Neighborhoods matter: a population-based study of provision of cardiopulmonary resuscitation

Abstract

STUDY OBJECTIVE: Cardiorespiratory resuscitation (CPR) nonprovision-the failure of bystanders to provide CPR to cardiac arrest victims-remains a well-documented public health problem associated with significant mortality. Multivariate data on failure to provide CPR are limited. Given the established independent contributions of neighborhoods to explaining many behaviors, we asked the following questions: Do neighborhood characteristics affect the likelihood of CPR nonprovision? In particular, we sought to identify the characteristics of areas that have had the most success in providing CPR. METHODS: We performed multivariable logistic regression analysis of a prospectively collected cohort of 4,379 cardiac arrests linked at an individual level to neighborhood data from the US Census. These arrests represent all out-of-hospital cardiac arrests in the City of Chicago in 1987 and 1988. RESULTS: In multivariate analysis, patients who had cardiac arrests who lived in neighborhoods where cardiac arrests were more common were significantly more likely to receive CPR. Patients with arrests in racially integrated neighborhoods were most likely to be provided with CPR, followed by those in predominately white neighborhoods, with the lowest rates of CPR provision in predominately black neighborhoods. Neither the socioeconomic status, number of elderly, nor the occupational characteristics of the neighborhood appeared to influence CPR provision. At the individual level, in-home arrests and arrests among middle-aged black residents (relative to older black and all white residents) were less likely to receive CPR. CONCLUSION: Substantial variation in rates of CPR nonprovision exists between neighborhoods; the variation is associated with neighborhood characteristics. Combining individual and neighborhood data allows identification of important factors associated with the failure to provide CPR.