Abstract
OBJECTIVE: Concentration of care and collaborations between hospitals increasingly reorganise oncological care into Comprehensive Cancer Networks (CCNs), aiming to improve care outcomes and reduce costs. This study aims to evaluate the effect of four CCNs on healthcare cost and outcomes for patients with colon or pancreatic cancer. DESIGN: We performed a retrospective cohort study based on claims data in the Netherlands. Data included patient characteristics, health insurance claims and healthcare activities. All costs were indexed to Euro 2023. We performed propensity score matching per CCN and applied regression models with a difference-in-difference design, adjusting for non-linear trends before the start of a CCN. SETTING: The study was conducted within the Dutch healthcare system, analysing claims data representative of hospital-based cancer care. PARTICIPANTS: A total of 92 309 patients with colon cancer and 25 630 patients with pancreatic cancer were included. Patients were identified through health insurance claims between January 2013 and June 2021. INTERVENTION: Implementation of four CCNs, which included structured collaboration between healthcare organisations. Follow-up duration was 2 years post-diagnosis. OUTCOME MEASURES: Primary outcomes included 2-year oncological healthcare costs and 2-year mortality rate. Secondary outcomes involved care process indicators: referral rates and double diagnostics (an identical diagnostic activity performed within 4 weeks after referral to a secondary hospital). RESULTS: For colon cancer, one CCN showed a significant decrease in 2-year oncological costs (-€1899). One CCN showed a significant decrease in referrals (-3.6%) and one a significant increase (+4.4%). No significant effect on 2-year mortality and double diagnostic activities was found. For pancreatic cancer, one CCN showed a significant decrease in 2-year oncological costs (-€3747) and one CCN showed a significant increase in double diagnostic activities (+8.6%). No significant effect on referrals and 2-year mortality was found. CONCLUSION: CCNs do not consistently reduce costs or affect referral patterns or redundant diagnostics. No impact on mortality was found. Additional insights into determinants of CCN success are required before broad implementation is warranted.