodenostomy (SDD, 10 infants), side-to-side duodenojejunostomy (SDJ, 9 infants), or diamond-shaped duodenoduodenostomy (DDD, 22 infants) in a nonrandomized series. One-layer anastomosis was used in each case, and gastrostomy was placed in most patients. Groups were compared for survival, prematurity, associated anomalies, time until feeding onset, total hospitalization time, and complications. The three groups were identical for survival (all 100%), prematurity (60% SDD, 67% SDJ, 59% DDD), and serious associated anomalies (cardiac, gastrointestinal, Down's; 33% SDD, 44% SDJ, 50% DDD). The time until feeding onset was shortest with DDD (4.1 +/- 0.4 days), compared with 8.0 +/- 1.1 days for SDD and 9.6 +/- 1.9 days for SDJ (both P less than .05 v DDD). Total hospitalization was significantly less in DDD (16.2 +/- 2.1 days) v 24.2 +/- 3.1 days for SDD (P less than .05) and 28.3 +/- 4.3 days for SDJ (P less than .01). One complication necessitating reoperation occurred in each group (SDD, adhesions; SDJ, stenotic anastomosis; and DDD, missed second atresia). These data suggest that DDD is superior to SDD and SDJ for repair of duodenal atresia, resulting in earlier feeding and discharge.