Cost-Benefit Analysis of Providing Non-Emergency Medical Transportation (2005)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Chapter 2: Literature Review To develop a better understanding of the status of non-emergency medical transportation in the United States, particularly in regards to unmet trip needs, a thorough literature search of the field was conducted. This was accomplished using local resources (e.g., University of Michigan libraries, including that belonging to the University of Michigan Transportation Research Institute) and national and global search engines (such as TRIS, GOOGLE, Wilson, ProQuest, and Medline/PubMed). This search resulted in more than 200 sources. While not all of these are referenced in this discussion of the main findings and views present, most are summarized in Appendix B: Annotated Bibliography. The results emerging from our literature review focus on the most recent studies and are organized into four major sub-units: • Identification of the transportation-disadvantaged population • Evidence of unmet non-emergency medical (NEM) trip needs • Consequences of this unmet need • Estimates of the costs and benefits of meeting this unmet need In addition to these four major thrusts culled from the literature, there is substantial use of published literature that addresses the cost-effectiveness of well- managed care for critical medical conditions affecting the transportation-disadvantaged population in Chapter 7. 2.1 Identification of the Transportation-Disadvantaged Population Because no single definition of the “transportation disadvantaged” has yet been universally accepted, the literature from the health and transportation sectors can be used to support differing estimates of the size of the transportation-disadvantaged population. Indeed, much of the literature avoids the term altogether and instead simply documents discrepancies in transportation access associated with socio- economic (e.g., household income), demographic (e.g., age), and geographic (e.g., urban vs. rural) factors. Thus, at one extreme, any household that does not own a vehicle might be defined as transportation disadvantaged, and this amounts to 8.3 percent of all households in the U.S. (Pucher and Renne, 2003). Furthermore, 88 percent of Americans aged 15 or higher report that they are drivers, leaving 12 percent who do not operate personal vehicles (U.S. DOT, Bureau of Transportation Statistics, 2003a). When they examined population subgroups, however, Pucher and Renne found that 26.5 percent of households with incomes less than $20,000 do not own a vehicle. They also found that members of this income group were far more likely to use public transit (4.6 percent of all trips compared with an average of 1.7 percent for all Americans) and non-motorized (walk or bicycle) modes (17.0 percent of all trips compared with 10.4 percent for all Americans). Further illuminating who the transportation disadvantaged are, the BTS reported that households with no vehicles also are disproportionately renters, located in urban areas, and made up of a single person (U.S. DOT, Bureau of Transportation Statistics, 2003a). Final Report 10

Race and ethnicity also are associated with being transportation disadvantaged. Using data from the 2001 National Household Travel Study (NHTS), Pucher and Renne (2003) found that African Americans and Hispanics have much lower mobility and use public transit at much higher rates than the general population. Furthermore, Schweitzer and Valenzuela (2004) found that both low-income and minority communities suffer from a lack of access to transportation and from a litany of other ill effects associated with transportation (such as air pollution, noise, and fewer jobs in the transportation sector) at higher rates than do other population groups. Beyond income and ownership barriers to transportation, studies also show that age and location are important factors in defining the transportation disadvantaged. Although studies have shown that older adults and residents of rural areas continue to rely on personal vehicles for the vast majority of their transportation needs (Rosenbloom, 2003; Pucher and Renne, 2004)—for example, Americans over 65 years of age make about 90 percent of their trips by car and 97 percent of rural households own at least one car—they often have few, if any, options when the car is not available. Indeed, Rosenbloom (2003), citing work done by the Community Transportation Association of America (CTAA), reports that about 40 percent of rural counties have no public transportation and only 14 percent of the rural elderly have transportation service available within one half-mile of their residence. 2.2 Evidence of Unmet Need for NEMT Published literature both convincingly documents the existence of unmet transportation needs and provides insights into the demographic and other factors associated with these unmet needs. These factors include age, poverty, disability, geographic location, and race. A study commissioned by the Children’s Health Fund (Zogby International, 2001) found that nine percent of children in families with incomes less than $50,000 per year miss essential medical appointments due to a lack of transportation, regardless of their insurance status. In fact, at least two other studies have also shown that lack of transportation is a problem even after accounting for insurance status (Aved et al., 1993; Braverman et al., 2000). Focusing on the population below age 65 in Dayton, Ohio, Ahmed et al. (2001) found that 16 percent of respondents reported that finding transportation for medical care was “hard” and another 15 percent reported that it was “very hard.” In a study that focused on the higher end of the age spectrum (defined as those over age 50), O’Malley and Mandelblatt (2003) found that patients who were over age 50 and whose household income was less than 200 percent of the poverty line were nearly twice as likely to delay care due to transportation and/or time issues compared to all patients above age 50. Furthermore, they also found that transportation and time issues were nearly as important a barrier as cost (14.3 percent vs. 18.8 percent) for the same group (>50 years and <200 percent of poverty level). Another study (Sipe et al., 2004), based on interviews with visitors to a pediatric clinic at a large, urban hospital, found that 60 percent of respondents had previously missed or arrived late for an appointment due to transportation difficulties. Finally, another study revealed that not showing up for appointments in the past due to transportation problems was predictive of not showing up in the future (Paul and Hanna, 1997), indicating a recurring problem for the same individuals. Final Report 11

In rural areas, access to healthcare can be even more difficult to obtain than it is elsewhere, and numerous studies (Flores et al., 1998; Ide et al., 1993; Larson et al., 2004; McClure et al., 1996; Mulder et al., 2000) have documented these difficulties. In a study of one rural county, for example, Walker (2002) found that 40 percent of patients missed medical appointments and 28 percent could not get to a pharmacy because of transportation barriers. Contributing to the problem, rural areas are less likely to have public transportation available; indeed, the American Public Transportation Association (APTA) reported that 41 percent of rural Americans have no access to public transit (APTA, 2003). Another contributor to access problems in rural areas is distance; medical facilities tend to be farther away from patients, on average, and distance is associated with missed care. In one study, patients who lived more than 20 miles from the site of care were twice as likely to miss scheduled appointments (Ide et al., 1993) as those who lived closer. While large-scale, nationally representative studies performed by, or on behalf of, the U.S. Department of Transportation have not specifically examined the nexus of non- emergency medical transportation (NEMT) and unmet trip needs for healthcare, two studies shed some light on the potential scale of the problem. One of these studies, the 2001 National Household Travel Survey (NHTS), found that 8.6 percent of respondents reported having a medical condition that limits their travel, regardless of trip purpose (U.S. DOT, Bureau of Transportation Statistics, 2003a). Perhaps more tellingly, a second study (U.S. DOT, Bureau of Transportation Statistics, 2003b) revealed that 3.5 million Americans never leave their homes. Of these, 1.9 million have disabilities, and roughly 528,000 of these have transportation difficulties. Race and ethnicity, too, are associated with transportation access problems. Research on the effects of race on access to care has shown that 10 to 20 percent more racial minorities than whites are transportation disadvantaged (Friedman et al., 2003; Moran et al., 2003). Looking specifically at access to healthcare, the Institute of Medicine (2002) cited access issues as a reason that minorities receive lower-quality healthcare than do non-minorities, even when the minorities have equal levels of insurance coverage. 2.3 Consequences of Unmet NEMT Needs The health of individuals who fail to obtain medical care due to transportation barriers depends to some extent on whether the missed care was preventive or treatment for an existing (or chronic) condition. In the preventive arena, lack of transportation can lead to under-immunization (Yawn et al., 2000), difficulties in administering screening programs (Lavizzo-Mourey et al., 1994), failure to attend pediatric check- ups (Specht and Bourguet, 1994), and lack of prenatal care (Aved et al., 1993; Braverman et al., 2000; McCray, 2000). In the realm of chronic problems, numerous studies have documented inadequate care due to lack of transportation. Conover and Whetten-Goldstein (2002), for example, found that 16.7 percent of AIDS and HIV patients reported difficulties in obtaining transportation and as a result were much less likely to have a primary-care physician or to get regular care. Patients suffering from diabetes who missed more than 30 percent of scheduled appointments suffered significantly worse health outcomes than those who kept their appointments (Karter et al., 2004). Furthermore, Final Report 12

an earlier study, concerned with adherence to standards of care for diabetes, found that transportation problems were among the most frequently cited reasons for missed care (Jorgensen et al., 2002). Additionally, asthmatic patients entering the emergency room have been shown to be much less likely to obtain follow-up care if they do not have access to transportation (Baren et al., 2001; Ebbinghaus and Bahrainwala, 2003; Smith et al., 2002). A detailed analysis sponsored by the U.S. Agency for Healthcare Research and Quality highlights healthcare access problems faced by rural residents (Gresenz et al., 2004). It finds that the distance between the rural uninsured and a variety of safety net providers (hospital emergency rooms, public hospitals, migrant health centers, etc.) is a significant variable explaining lower healthcare service utilization. The authors of this study present strong evidence that facilitating transport will improve access to care by the rural uninsured. Importantly, this study relies on a preferred dataset derived from the Medical Expenditure Panel Survey—a data source that is also heavily used in the current study. Not having a car is one factor that hinders access. Urban and rural areas that have some form of public transportation may not have routes to medical care, especially for the most economically disadvantaged neighborhoods (Hobson and Quiroz- Martinez, 2002). Of patients riding public transportation to get medical care, 86 percent reported missing an appointment due to transportation barriers, and 95 percent reported arriving late, as compared with 27 percent and 43 percent, respectively, among patients with cars (Sipe et al., 2004). Another study showed that patients diagnosed with asthma were much less likely to return for a follow-up appointment with a primary-care physician if they relied on public transportation, friends, or walking to access appointments than were patients with their own transportation (Baren et al., 2001). Older adults are distinctly affected by problems accessing NEMT. Of Americans over age 65, 21 percent do not drive any longer, and these reported taking 15 percent fewer trips to the doctor than did older adults who still drive (Bailey, 2004). In addition to documenting age-related access problems, studies have also shown that the prevalence of more than one health problem (co-morbidity) is also age dependent, suggesting that many older Americans who experience transportation barriers also suffer from multiple chronic health conditions (Bayliss et al., 2003). Indeed, a new body of literature addresses the issues confronted by the roughly 125 million Americans with a chronic illness and 60 million with multiple chronic conditions (Anderson, 2002; Anderson and Knickman, 2001; Burton et al., 2004; Partnership for Solutions, 2004; Partnership for Solutions, 2002). This literature demonstrates that these individuals are more likely to be hospitalized, see a variety of physicians, take several prescription drugs, and be visited at home by health workers. Due to uncoordinated care, they experience unnecessary hospitalization, duplicate tests, conflicting clinical advice, and adverse drug reactions. Furthermore, this group with multiple chronic conditions is estimated to account for 57 percent of total healthcare spending nationwide (Burton et al., 2004). Final Report 13

2.4 Estimates of Costs and Benefits of Meeting Unmet NEMT Needs Meeting the needs of the transportation-disadvantaged population requires enhanced transportation resources plus additional healthcare—both of which increase costs. These services, however, are hypothesized to result in improved health outcomes, such as reduced need for high-priced emergency care, and better quality of life for affected individuals. As evidence of the former claim, Moran (2003) found that 61 percent of caretakers who experienced transportation problems associated with bringing children to the emergency room reported missing care that the caretaker believed was necessary. Other studies have shown that a high percentage of emergency room visits are actually for non-emergency conditions (Beland et al., 1998; U.S. GAO, 2000a; Burkhardt and McGavock, 2002). An international comparison shows weak primary care coordination, lower access to a usual source of care, and a lack of long-standing patient-physician relationships in the U.S. compared with other countries (Schoen et al., 2004). As discussed above, these shortcomings can lead to emergency care for non-emergency conditions. Nationally, the General Accounting Office (U.S. GAO, 2003) identified 62 federal programs that “fund transportation services for the transportation disadvantaged,” accounting for about $2.4 billion in federal expenditures alone in fiscal year 2001, plus whatever is spent by states and localities, an amount that GAO estimates to be in the hundreds of millions of dollars. Of course, not all of this transportation is for non-emergency medical trips. Looking only at Medicaid transportation assistance, a comprehensive study of state programs (Stefl and Newsom, 2003) showed that California, the most populous state, alone spent about $95 million of federal and state money (50-50 match) in FY 2002. Florida, another populous state, spent about $69 million in FY 2002, about 56 percent of which were federal dollars; with these funds, Florida provided roughly 3.4 million one-way trips, at an average cost of about $19.65 per trip. Delaware, a much smaller and less populous state than either California or Florida, spent $8.5 million of federal and state funds (again, 50-50 match), and provided 544,000 one-way trips for this money (about $15.65 per one-way trip). Measuring the benefits of providing transportation is far more difficult than measuring its costs. Nonetheless, some studies have succeeded in developing estimates. These studies have shown that interventions that include transportation to increase attendance at appointments often reported positive results, including fewer missed appointments, reduced length of stay, and fewer emergency room visits (Block and Branham, 1998; Baren et al., 2001; Ebbinghaus and Bahrainwala, 2003; Friedhoff, 1999; Friedmann, Lemon, and Stein, 2001; Messeri et al., 2002; Rimmer et al., 2002; Sherer et al., 2002). Examining dollar benefits of improved access, Burkhardt and colleagues (1998) focused on transportation for dialysis patients in Charlottesville, Virginia. They found significant benefits, even though they considered only the cost differential between trip provision by public transit and costs for the same trips provided by private, wheelchair van services. Thus, they did not consider the quality-of-life benefits from these trips, assuming that patients would have made these trips one way or the other due to the life-and-death nature of dialysis. This study inherently failed to investigate the effects of providing otherwise Final Report 14

missed trips—its authors assumed all trips would be made no matter what—so the costs of transportation provision did not need to be balanced against the net benefits of treatment. For near-emergency services such as dialysis, however, one would expect even greater benefits for providing otherwise missed trips. The literature points to a clear, policy-relevant problem of inadequate transportation serving as a key, contributing factor to a lack of access to medical care. Nevertheless, despite a burgeoning literature on the healthcare access problems in America—typically labeled as a “crisis” of un- or under-insured—only a fraction of the literature specifically relates transportation barriers to the larger problem of healthcare access. Indeed, two articles appearing in a prestigious health journal (Inquiry) that attempt to comprehensively link community factors with the health access problems faced by lower-income adults do not even mention the word “transportation” (Davidson et al., 2004; Brown et al., 2004). This gap in the professional literature strongly attests to the value and novelty of the present study. 2.5 Additional Use of Literature in This Report The literature review is intended to be representative of the existing literature and the views and findings presented therein. It is not, therefore, meant to be comprehensive of all literature in the healthcare and transportation fields bearing on non-emergency medical transportation (NEMT). In particular, a large swath of the literature addressing the cost-effectiveness of care for specific disease conditions is referenced in Chapter 7. Furthermore, literature specific to other aspects of this study is referenced in other chapters, where these citations are most relevant and useful. Finally, there is an extensive annotated bibliography of literature included as Appendix B. Final Report 15