Mortality among homeless shelter residents in New York City.

OBJECTIVES: This study examined the rates and predictors of mortality among sheltered homeless men and women in New York City. METHODS: Identifying data on a representative sample of shelter residents surveyed in 1987 were matched against national mortality records for 1987 through 1994. Standardized mortality ratios were computed to compare death rates among homeless people with those of the general US and New York City populations. Logistic regression analysis was used to examine predictors of mortality within the homeless sample. RESULTS: Age-adjusted death rates of homeless men and women were 4 times those of the general US population and 2 to 3 times those of the general population of New York City. Among homeless men, prior use of injectable drugs, incarceration, and chronic homelessness increased the likelihood of death. CONCLUSIONS: For homeless shelter users, chronic homelessness itself compounds the high risk of death associated with disease/disability and intravenous drug use. Interventions must address not only the health conditions of the homeless but also the societal conditions that perpetuate homelessness.