Self-Declared Roma Ethnicity and Health Insurance Expenditures: A Nationwide Cross-Sectional Investigation at the General Medical Practice Level in Hungary

The inevitable rising costs of health care and the accompanying risk of increasing inequalities raise concerns. In order to make tailored policies and interventions that can reduce this risk, it is necessary to investigate whether vulnerable groups (such as Roma, the largest ethnic minority in Europe) are being left out of access to medical advances. Objectives: The study aimed to describe the association between general medical practice (GMP) level of average per capita expenditure of the National Health Insurance Fund (NHIF), and the proportion of Roma people receiving GMP in Hungary, controlled for other socioeconomic and structural factors. Methods: A cross-sectional study that included all GMPs providing care for adults in Hungary (N = 4818) was conducted for the period 2012–2016. GMP specific data on health expenditures and structural indicators (GMP list size, providing care for adults only or children also, type and geographical location of settlement, age of GP, vacancy) for secondary analysis were obtained from the NHIF. Data for the socioeconomic variables were from the last census. Age and sex standardized specific socioeconomic status indicators (standardized relative education, srEDU; standardized relative employment, srEMP; relative housing density, rHD; relative Roma proportion based on self-reported data, rRP) and average per capita health expenditure (standardized relative health expenditure, srEXP) were computed. Multivariate linear regression model was applied to evaluate the relationship of socioeconomic and structural indicators with srEXP. Results: The srEDU had significant positive (b = 0.199, 95% CI: 0.128; 0.271) and the srEMP had significant negative (b = −0.282, 95% CI: −0.359; −0.204) effect on srEXP. GP age > 65 (b = −0.026, 95% CI: −0.036; −0.016), list size <800 (b = −0.043, 95% CI: −0.066; −0.020) and 800–1200 (b = −0.018, 95% CI: −0.031; −0.004]), had significant negative association with srEXP, and GMP providing adults only (b = 0.016, 95% CI: 0.001;0.032) had a positive effect. There was also significant expenditure variability across counties. However, rRP proved not to be a significant influencing factor (b = 0.002, 95% CI: −0.001; 0.005). Conclusion: As was expected, lower education, employment, and small practice size were associated with lower NHIF expenditures in Hungary, while the share of self-reported Roma did not significantly affect health expenditures according to our GMP level study. These findings do not suggest the necessity for Roma specific indicators elaborating health policy to control for the risk of widening inequalities imposed by rising health expenses.