The classic jejunostomy for enteral feedings has several major complications including intraperitoneal leakage and intestinal obstruction. In some children, severe adhesions from previous surgery makes the finding and isolation of the proximal jejunum both tedious and hazardous. A jejunostomy may be easily placed into the retroperitoneal duodenojejunal flexure without entering the peritoneum. This is accomplished by a flank muscle splitting incision displacing the descending colon anteriorly and the kidney posteriorly. A Broviac catheter with a Dacron cuff is used as the conduit. The catheter is secured to the skin until fibrous adhesions have anchored the Dacron cuff. Four retroperitoneal jejunostomy tubes have been placed. The first tube was inserted with an incomplete Dacron cuff and "accidently" extracted on the day after surgery. The second tube was placed for temporary nutritional support until definitive surgery was performed. The tube was in place for 3 months with the patient gaining 2.4 kg. The third and fourth tubes were placed for permanent nutritional support and have been in for 4 and 2.5 months. The patients have gained 1.3 and 3.0 kg, respectively. Both tubes have been transected through carelessness, but easily repaired using the Broviac repair kit.

A prospective study in healthy infants predefining both diet fatty acid and cholesterol, from birth to age 1 year, compared response of cholesterol fractions in three groups: ran- dom assignment to 1) monounsaturated-(Hi-Mono) (n = 20), or 2) polyunsaturated-(Hi-Poly) (n = 22) fatty acid-enriched diets, or 3) non-randomized selection to breast feeding (Human Milk) (n = 25). In each group, designated weaning foods and supplements maintained fatty acid and cholesterol intake similar to that of each group's defined formulas, with long-term compli- ance confirmed by plasma phospholipid fatty acid concentra- tions. By 12 months, total cholesterol was significantly lower in the Hi-Poly group compared to either of the other groups (P < 0.05). Low density lipoprotein (LDL)- and high density lipoprotein (HDL)-cholesterol concentrations were significantly lower by 12 months in the Hi-Poly group, compared to the Hi- Mono groups. However, at the earlier 4-month interval, total cholesterol and LDL-cholesterol in both Hi-Mono and Hi-Poly groups were not different from each other, although each was significantly lower than the parallel Human Milk-group (P < 0.05). The Hi-Mono group increased gradually in total and LDL-cholesterol such that, after 12 months' feedings, all lipid fractions of this Hi-Mono group were no different from those of the Human Milk group. In independent group compar- isons, there were no significant differences in HDL-cholesterol concentrations after 4 and 9 months on these diets. Independent of diet, HDL-cholesterol showed a falling trend as an overall time-effect across all groups (P < 0.001). These data sug- gest that prolonged feeding of a diet enriched in polyunsaturated acids in early infancy has a significant cholesterol-lowering effect compared to monounsaturates. These differences in total, LDL-, and HDL-cholesterol plasma concentrations between polyun- saturates and monounsaturates were not significantly evident until feedings had continued for a year.-Mize, C. E., R. Uauy, R. Kramer, M. Benser, S. Allen, and S. M. Grundy. Lipoprotein-cholesterol responses in healthy infants fed defined diets from ages 1 to 12 months: comparison of diets predominant in oleic acid versus linoleic acid, with parallel observations in in- fants fed a human &-based diet.J Lijndh. 1995. 36 1178-1187.