A preoperative prediction model designed to forecast conversion during minimally invasive colorectal cancer surgery did not maintain adequate discrimination or calibration during internal or external validation, according to an observational study published in the International Journal of Colorectal Disease. Researchers evaluated national register data from Sweden and Denmark to explore preoperative prediction of conversion in minimally invasive colorectal cancer surgery.
Conversion from minimally invasive surgery to open procedures presents important challenges, as it often adds operative time and may lead to post-surgical complications. The primary aim of the investigation was to develop an effective preoperative prediction model for conversion and identify key predictive risk factors during minimally invasive colorectal cancer resections, intended to assist surgical teams during clinical consideration and preoperative patient counselling.
The observational investigation drew patient records from Swedish national registries within the Colorectal Cancer Database (CRCBaSE) covering the period from 2019 to 2021. Within this registry cohort, minimally invasive surgery was performed on a total of 7,811 patients. To establish the prediction framework, researchers constructed multivariable logistic regression models to predict conversion, determining final variable selection based on statistical performance criteria.
The methodological assessment incorporated the Brier score, the area under the receiver operating characteristic curve (AUC), calibration plots, and bootstrap resampling methods. For independent external validation, the research team subsequently tested the modeling approach using records from the Danish Colorectal Cancer Registry from 2016 through 2020.
Statistical evaluation identified seven preoperative predictors as risk factors for surgical conversion: body mass index (BMI), clinical T stage, age, previous abdominal surgery, sex, robotic approach, and left-sided tumour location. Conversely, female sex and the use of a robotic surgical approach reduced the risk of conversion.
A multivariable model incorporating all seven variables was developed and achieved an AUC of 0.730 in the initial Swedish test set. However, upon internal validation, the 7-variable model demonstrated reduced discrimination and calibration, registering an AUC of 0.681. External validation within the Danish national cohort confirmed consistent directions for all seven predictors but showed further reduced predictive performance, with the discrimination metric falling to an AUC of 0.608.
In their conclusions, the authors emphasized that internal validation did not maintain adequate discrimination or calibration, and external validation similarly failed to maintain acceptable predictive capacity. Nonetheless, several individual variables exhibited consistent associations with conversion across both national cohorts and may be regarded as candidate predictors.
Subject of Research: Preoperative prediction of conversion in minimally invasive colorectal cancer surgery
Article Title: Exploring preoperative prediction of conversion in minimally invasive colorectal cancer surgery using Swedish and Danish national register data
Article References: Afshari, K., Grönkvist, R., Pettersson, J., Larsson, C., Ehrencrona, C., Rosenberg, J., Angenete, E., & Haglind, E. (2026). Exploring preoperative prediction of conversion in minimally invasive colorectal cancer surgery using Swedish and Danish national register data. International Journal of Colorectal Disease, 41(1), Article 187. https://doi.org/10.1007/s00384-026-05273-w
Image Credits: AI Generated
DOI: 10.1007/s00384-026-05273-w
Keywords: colorectal cancer surgery, minimally invasive surgery, surgical conversion, preoperative prediction, logistic regression, internal validation, external validation, model calibration, model discrimination, Brier score, receiver operating characteristic, robotic surgery, body mass index, clinical T stage, CRCBaSE registry, Danish Colorectal Cancer Registry, bootstrap resampling
News Source: Ophelia Keating. (November 19, 2026). Colorectal Cancer Surgery Conversion Model Shows Reduced Discrimination in External Validation. Scienmag.



