A scoring system to predict a prolonged length of stay after surgery for Crohn's disease

Aim: Many factors influence the postoperative length of stay (LOS) in Crohn's disease (CD). This study aims to identify the factors associated with a prolonged LOS after ileocolic resection (ICR) for CD and to develop a scoring system to predict the postoperative LOS in CD. Method: Patient data were collected from St Marks Hospital, London, UK, and the Humanitas Clinical and Research Center Milan, Italy, for all patients who underwent an ICR for CD from 2005 to 2017. Logistic regression was used for multivariate analysis. The scoring system was developed from the logistic regression model. The performance of the scoring system was evaluated using the area under the receiver operating characteristic curve (AUROC). Results: A total of 628 surgeries were included in the analysis. Eighty eight surgeries were excluded due to missing data. The remaining 543 were divided into two cohorts for the development (n = 418) and validation (n = 125) of the scoring system. The regression model was statistically significant (p < 0.0001). The statistically significant independent variables included the time since diagnosis, American Society of Anesthesiologists (ASA) grade, perioperative use of steroids, surgical access, strictureplasty and platelet count. The AUROCs for the development and validation cohorts were 0.732 and 0.7, respectively (p < 0.0001). The cut-off score suggested by Youden's index was 50, with a sensitivity of 65.6% and a specificity of 73.3%. Conclusion: The time since diagnosis, ASA grade, steroid use, surgical access, strictureplasty and platelet count were associated with a prolonged LOS and were used to develop a scoring system. The calculator is available online at https://rebrand.ly/Crohnscal.

Aim Many factors influence the postoperative length of stay (LOS) in Crohn's disease (CD). This study aims to identify the factors associated with a prolonged LOS after ileocolic resection (ICR) for CD and to develop a scoring system to predict the postoperative LOS in CD.Method Patient data were collected from St Marks Hospital, London, UK, and the Humanitas Clinical and Research Center Milan, Italy, for all patients who underwent an ICR for CD from 2005 to 2017. Logistic regression was used for multivariate analysis. The scoring system was developed from the logistic regression model. The performance of the scoring system was evaluated using the area under the receiver operating characteristic curve (AUROC).Results A total of 628 surgeries were included in the analysis. Eighty eight surgeries were excluded due to missing data. The remaining 543 were divided into two cohorts for the development (n = 418) and validation (n = 125) of the scoring system. The regression model was statistically significant (p < 0.0001). The statistically significant independent variables included the time since diagnosis, American Society of Anesthesiologists (ASA) grade, perioperative use of steroids, surgical access, strictureplasty and platelet count. The AUROCs for the development and validation cohorts were 0.732 and 0.7, respectively (p < 0.0001). The cut-off score suggested by Youden's index was 50, with a sensitivity of 65.6% and a specificity of 73.3%.Conclusion The time since diagnosis, ASA grade, steroid use, surgical access, strictureplasty and platelet count were associated with a prolonged LOS and were used to develop a scoring system. The calculator is available online at .