Published online Mar 27, 2024. doi: 10.4240/wjgs.v16.i3.717
Peer-review started: September 27, 2023
First decision: December 28, 2023
Revised: January 12, 2024
Accepted: February 18, 2024
Article in press: February 18, 2024
Published online: March 27, 2024
Due to the complexity and numerous comorbidities associated with Crohn’s disease (CD), the incidence of postoperative complications is high, significantly impacting the recovery and prognosis of patients. Consequently, additional stu
To construct novel models based on machine learning (ML) to predict short-term major postoperative complications in patients with CD following IR.
A retrospective analysis was performed on clinical data derived from a patient cohort that underwent IR for CD from January 2017 to December 2022. The study participants were randomly allocated to either a training cohort or a validation cohort. The logistic regression and random forest (RF) were applied to construct models in the training cohort, with model discrimination evaluated using the area under the curves (AUC). The validation cohort assessed the performance of the constructed models.
Out of the 259 patients encompassed in the study, 5.0% encountered major postoperative complications (Clavien-Dindo ≥ III) within 30 d following IR for CD. The AUC for the logistic model was 0.916, significantly lower than the AUC of 0.965 for the RF model. The logistic model incorporated a preoperative CD activity index (CDAI) of ≥ 220, a diminished preoperative serum albumin level, conversion to laparotomy surgery, and an extended operation time. A nomogram for the logistic model was plotted. Except for the surgical approach, the other three variables ranked among the top four important variables in the novel ML model.
Both the nomogram and RF exhibited good performance in predicting short-term major postoperative complications in patients with CD, with the RF model showing more superiority. A preoperative CDAI of ≥ 220, a di
Core Tip: Given the unique characteristics of Crohn’s disease (CD), the incidence of postoperative complications is notably high. Previous studies, while identifying risk factors influencing these complications, often yield inconsistent results due to the heterogeneity of patient populations. This machine learning-based study included data from a single center over a relatively short period in China. Novel models employing logistic regression and random forest were developed to inform individualized perioperative management of patients with CD. The models, particularly the random forest, demonstrated robust performance, highlighting the significance of preoperative CD activity index, serum albumin levels, and operation time as crucial predictors.