EP781 - Are High-Volume Hospitals Worth the Trip for Patients with Non-metastatic Colon Cancer?

Image or Table (Do Not add text here):

Purpose/Background: Surgery at high-volume hospitals (HVHs) is associated with improved perioperative outcomes and overall survival (OS) for colon cancer (CC) patients. Limited research assesses whether travelling to distant HVHs provides better outcomes than nearby lower-volume hospitals (LVHs), especially considering potential care fragmentation and surgical delays. This study evaluates the impact of travel distance and hospital volume on perioperative outcomes and 5-year OS for patients with non-metastatic CC.

Methods/Interventions: We queried the 2004-2020 National Cancer Database for stage I-III CC patients who underwent surgery. Hospitals were classified by average annual volume as LVH ( < 25th percentile: < 37 cases/year), intermediate- (IVH; 25-75th percentile; 37-81 cases/year), or HVH (>75th percentile: >81 cases/year). Haversine distance was used to classify hospitals as near (≤75th percentile) or distant (>75th percentile). Patients living outside large metropolitan state areas (MSAs; >1 million population) were included if they had surgery at a near LVH, near IVH, or distant HVH. Fragmented care was defined as receiving care at multiple hospitals. Multivariable regression models were computed to assess 5-year OS, delayed surgery (>30 days post-diagnosis), and 30-day readmission or mortality.

Results/Outcomes: Among 116,598 patients, 34.2% had surgery at near LVHs, 49.5% at near IVHs, and 16.3% at distant HVHs. Haversine distance to HVHs was 42.5 miles vs. 5.0 miles for LVHs/IVHs (p < 0.001). From 2004-2020, distant HVH surgeries increased from 12.4% to 20.6%, while near LVH surgeries decreased from 37.0% to 34.2%. Asian patients were least likely to have surgery at distant HVHs (5.3% vs. 15.4% for Black and 16.7% for White patients; p< 0.001). Patients living in rural areas most often received surgery at distant HVHs (60.0% vs. 39.8% and 7.4% for patients living in urban areas and small MSAs, respectively; p< 0.001).

Care fragmentation was more frequent for patients traveling to distant HVHs (15.0% vs. IVHs 9.5% and LVHs 10.2%; p< 0.001), as were delayed surgeries (26.5% vs. IVHs 21.3% and LVHs 18.1%; p< 0.001). HVHs had lower rates of 30-day readmission (6.4% vs. IVHs 8.0% and LVHs 7.7%; p< 0.001) and mortality (2.5% vs. IVHs 3.4% and LVHs 4.1%; p< 0.001). Five-year OS was higher at distant HVHs (72.3% vs. IVHs 68.6% and LVHs 66.2%; p< 0.001).

Compared to near LVHs, distant HVHs had better 5-year OS (HR 0.930 [95% CI 0.896-0.966]) (Figure), lower 30-day readmission (OR 0.812 [0.748-0.882]), and lower 30-day mortality (OR 0.726 [0.640-0.824]) on regression analysis. The survival benefit for HVHs vs. LVHs was less pronounced when patients with 30-day mortality were excluded (HR 0.957 [0.920-0.996]). Surgery at HVHs showed no OS benefit when care was fragmented (HR 1.074 [0.957-1.205]). Delayed surgery at distant HVHs was still associated with better 5-year OS (HR 0.802 [0.750-0.858]) and lower 30-day mortality (OR 0.528 [0.413-0.675]) than non-delayed surgery at near LVHs. Excluding patients with < 7 days between diagnosis and surgery (i.e., urgent or emergent), we found no difference in 5-year OS and 30-day mortality between delayed surgery at HVHs and non-delayed surgery at LVHs.

Conclusion/Discussion: Surgery at distant HVHs provides modest 5-year OS benefits over near LVHs and similar outcomes to near IVHs. Distant HVHs offer no survival advantage over near LVHs in cases of fragmented care, and delayed surgery at distant HVHs shows comparable survival to non-delayed surgery at near LVHs.