The association between heart disease mortality and geographic access to hospitals: County level comparisons in Ohio, USA

Abstract

Greater distance to health care facilities is associated with poorer health care service utilization, yet little is known about how the ‘decay effect’ of distance influences the outcome of heart disease that requires frequent medical care. Heart disease has been a leading cause of death in the United States for a last few decades, even with significant improvements in treatment and management. In this study, we examined the association between physical distance to hospitals and heart disease mortality. The geographic information system (GIS) approach was taken to integrate, visualize and analyze data from multiple sources. Hospitals in the state of Ohio were geocoded and zonal statistics were computed to quantify geographical access to hospitals at the level of Ohio's 88 counties. Whereas the results of bivariate analysis showed a significant association between distance to hospitals and heart disease mortality, this relationship was not significant when accounting for socioeconomic and socio-demographic factors. This study demonstrates the usefulness of visualized health data and makes a case for further research on associations between disease outcomes and access to health care services.

Introduction

Heart disease remains a leading cause of death in the United States, even though there have been significant improvements in heart disease mortality rates. “Death rates from heart disease have fallen dramatically in the last 50 years, from over 589 age-adjusted deaths per 100,000 in 1950 to less than half that number in 2000 (258 per 100,000) (U.S. Bureau of the Census, 2001a: p. 1)”. Even with these improvements in heart disease mortality rates, heart disease is still the leading cause of death in the U.S. In 2005, 26.6% of deaths were accounted for by heart disease (Kung, Hoyert, Xu, & Murphy, 2008). In addition to its rank as a leading cause of death, heart disease carries significant human and economic costs. Currently, nearly 1 in 3 Americans live with one or more type of cardiovascular disease (Centers for Disease Control and Prevention, 2008). It is estimated that health care expenditures and lost productivity linked to disability or death due to heart disease or stroke will total $448 billion in 2008 (Centers for Disease Control and Prevention, 2008). Heart disease will remain a significant health issue in the U.S. for the foreseeable future; in fact, some researchers and public health advocates suggest that the recent trends in declining heart disease mortality might be reversed as a result of physical inactivity, and alarming increases in obesity (American Heart Association, 2007). Public health measures to prevent heart disease are essential. To that same end, understanding the paths through which heart disease results in mortality can inform public policy and practice.

As with many diseases, outcomes of heart disease partly depend on prompt action in primary treatment, as well as prevention. The decline in heart disease mortality mentioned above can, in part, be attributed to better emergency care (U.S. Census Bureau, 2002). In 2008, an estimated 920,000 people had a heart attack in the U.S.(Centers for Disease Control, 2008). Early access to emergency medical treatment is the first step in survival (Cummins, Ornato, Thies, & Pepe, 1991). According to the American Heart Association (2007), every second is critical for heart conditions; delayed treatment is related to residual disabilities and to the likelihood of death.

Citing research on the high number of cardiac deaths that occur within the first hour of symptom onset, Fang et al. (2008) emphasize the importance of early recognition of symptoms, and immediately seeking emergency medical assistance. We suggest that an additional factor—distance from a hospital—may play a crucial role in preventing death from a cardiac event, and thus may be related to overall heart disease mortality rates.

A number of past studies have identified a “decay effect” of distance on health care service utilization (Arcury et al., 2005, Higgs, 1999, Hyndman and Holman, 2001, Love and Lindquist, 1995, Luo and Wang, 2003). For instance, Nemet and Bailey (2000) found that increased distance to health care facilities was associated with reduced use of health care service among rural population in the United Kingdom. Greater distance to hospitals adds more cost, time and effort to the decisions and actions an individual makes in seeking health care (Cromley & McLafferty, 2002). Considering the time that is taken to recognize symptoms and to call paramedics, and the time for ambulance response and transfer of patients to hospitals, distance is logically an important factor in speed of treatment and, ultimately, outcome of the event.

Another important aspect of accessibility to health care is accounting for social factors that can be related to geographical factors. For example, Luo and Wang (2003) reported a tendency for areas with high proportions of vehicle owners and a socioeconomically advantaged population to have better access to health care. Also, Cromley and McLafferty (2002) indicated that lack of insurance and low income, along with greater distance, are associated with unwillingness and inability to use health care service. Having a regular doctor and having hospitals located near one's activity area are other significant predictors of health care service utilization. Thus, accessibility to health care is related to both physical distance and social “distance” factors.

The inclusion of socioeconomic and demographic variables in an exploration of the role of distance as a measure of accessibility of health care is conceptually supported by the Andersen framework (1995), which explicitly argues that such variables are measures of equitable and inequitable access to health care. Indeed, the emerging empirical evidence on the importance of distance from a hospital for health utilization and outcomes, and the rationale for the present study, link directly to Andersen's classic behavioral model of health care. In the article in which he revisits and expands the model to embed individual behaviors and outcomes in a more explicitly structural model, Andersen acknowledges the need to include organizational and community-level enabling resources. Travel time (a function of distance and access to transportation) is one of those factors. “Health personnel and facilities must be available where people live and work. Then, people must have the means… to get to those services…. Travel and waiting times are some of the measures that are important here” (p. 3) (Andersen, 1995). Explicit modeling of the impact of distance is well-supported by this conceptual framework, as is the inclusion of socioeconomic and demographic variables which Andersen (1995) explicitly mentions as measures of equitable and inequitable access to health care.

Based on the Andersen model, and seeking to add to our understanding of health outcomes, this research addresses two specific questions:

Is heart disease mortality (HDM) related to distance from a hospital?

Are HDM, and the relationship between HDM and distance, affected by social factors such as poverty, education, and insurance?

In tackling these questions, this research builds from the Andersen model by measuring an under-examined enabling resource (spatial proximity to services), extends the growing body of literature examining the role of distance in health care utilization by focusing on health outcomes, and will add to the literature derived from Andersen's (1995) revised model of structural and behavioral influences on health behaviors and health outcomes. Finally, this research contributes to the growing use of geographic information systems (GIS) as an analytical approach. (See, for example, Arcury et al., 2005, Hyndman and Holman, 2001, Love and Lindquist, 1995, Luo and Wang, 2003; and, Nemet & Bailey, 2000).