Survival of patients with colorectal peritoneal metastases is affected by treatment disparities among hospitals of diagnosis: A nationwide population-based study.

ABSTRACT

BACKGROUND: In the Netherlands, surgery for peritoneal metastases of colorectal cancer (PMCRC) is centralised, whereas PMCRC is diagnosed in all hospitals. This study assessed whether hospital of diagnosis affects treatment selection and overall survival (OS). METHODS: Between 2005 and 2015, all patients with synchronous PMCRC without systemic metastases were selected from the Netherlands Cancer Registry. Treatment was classified as cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (CRS/HIPEC), systemic therapy or other/no treatment. Hospitals of diagnosis were classified as (1) non-teaching or academic/teaching hospital and (2) HIPEC centre or referring hospital. Referring hospitals were further classified based on the frequency of CRS/HIPEC as high-, medium- or low-frequency hospital. Multivariable regression analyses were used to assess the independent influence of hospital categories on the likelihood of CRS/HIPEC and OS. RESULTS: A total of 2661 patients, diagnosed in 89 hospitals, were included. At individual hospital level, CRS/HIPEC and systemic therapy ranged from 0% to 50% and 6% to 67%, respectively. Hospital of diagnosis influenced the likelihood of CRS/HIPEC 33% versus 13% for HIPEC centres versus referring hospitals (odds ratio (OR) 3.66 [2.40-5.58]) and 11% versus 17% for non-teaching hospitals versus academic/teaching hospitals (OR 0.60 [0.47-0.77]). Hospital of diagnosis affected median OS 14.1 versus 9.6 months for HIPEC centres versus referring hospitals (hazard ratio (HR) 0.82 [0.67-0.99]) and 8.7 versus 11.5 months for non-teaching hospitals versus academic/teaching hospitals (HR 1.15 [1.06-1.26]). Compared with diagnosis in medium-frequency referring hospitals, median OS was increased in high-frequency referring hospitals (12.6 months, HR 0.82 [0.73-0.91]) and reduced in low-frequency referring hospitals (8.1 months, HR 1.12 [1.01-1.24]). CONCLUSION: Treatment disparities among hospitals of diagnosis and their impact on survival indicate suboptimal treatment selection for PMCRC.