Preoperative assessment of severe complications before bariatric surgery by risk prediction models: Is it worth it?
EAES Academy. Karpińska I. 07/05/22; 363191; P255
Dr. Izabela Karpińska
Contributions
Contributions
Abstract
Introduction
Bariatric surgery was proven to be effective and safe obesity treatment. Sleeve gastrectomy is one of the most commonly performed procedure worldwide. Obesity-related comorbidities contribute to the occurrence of complications after intervention. Preoperative assessment of possible outcomes seems to be crucial for surgeons in qualification process and perioperative care. Over past decade various tools predicting complications after bariatric surgery has been proposed.
Aim
We aimed to identify and validate available risk stratification models as the predictors of severe complications after sleeve gastrectomy.
Material and methods
The retrospective analysis included patients who underwent sleeve gastrectomy (SG) and completed 30-day of follow-up. The literature review was done to identify available risk stratification models. The score or odds of postoperative complications were calculated for each patient. Postoperative complications were defined as any abnormality reaching 3 or more stage in Clavien-Dindo Classification and occurring within 30 days after the operation. The relationship between predicted and actual outcomes were assessed by logistic regression analysis. Discrimination was evaluated by area under the receiver operating characteristic (AUROC) whereas calibration by Hosmer–Lemeshow test.
Results:
Out of 997 patients enrolled 662 (66.40%) were women whereas 335 (33.60%) were men with mean age 41 years. The most common comorbidities were: dyslipidemia (69.71%), hypertension (63.29%) and obstructive sleep apnea (28.28%). Severe postoperative complications occurred in 1.71% of patients.
Literature review identified ten models. In logistic regression analysis only three of them had statistically significant capability of identifying severe complications (OR: 1.26-1.52). Two models including MBSAQIP Calculator and model proposed by Gupta reached reasonable discrimination with AUROC=0.73 and 0.72 respectively. Moreover both of them did not lose their goodness-of-fit in Hosmer–Lemeshow test.
Conclusions:
There are three tools which seem to be helpful in preoperative assessment of severe complications after sleeve gastrectomy. MBSAQIP Calculator and Gupta’s model seems to be more accurate than others. Further external validation of analysed models at the international level is needed.
Keywords
predictive models; external validation; severe complications; bariatric surgery; sleeve gastrectomy,
Bariatric surgery was proven to be effective and safe obesity treatment. Sleeve gastrectomy is one of the most commonly performed procedure worldwide. Obesity-related comorbidities contribute to the occurrence of complications after intervention. Preoperative assessment of possible outcomes seems to be crucial for surgeons in qualification process and perioperative care. Over past decade various tools predicting complications after bariatric surgery has been proposed.
Aim
We aimed to identify and validate available risk stratification models as the predictors of severe complications after sleeve gastrectomy.
Material and methods
The retrospective analysis included patients who underwent sleeve gastrectomy (SG) and completed 30-day of follow-up. The literature review was done to identify available risk stratification models. The score or odds of postoperative complications were calculated for each patient. Postoperative complications were defined as any abnormality reaching 3 or more stage in Clavien-Dindo Classification and occurring within 30 days after the operation. The relationship between predicted and actual outcomes were assessed by logistic regression analysis. Discrimination was evaluated by area under the receiver operating characteristic (AUROC) whereas calibration by Hosmer–Lemeshow test.
Results:
Out of 997 patients enrolled 662 (66.40%) were women whereas 335 (33.60%) were men with mean age 41 years. The most common comorbidities were: dyslipidemia (69.71%), hypertension (63.29%) and obstructive sleep apnea (28.28%). Severe postoperative complications occurred in 1.71% of patients.
Literature review identified ten models. In logistic regression analysis only three of them had statistically significant capability of identifying severe complications (OR: 1.26-1.52). Two models including MBSAQIP Calculator and model proposed by Gupta reached reasonable discrimination with AUROC=0.73 and 0.72 respectively. Moreover both of them did not lose their goodness-of-fit in Hosmer–Lemeshow test.
Conclusions:
There are three tools which seem to be helpful in preoperative assessment of severe complications after sleeve gastrectomy. MBSAQIP Calculator and Gupta’s model seems to be more accurate than others. Further external validation of analysed models at the international level is needed.
Keywords
predictive models; external validation; severe complications; bariatric surgery; sleeve gastrectomy,
Introduction
Bariatric surgery was proven to be effective and safe obesity treatment. Sleeve gastrectomy is one of the most commonly performed procedure worldwide. Obesity-related comorbidities contribute to the occurrence of complications after intervention. Preoperative assessment of possible outcomes seems to be crucial for surgeons in qualification process and perioperative care. Over past decade various tools predicting complications after bariatric surgery has been proposed.
Aim
We aimed to identify and validate available risk stratification models as the predictors of severe complications after sleeve gastrectomy.
Material and methods
The retrospective analysis included patients who underwent sleeve gastrectomy (SG) and completed 30-day of follow-up. The literature review was done to identify available risk stratification models. The score or odds of postoperative complications were calculated for each patient. Postoperative complications were defined as any abnormality reaching 3 or more stage in Clavien-Dindo Classification and occurring within 30 days after the operation. The relationship between predicted and actual outcomes were assessed by logistic regression analysis. Discrimination was evaluated by area under the receiver operating characteristic (AUROC) whereas calibration by Hosmer–Lemeshow test.
Results:
Out of 997 patients enrolled 662 (66.40%) were women whereas 335 (33.60%) were men with mean age 41 years. The most common comorbidities were: dyslipidemia (69.71%), hypertension (63.29%) and obstructive sleep apnea (28.28%). Severe postoperative complications occurred in 1.71% of patients.
Literature review identified ten models. In logistic regression analysis only three of them had statistically significant capability of identifying severe complications (OR: 1.26-1.52). Two models including MBSAQIP Calculator and model proposed by Gupta reached reasonable discrimination with AUROC=0.73 and 0.72 respectively. Moreover both of them did not lose their goodness-of-fit in Hosmer–Lemeshow test.
Conclusions:
There are three tools which seem to be helpful in preoperative assessment of severe complications after sleeve gastrectomy. MBSAQIP Calculator and Gupta’s model seems to be more accurate than others. Further external validation of analysed models at the international level is needed.
Keywords
predictive models; external validation; severe complications; bariatric surgery; sleeve gastrectomy,
Bariatric surgery was proven to be effective and safe obesity treatment. Sleeve gastrectomy is one of the most commonly performed procedure worldwide. Obesity-related comorbidities contribute to the occurrence of complications after intervention. Preoperative assessment of possible outcomes seems to be crucial for surgeons in qualification process and perioperative care. Over past decade various tools predicting complications after bariatric surgery has been proposed.
Aim
We aimed to identify and validate available risk stratification models as the predictors of severe complications after sleeve gastrectomy.
Material and methods
The retrospective analysis included patients who underwent sleeve gastrectomy (SG) and completed 30-day of follow-up. The literature review was done to identify available risk stratification models. The score or odds of postoperative complications were calculated for each patient. Postoperative complications were defined as any abnormality reaching 3 or more stage in Clavien-Dindo Classification and occurring within 30 days after the operation. The relationship between predicted and actual outcomes were assessed by logistic regression analysis. Discrimination was evaluated by area under the receiver operating characteristic (AUROC) whereas calibration by Hosmer–Lemeshow test.
Results:
Out of 997 patients enrolled 662 (66.40%) were women whereas 335 (33.60%) were men with mean age 41 years. The most common comorbidities were: dyslipidemia (69.71%), hypertension (63.29%) and obstructive sleep apnea (28.28%). Severe postoperative complications occurred in 1.71% of patients.
Literature review identified ten models. In logistic regression analysis only three of them had statistically significant capability of identifying severe complications (OR: 1.26-1.52). Two models including MBSAQIP Calculator and model proposed by Gupta reached reasonable discrimination with AUROC=0.73 and 0.72 respectively. Moreover both of them did not lose their goodness-of-fit in Hosmer–Lemeshow test.
Conclusions:
There are three tools which seem to be helpful in preoperative assessment of severe complications after sleeve gastrectomy. MBSAQIP Calculator and Gupta’s model seems to be more accurate than others. Further external validation of analysed models at the international level is needed.
Keywords
predictive models; external validation; severe complications; bariatric surgery; sleeve gastrectomy,
{{ help_message }}
{{filter}}