Geriatric Risk Prediction Models for Major Gastroenterological Surgery Using the National Clinical Database in Japan

Objectives: To investigate the effect of geriatric variables on 5 newly added outcomes and create risk models for predicting these outcomes. Summary of Background Data: Because there is a current lack of geriatric research focusing on geriatric outcomes using a national surgical database in Japan, there is a need to investigate outcomes associated with major gastro-enterological surgery using these data. Methods: This multicenter prospective cohort study was conducted at 26 surgery departments across 21 institutions in Japan using the NCD surgical registry. in total, 22 new geriatric variables were imported from the ACS National Surgical Quality Improvement Program geriatric pilot study. The following 5 geriatric outcomes were defined: (1) postoperative delirium, (2) physical function on postoperative day 30, (3) fall risk on discharge, (4) discharge other than home with social service, and (5) functional decline on discharge, and geriatric risk prediction models for major gastroenterological surgery were created. Results: Between January 2018 and December 2018, data on 3981 procedures from 7 major gastroenterological surgeries were collected and analyzed. Older age and preoperative geriatric variables (Origin status from home, History of dementia, Use of mobility aid, fall history, and not competent on admission) were strongly associated with postoperative outcomes. Geriatric risk prediction models for these outcomes were created, with C-statistic values ranging from 0.74 to 0.90, demonstrating model validity and sufficiency of fit. Conclusions: The risk models for the newly defined 5 geriatric outcomes that we created can be used in the decision-making process or provision of care in geriatric patients.