Background: We hypothesized that major co-morbidities affect survival and complications after gastric bypass.

Methods: A total of 1465 patients undergoing laparoscopic and open gastric bypass between 1995 and 2002 were studied. Patients with a body mass index >or= 35 kg/m(2) and major co-morbidities (group 1, n = 1045) were compared with patients with a body mass index >or= 40 kg/m(2) with minor/no co-morbidities (group 2, n = 420).

Results: Group 1 patients were older (43 versus 36 years, P < 0.001) and had a greater BMI (53 versus 50 kg/m(2), P < 0.001). Early postoperative complications were greater in group 1 than in group 2 and included leaks (4.1% versus 1.2%, P < 0.0032) and wound infections (3.9% versus 1.4%, P < 0.0133). Procedure-related mortality in the series was 1.7%. Mortality was 10-fold greater in group 1 (2.3% versus 0.2%, P < 0.0032). The incidence of small bowel obstruction, incisional hernia, and pulmonary embolism was similar in the two groups. Excess weight loss was significantly greater in group 2 (68% versus 62%, P < 0.001) at 1 year. Resolution of group 1 co-morbidities was great, including hypertension in 62%, diabetes in 75%, venous stasis disease in 96%, and pseudotumor cerebri in 98%.

Conclusion: Outcomes analysis of obesity surgery requires risk stratification. The very low mortality rates in published studies are likely explained by surgical treatment of low-risk patients with minor co-morbidities, such as those seen in group 2. However, despite the increased perioperative risk, the group 1 patients (with major co-morbidities) demonstrated dramatic resolution of their co-morbid conditions, justifying the decision to go forward with surgery. The data support a radical change in treatment philosophy in which morbidly obese individuals should be offered bariatric surgery before major co-morbid conditions develop as a strategy to decrease the operative risk.