The 14 studies included (n=6,599) comprised one randomised controlled trial (n=51) and 13 retrospective studies, of which two matched at least by age, sex and body mass index (n=568) and 11 were without matching study groups (n=5,980). Study quality was graded very low for 11 studies, low for two and moderate for one.
The median absolute difference in loss of excess body weight at one year was consistently greater for Roux-en Y gastric bypass than laparoscopic gastric banding (median difference 26%, range 19% to 34%, p<0.001; 12 studies). Resolution of comorbidities, where reported, was greater after a Roux-en-Y bypass. In the RCT, weight loss at one year was maintained through five years of follow-up.
Patients that had laparoscopic gastric banding generally had lower rates of short-term complications compared with Roux-en Y gastric bypass; operative times were shorter by a median of 68 minutes (eight studies), length of hospital stay was approximately two days shorter (11 studies) and there were fewer deaths (rates were low in both groups at 0.06% versus 0.17%) (13 studies). Rates of conversion to open procedures, perforation, bleeding and anastomotic leaks were low in both groups. Overall difference in major early complications ranged between 1.1% and 6.3% and were less common with laparoscopic gastric banding. Laparoscopic gastric banding was associated with more long-term complications than Roux-en Y gastric bypass and the complication rates differed markedly between studies.
Patient satisfaction was greater in the Roux-en Y gastric bypass group compared with the laparoscopic gastric banding group (80% versus 46% were very satisfied and 0% versus 19% were unsatisfied, p=0.006; one study).