Objective This study investigates the relationship between individual and neighborhood socioeconomic status (SES) and mortality among patients with schizophrenia. Methods A study population was identified from the National Health Insurance Research Database (NHIRD) prior to the end of 1999 that included 60,402 patients with schizophrenia. Each patient was tracked until death or to the end of 2009. Individual SESs were defined by enrollee category. Neighborhood SES was defined by enrollee category (as a proxy for occupation) and education, which were classified according to the conventions of Hollingshead.
Neighborhoods were also grouped into advantaged and disadvantaged areas. The Cox proportional hazards model was used to compare the death-free survival rate between the different SES groups after adjusting for possible confounding factors and risk factors.
Results During the 10-year follow-up period, the mortality rates among high, moderate, and low individual SES groups were 12.22, 14.75, and 18.48 %, respectively (P\0.001). Schizophrenia patients with low individual SESs in disadvantaged neighborhoods had a risk of death that was 18–22 % higher than that of those with high individual SES in advantaged neighborhoods. The analysis of the combined effect of individual SES and neighborhood SES revealed that the death rates were highest among those with low individual SES and low neighborhood SES (P\0.001).
Conclusions Schizophrenia patients with low individual SES in disadvantaged neighborhoods have the highest risk of mortality despite a universal health-care system. Public health strategies and welfare policies must continue to focus on this vulnerable group.
