Travel burden is a key element in conceptualizing geographic access to health care. A better understanding of distances and mode of travel for individuals seeking health care is particularly important for vulnerable populations, such as rural residents and racial and ethnic minorities, who are more likely to experience barriers to transportation. Rural residents face travel barriers stemming from distance and the lack of public transportation systems in rural areas. Rural households are more likely than urban households to own at least one car . Rural households tend to make fewer trips per day, but travel 38% more miles . Poorer people living in rural areas travel 59% more miles per day than their urban counterparts . Rural residents unable to own or operate cars often depend on friends and family for transportation, limiting their trip timing, route, flexibility, and preferred mode of travel. This dependence has been shown to be associated with reduced numbers of physician visits for chronic care . Public transportation is limited in rural areas; even in rural households without cars, only 1% of trips are made by public transportation . Rural residents with more complex medical conditions are more likely to travel further for care than those living in urban areas, as are children and older people living in rural areas [3–7]. Compared with persons living in urban areas, rural residents reported longer travel time to see a physician, particularly specialists .
Barriers to transportation in rural areas compound access problems traditionally experienced by minorities [9, 10]. In both urban and rural areas, minorities are more likely to use public transportation for all non-work related trips, even after adjusting for socioeconomic characteristics . African-Americans report longer travel distances for non-work related trips than whites; Hispanics report that non-work related trips are longer in duration than those made by other racial and ethnic groups . Utilization of health care tends to decrease as the distance traveled to care increases. Uninsured Americans living closer to safety-net providers, for example, report fewer unmet health needs and are more likely to have a usual source of care than those who live further away . Transportation barriers to care are also associated with reduced compliance to treatment regimens and lower rates of preventive care, as well as greater difficulties in accessing emergency health care [13, 14].
Most previous studies of travel for care have been limited to specific geographic regions or specific populations such as Medicare beneficiaries [3, 15], use of mammogram services , rural residents with a diagnosis of human immunodeficiency virus , follow up care after a myocardial infarction among patients insured through the Veteran's Administration , failure to keep physician appointments [18, 19] and use of pharmacy services . To the authors' knowledge, no previous studies have examined travel for medical care using a nationally representative population, and examining actual distance information. The research reported here sought to address this gap by using a transportation planning resource, the National Household Travel Survey, to provide a detailed description of travel to care patterns by residence and race and ethnicity. The purpose of this study is to provide nationally representative estimates of the distance traveled along roads and time spent in travel for medical or dental care, comparing differences among rural and urban residents and by race and ethnicity.
Transportation is linked to health through the concept of access. It is generally accepted that access to health care is an important determinant of health status. Aided by advances in geographic information science and technology, the conceptualization and measurement of access has evolved to include spatial measurement. One of the earliest attempts to model the concept of access was proposed by Andersen  as the "Behavioral Model of Health Services Use." Anderson suggested that access was determined by predisposing, enabling, and need-based factors. This was later expanded to classify access as potential or realized . Penchansky and Thomas  described access in five dimensions: availability, accessibility, accommodation, affordability and acceptability. Kahn  noted that access measures could be sorted into a two-way framework: potential or realized, and spatial or aspatial. Subsequently, Guagliardo  partitioned Penchansky and Thomas' dimensions of access spatially, with availability and accessibility (in a geographic sense) collectively grouped as spatial accessibility, with the remaining factors characterized as aspatial. Guagliardo  also delineated four categories of spatial accessibility measurements: provider-to-population ratios, distances to the nearest provider, average travel impedance to a provider, and gravity models. Talen and Anselin  note that differing methods yield differing results, requiring the researcher to choose the measure of accessibility most suited to the service being measured and the way the population is likely to travel to the service. In the present study, we use reported measurements of distance and time traveled for health or dental care purposes as a measure of geographic/spatial accessibility to health care.
"Travel impedance"  includes measures of Euclidean (straight-line) distance, travel distance along a given path (over a road network, for instance), or travel time between points. By virtue of their point-to-point nature, travel impedance measures have an advantage over provider/population ratios, as they are able to account for border-crossing behaviors [27, 28] and intra-area/local provider variations . Travel time analyses often assume optimal driving conditions, but weather disturbances, rural terrain, and urban traffic congestion can all inflate estimated travel times that are based on observed measurements of distance . Travel impedance measurements are particularly appropriate for rural areas, where provider choices are limited and the nearest provider is usually the one most likely to be utilized.
A recent study by Collia, Sharp, and Giesbrecht  tapped a resource not previously used by health services researchers, the 2001 National Household Travel Survey (NHTS) conducted by the US Department of Transportation (USDOT). The NHTS is used extensively to plan roads and public transportation. The NHTS constitutes the only nationally representative dataset that includes travel for medical or dental care. Further, it includes many measures not included in previous studies of travel for care, including time spent in travel, mode of travel, and perceived barriers stemming from traffic or road conditions. Collia and colleagues  compared travel patterns among younger and older adults using the NHTS. Among their results, Collia and colleagues reported that 1.3% of working age (age 18–64) adults and 2.9% of older (age 65 and over) adults traveled for "medical/dental care." Our research builds on that of Collia and colleagues. Ours is the first study to provide nationally representative estimates of travel distance and travel time to care for rural and urban residents, and among members of racial and ethnic minority groups.