[Perioperative artificial nutrition in elective surgery. A descriptive study of practice patterns in France]

To analyse the practice patterns of perioperative artificial nutrition in France before the Consensus Conference held on 16 december 1994 and to compare them with the recommendations produced by this Conference.

Retrospective cross-sectional practice patterns study.

Sample of patients collected from general or digestive tract surgeons, including for each one the last two patients having had before 10 December 1994 either elective upper (oesophagectomy, duodenopancreatectomy) or elective lower (colectomy) abdominal surgery. The response rate to the 2, 150 circulated questionnaires was 14% (309), producing a sample of 601 patients (49% with upper abdominal surgery, 51% with lower abdominal surgery).

The geographical distribution of surgeons who replied and their type of activity (private or public) was similar to the national pattern, however with some underrepresentation of the larger Paris area. The duration of pre- or postoperative artificial nutrition was bellow seven days in 36% of the cases, and above this delay in 64% of the cases. Artificial feeding rates for upper abdominal surgery, were 9 and 75%, respectively (duration of artificial feeding of 7 days and more being only considered). For lower abdominal surgery, these rates were 5 and 41%, respectively. Enteral nutrition was given in 13% of preoperative cases and 19% of postoperative ones; the others received parenteral nutrition. When compared to the recommendations by the Consensus Conference, these results show an insufficient use of preoperative artificial nutrition in patients with malnutrition (only 22% of them received it) and an excessive use in postoperative patients, particularly after lower abdominal surgery. Furthermore, the caloric intake was in accordance with the recommendations in only 20% of the cases. It was too high in 38% of the cases and insufficient in 47% of them. Similarly, the postoperative nitrogen intake complied with the recommended figures in only 5% of the cases.

This study highlights large discrepancies between the recommendations by the Consensus Conference and current practice patterns. It questions the efficiency of this therapy as it is implemented today and generates high and unwarranted expenses.