Page 113 - Risk quantification and modification in older patients with colorectal cancer

GerCRC risk prediction model for severe complications
Background
Older patients make up the majority of newly diagnosed patients with colorectal cancer (CRC)1 and for this heterogeneous population, risks and benefits of treatment must be weighted at an individual level.2-5 Prediction models can be used to facilitate decision-making and estimate outcomes of treatment such as surgery-related morbidity and mortality. Especially severe complications are of interest, because they hinder the postoperative course and impact postoperative functioning and quality of life of older patients.6-8
For older patients with CRC potential predictors for these outcomes include physical performance measures,9-11 falls and cognitive impairments.12,13 However, in currently available prediction models, there is a focus on cancer- and surgery- related predictors. At the same time, the inclusion of perioperative predictors in many models limits their use for preoperative decision making.14-16
We have previously shown that most available CRC prediction models have a moderate to high risk of bias, especially in older adults.17 That also applies to the three surgical risk prediction models for prediction of severe complications; The Physiology and Operative Severity Score for the enumeration of Mortality and Morbidity (POSSUM),14 Colorectal Biochemical and Hematological Outcome Model (CR-BHOM)18 and The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP).10,19 Predictors related to geriatric characteristics might improve a prediction model’s performance for older CRC patients.20,21
This study aimed to develop and internally validate a prognostic preoperative clinical model for severe postoperative complications after elective surgery for stage I-III CRC, intended to support shared decision making with older patients. We hereto analyzed data from a large population-based cohort of patients ≥70 years.
111
6