Hospital Service Volume as an Indicator of Treatment Patterns for Colorectal Cancer

Introduction: A hospital's approach (volume of cancer treatment services provided) to treating metastatic colorectal cancer influences a patient's treatment as strongly as patient disease status. The implications of hospital-level treatment approaches across disease stages remain understudied. We sought to determine if hospital service volume (SV) for metastatic colorectal cancer could be predictive of nonstandard treatment patterns in stages I-III colon cancer. Materials and methods: Using the National Cancer Database, we examined rates of nonstandard treatment patterns among patients with colon cancer between 2010 and 2017. After adjusting for clinicopathological characteristics using multivariable logistic regression, we evaluated the relationship between hospital-level SV for metastatic colorectal cancer and nonstandard treatment approaches for patients with stages I-III colon cancer. Results: There were significant associations between hospital-level SV for metastatic colorectal cancer and the odds of chemotherapy overtreatment among patients with stage I-III colon cancer, as well as undertreatment among patients with stages II-III disease after adjusting for hospital-, patient-, and tumor-level covariates. Patients at the highest-level SV hospitals for metastatic disease had 1.29 higher odds (95% CI = 1.18-1.41; P <0.0001) of receiving overtreatment compared to patients from lowest SV hospitals. The odds ratio of undertreatment in highest SV compared to lowest SV was 0.64 (95% CI 0.560.72; P < 0.0001). Conclusions: Hospital-level SV of patients with metastatic colon cancer is a significant indicator of nonstandard treatment patterns among patients with stage I-III colon cancer. Hospitals with the highest volume of cancer treatments have higher odds of providing overtreatment, while low SVs are associated with higher odds of undertreatment. (c) 2024 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.