Back to 2025 Abstracts
DETERMINANTS OF SURGERY REFUSAL IN STAGE I-III GASTRIC ADENOCARCINOMA AND THEIR EFFECT ON SURVIVAL
Christina H. Bae
*, Ryan Quisling, Malek Moumne, Jonathan Palmer, Luke Guy, Gustavo Capo, Daniel Milgrom, Steven Colquhoun, Alicia Arnold, Danny Yakoub
Surgical Oncology, Augusta University, Augusta, GA
Introduction: Surgical resection in gastric adenocarcinoma (GA) has been associated with improved survival; however, some patients refuse surgery against physician recommendations. We examined the factors associated with surgery refusal and its effect on overall survival (OS).
Methods: The National Cancer Database (NCDB, 2004-2019) was queried for patients with stage I-III GA who were recommended surgery. Patients who underwent surgery were compared to those who refused it. Logistic regression was used to identify predictors of surgery refusal. Linear regression analysis was used to investigate trends in rates of surgical refusal between 2004 and 2019. Kaplan-Meier and Log-Rank analyses were also used to determine the effect of surgery refusal on OS. Lastly, propensity score match (PSM) was used to produce comparable groups regarding demographics and disease characteristics.
Results: Out of the 59,662 patients with stage I-III GA who were offered surgery, 1382 patients (2.32%) refused it. Multivariate logistic regression showed that female sex (OR=1.59), non-Hispanic Black race (OR=1.59), age 50+ (OR=1.53, p=0.02), refusal of chemotherapy (OR=11.02) or radiation (OR=1.48), and acceptance of palliative care (OR=5.31) were all associated with an increased probability of surgery refusal. Factors associated with a decreased probability of surgery refusal included having private insurance (OR=0.57), TNM Stage 2 disease (OR=0.86), TNM Stage 3 disease (OR=0.61), and residing 50-200 miles from the treatment center (OR=0.63). All variables met statistical significance (p<0.01). Linear regression analysis showed an increasing trend of surgery refusal from 2004-2020 (R
2=0.87, p<0.01), from 1% in 2004 to 3% in 2019. Kaplan-Meier and log-rank analysis showed a decrease in OS in those who refused surgery across all stages (Figure-1): Stage-1 (16.11 vs. 99.41 mos. HR=24.97, p<0.01), Stage-2 (20.94 vs. 51.64 mos. HR=2.82, p<0.01), and Stage-3 (19.33 vs. 25.31 mos. HR=1.39, p<0.01). Kaplan-Meier and log-rank analyses utilizing PSM groups showed a decrease in median survival in those who refused surgery (15.61 vs. 76.75 mos., p<0.01). On multivariate analysis of survival in patients who refused surgery, advanced age (HR=1.02, p=0.02), higher Charlson Comorbidity Index (CCI) (HR=1.57, p=0.006), and opting for palliative care (HR=1.75, p=0.035) were all associated with decreased OS.
Conclusion: Surgery is associated with improved survival in stage I-III GA. Targeted counseling and interventions can help mitigate the identified factors associated with surgery refusal to improve outcomes.
Kaplan-Meier curves comparing survival between surgery vs. non-surgery Stage I-III GA
Back to 2025 Abstracts
