Gus J Slotman1. 1. Inspira Health Network, Vineland, New Jersey. Electronic address: slotmang@ihn.org.
Abstract
BACKGROUND: No method preoperatively predicts the postoperative bariatric surgery outcomes in individual patients. Decisions for or against surgery and operation choice remain subjective. Only 1% of qualifying patients embrace bariatric surgery. OBJECTIVE: To predict preoperatively and validate prospectively the weight and co-morbidity resolution in individual patients afteropen Roux-en-Y gastric bypass (RYGB) and laparoscopic Roux-en-Y gastric bypass (LRYGB), laparoscopic adjustable gastric band (LAGB), sleeve gastrectomy (SG), and biliopancreatic diversion/duodenal switch (BPD/DS). SETTING: Surgical Review Corporation BOLD database, 2007-2010. METHODS: A total of 166,601 patients who had undergone RYGB (n = 5389), LRYGB (n = 83,059), LAGB (n = 67,514), SG (n = 8966), or BPD/DS (n = 1673) were randomized into modeling (n = 124,053) and validation (n = 42,548) groups. From preoperative data, multivariate linear and logistic regression predicted weight and co-morbidities at 2, 6, 12, 18, and 24 months postoperatively. Model fit was examined by R2 and receiver operating characteristic/area under the curve and predicted versus observed results via Pearson correlation coefficient and sensitivity/specificity. RESULTS: Follow-up at 2/24 months was 120,909/11,014 for the modeling group and 41,528/3703 for validation. Weight models' R2 was .910, .813, .725, .638, and .613 at 2, 6, 12, 18, and 24 months, respectively. The categorical receiver operating characteristic/area under the curve was .617 to .949 for 24-month predictions. Pearson continuous coefficients were .969 and .811 at 2 and 24 months, respectively. The median 24-month sensitivity and specificity of co-morbidity resolution were 79.2% and 97.42%, respectively. CONCLUSIONS: Prospectively validated preoperative models predict, in individual patients, weight and obesity co-morbidities 2 years in advance for RYGB, LRYGB, LAGB, SG, and BPD/DS. This advance knowledge facilitates choosing the operation that is best for each individual and may encourage more patients to choose bariatric surgery.
RCT Entities:
BACKGROUND: No method preoperatively predicts the postoperative bariatric surgery outcomes in individual patients. Decisions for or against surgery and operation choice remain subjective. Only 1% of qualifying patients embrace bariatric surgery. OBJECTIVE: To predict preoperatively and validate prospectively the weight and co-morbidity resolution in individual patients after open Roux-en-Y gastric bypass (RYGB) and laparoscopic Roux-en-Y gastric bypass (LRYGB), laparoscopic adjustable gastric band (LAGB), sleeve gastrectomy (SG), and biliopancreatic diversion/duodenal switch (BPD/DS). SETTING: Surgical Review Corporation BOLD database, 2007-2010. METHODS: A total of 166,601 patients who had undergone RYGB (n = 5389), LRYGB (n = 83,059), LAGB (n = 67,514), SG (n = 8966), or BPD/DS (n = 1673) were randomized into modeling (n = 124,053) and validation (n = 42,548) groups. From preoperative data, multivariate linear and logistic regression predicted weight and co-morbidities at 2, 6, 12, 18, and 24 months postoperatively. Model fit was examined by R2 and receiver operating characteristic/area under the curve and predicted versus observed results via Pearson correlation coefficient and sensitivity/specificity. RESULTS: Follow-up at 2/24 months was 120,909/11,014 for the modeling group and 41,528/3703 for validation. Weight models' R2 was .910, .813, .725, .638, and .613 at 2, 6, 12, 18, and 24 months, respectively. The categorical receiver operating characteristic/area under the curve was .617 to .949 for 24-month predictions. Pearson continuous coefficients were .969 and .811 at 2 and 24 months, respectively. The median 24-month sensitivity and specificity of co-morbidity resolution were 79.2% and 97.42%, respectively. CONCLUSIONS: Prospectively validated preoperative models predict, in individual patients, weight and obesity co-morbidities 2 years in advance for RYGB, LRYGB, LAGB, SG, and BPD/DS. This advance knowledge facilitates choosing the operation that is best for each individual and may encourage more patients to choose bariatric surgery.