Homelessness and Cancer: Prevalence, Mortality, and Healthcare Costs in Hungary

Abstract Background Cancer is among the leading causes of death for people experiencing homelessness (PEH) but population-level health and care impacts are not well researched. This study is an exploratory analysis of differences in cancer cases between PEH and non-homeless individuals in Hungary, using a nationwide dataset. Methods The study included all PEH registered with the insurance between 2015-2021 (n = 11,857) and 59,285 matched non-homeless controls. We analyzed all-cancer prevalence, yearly cancer-related costs in purchase power parity euros (PPP€), and survival rates. We used Kaplan-Meier estimates and Cox proportional hazards (CPH) models, adjusting for costs. Network analysis with the Kamada-Kawai algorithm was used to examine the distribution and co-occurrence of different cancer types. Results PEH had a higher all-cancer prevalence (6.4% vs. 4.7%). Five-year survival rates were significantly lower for PEH (0.26 vs. 0.56, p < 0.01) and mortality higher (HR 2.67, p < 0.01). Female sex was associated with lower mortality (HR 0.63, p < 0.01), this effect persists when looking at the PEH group separately and when excluding sex specific cancers. Yearly cancer costs were 33% lower for PEH (PPP€ 5180 vs. PPP€ 7795), but only had a minimal effect on mortality (HR 0.99 per €1000, p < 0.01). Metastatic and lung cancers were the most common in both groups. Bone cancer was prominent for PEH, while skin cancer for controls. Conclusions The difference in outcomes between PEH and controls persists even after adjusting for age, sex and cost differences. This shows that homelessness is a key determinant of cancer outcomes. Addressing higher cancer mortality in PEH requires decreasing healthcare utilization barriers and targeted prevention strategies to increase health equality. Key messages • Cancer types, survival rates, and healthcare utilization differ significantly between PEH and controls. • Interestingly, sex and age effects are very similar in a matched control sample.