Comparative analysis of short-term outcomes after bariatric surgery between two disparate populations

Albert Einstein Healthcare Network, Philadelphia, Pennsylvania 19027, USA.
Surgery for Obesity and Related Diseases (Impact Factor: 4.94). 03/2008; 4(2):110-4. DOI: 10.1016/j.soard.2007.04.007
Source: PubMed

ABSTRACT Risk adjustment is a critically important aspect of outcomes research. Racial, geographic, cultural, and socioeconomic differences are nonclinical parameters that can affect clinical outcomes measurement after gastric bypass surgery.
A single surgeon's experience with 217 consecutive laparoscopic gastric bypass patients in private practice in Southern California was compared with the same surgeon's experience with 124 consecutive patients in an academic institution in Philadelphia.
Of the Southern California and Philadelphia groups, 89%, 1%, 9%, and 1% and 55%, 38%, 6%, and 0% were white, black, Hispanic, and Asian, respectively. The average number of co-morbidities was 7.8 in the Southern California group versus 14.4 in the Philadelphia group (P <.001). The 60-day readmission to the hospital rate and emergency room admission rate was 1.4% versus 10.4% and 1.4% versus 18.5%. The insurer mix of private pay, private insurer, and federally funded insurer was 20%, 80%, and 0% in the Southern California group and 0.8%, 71%, and 28% in the Philadelphia group, respectively. Multivariate logistic regression analysis found Medicaid status and practice location independently predicted for the 60-day readmission rate (odds ratio [OR] 3.7, P = .04 and OR 5.6, P = .04, respectively) and a return to the emergency room (OR 3.2, P = .03 and OR 16.3, P <.001). Race, income, and the presence of diabetes were not independent predictors. Variables with nonsignificant differences between the Southern California and Philadelphia cohorts included average age, average body mass index, and major complications (return to surgery and intensive care unit admissions).
The results of this study have shown that in comparing and predicting the outcomes after bariatric surgery, adjustment for demographic and insurance variables might be necessary to improve accuracy.