The impact of socioeconomic status on stage at presentation, receipt of diagnostic imaging, receipt of treatment and overall survival in colorectal cancer patients.

ABSTRACT

Socioeconomic factors influence patterns of care in colorectal cancer. Our study investigates the impact of socioeconomic status (SES) on stage at presentation, receipt of diagnostic imaging, receipt of treatment and overall survival (OS) in a universal healthcare system. The Ontario Cancer Registry (OCR) was accessed to identify a cohort of patients diagnosed with colorectal adenocarcinoma from 2007 to 2016 in Ontario, Canada. SES was measured using median neighborhood income divided into quintiles (Q1-Q5; Q1 = lowest income). Logistic regression analyses were used to evaluate stage, imaging and treatment. Cox proportional hazards models were used to evaluate OS. All endpoints were adjusted for demographics and comorbidities with OS models also adjusting for stage, imaging and treatment. In total, 39 802 colon and 13 164 rectal patients were identified. Lower SES was associated with advanced stage at presentation in both cohorts (Q1 vs Q5 Colon odds ratio [OR] = 1.08, P = .046, rectal OR = 1.25, P < .0001). Lower SES colon patients were less likely to receive adjuvant oxaliplatin (Q1 vs Q5 OR = 0.78, P < .001) and all palliative chemotherapies studied including oxaliplatin (Q1 vs Q5 OR = 0.60, P < 0.0001), irinotecan (Q1 vs Q5 OR = 0.65, P < .0001), bevacizumab (Q1 vs Q5 OR = 0.70, P < .001), cetuximab (Q1 vs Q5 OR = 0.40, P = .0053) and panitumumab (Q1 vs Q5 OR = 0.54, P = .0036). In rectal patients, lower SES was associated with decreased receipt of rectal cancer resection for stages I-III (Q1 vs Q5 OR = 0.78, P < .001), adjuvant oxaliplatin (Q1 vs Q5 OR = 0.72, P = .0020) and palliative chemotherapies including oxaliplatin (Q1 vs Q5 OR = 0.59, P < .001), irinotecan (Q1 vs Q5 OR = 0.53, P < .001) and bevacizumab (Q1 vs Q5 OR = 0.71, P = .046). All survival models identified poorer OS for lower SES patients (total colorectal; Q1 vs Q5 Hazard ratio [HR] = 1.25, P < .0001). These findings suggest disparities persist even within universal healthcare.