Increasing use of Medicare services by veterans with acute myocardial infarction.

OBJECTIVES: Some of the nation's 26 million veterans have two government-financed health care entitlements: Medicare and the Department of Veterans Affairs (VA). The aims of this investigation were to examine trends where Medicare-eligible VA users are initially hospitalized for acute myocardial infarction (AMI) and then to assess rates of cardiac procedure use and mortality for veterans initially admitted to each system of care. METHODS: We used VA and HCFA national databases to identify VA users (age range, > or = 65 years) who were initially admitted to a VAMC or Medicare financed hospital (Medicare hospital) with a primary diagnosis of AMI between January 1, 1992, and December 31, 1995, (n = 47,598). We examined the use of cardiac procedures (cardiac catheterization [CC], coronary artery bypass surgery [CABG], and coronary angioplasty [CA] and mortality (30-day and 1-year) by the type of initial admitting hospital within each system of care. RESULTS: Almost 70% of VA users hospitalized for AMI were initially admitted to Medicare hospitals versus VAMCs between 1992 (64%) and 1995 (72%). After adjusting for patient characteristics in logistic models, VA users initially hospitalized in Medicare hospitals were significantly more likely to undergo cardiac procedures than were VA users hospitalized in VAMCs. Differences in the odds of receiving a procedure were most significant when comparing Medicare hospitals with on-site cardiac technology to VA hospitals without on-site cardiac technology (CC: OR 4.34, 95% CI 3.98-4.73; CABG: OR 2.16, 95% CI 1.92-2.43; CA: OR 4.56, 95% CI 3.98-5.25). We found no significant differences in 30-day and 1-year adjusted mortality rates between VA users initially admitted to VAMCs or Medicare hospitals. CONCLUSIONS: Medicare-eligible VA users are increasingly hospitalized in Medicare hospitals for AMI. VA users cared for in Medicare hospitals receive more cardiac procedures but have the same survival as VA users cared for in VAMCs. These findings have policy implications for access, quality, and costs in both systems of care.

School leaving: a longitudinal perspective including neighborhood effects.

Using 1970 and 1980 census data from 202 tracts in the Chicago metropolitan region, we examine whether neighborhoods influence the likelihood of high school graduation for a cohort of African-American children followed from 1966 to 1993. Neighborhood-level variables included percent living below poverty and percent in white collar occupations. We test for the possible direct, indirect, and interactive effects of these neighborhood indicators on the likelihood of school dropout. Our examination found the advantage of living in a neighborhood characterized by a high percentage of residents who work in white-collar occupations. Male adolescents who lived in a middle-class neighborhood were more likely to graduate from high school, even with family background, early school performance, adolescent family supervision, and adolescent marijuana use controlled. These findings are consistent with findings from three other studies. However, living in a poverty census tract did not seem to influence the likelihood of high school graduation or school leaving over and above the impact of family and individual characteristics. There also were no neighborhood effects for females.