The association between socioeconomic position and tumour size, grade, stage, and mortality in Danish sarcoma patients - A national, observational study from 2000 to 2013

Background: Survival in sarcoma patients depends on a range of prognostic factors. An association between cancer survival and socioeconomic position is known for several other cancers. We therefore examined the relations between three socioeconomic factors and risk of presenting with known tumour related prognostic factors, and the overall mortality of the different socioeconomic and prognostic factors in 1919 patients diagnosed with sarcoma in Denmark 2000-2013.Material and methods: Patients with sarcoma in extremities or trunk wall aged 30years or more at diagnosis were identified in the Danish Sarcoma Registry and linked on an individual level to Danish national registries. We obtained data on educational level, disposable income and cohabitation status. Odds ratios (ORs) were estimated for the association between the socioeconomic factors and grade, stage and tumour size. Hazard ratios (HRs) were estimated using Cox proportional hazard models.Results: In adjusted analyses, educational level, income and cohabitation status were not associated with high grade or dissiminated stage at time of diagnosis. However, living alone was statistically significantly associated with having a large soft tissue sarcoma (>= 5cm) at time of diagnosis (OR 1.51; CI1.12-2.03). The overall mortality was statistically significantly increased in the group of patients with <= 10years of education (HR 1.27; CI 1.02-1.57), in patients with the 20% lowest income (HR 1.30; CI 1.00-1.67) and nearly in patients living alone (HR 1.16; CI 0.99-1.36).Conclusion: In this nationwide, multicentre, population-based study, soft tissue sarcoma patients living alone had greater risk of having a large tumour at time of diagnosis. Soft tissue and bone sarcoma patients with a short education, low income, or living alone, had a higher mortality. This might indicate that the social differences in mortality might be related to treatment aspects and the biology of the disease rather that the diagnostic process.