ARTICLE IN PRESS

Disability and Health Journal

-

(2016)

-

www.disabilityandhealthjnl.com

Brief Report

Trips to medical care among persons with disabilities: Evidence from the 2009 National Household Travel Survey Debra L. Brucker, M.P.A., Ph.D.*, and Nicholas G. Rollins, M.S. Institute on Disability, University of New Hampshire, 10 West Edge Drive, Suite 101, Durham, NH 03824, USA

Abstract Background: Persons with disabilities experience multiple barriers to obtaining necessary medical care. Problems with access to transportation and provider choice could lead to longer travel distances and longer travel times to medical appointments. Objective/hypothesis: 1) Persons with disabilities travel further distances to receive necessary care, holding other variables constant. 2) Travel to medical appointments takes a longer amount of time for persons with disabilities, controlling for distance, mode of transportation and other factors. 3) Disability is the key factor influencing access to transportation options, holding other variables constant. Methods: The 2009 National Household Travel Survey (NHTS) is used to examine travel patterns of persons with disabilities as they access medical care. Logistic regressions are run on distance to medical appointments, time taken for travel to medical appointments, and access to private vehicle. Results: There is no difference in the distance traveled, but trips to medical care by persons with disabilities take longer amounts of time than trips taken by persons without disabilities, holding other variables constant. Access to private transportation is similar for both persons with and without disabilities. Conclusions: Persons with disabilities experience longer travel times to receive medical care, despite traveling similar distances and having similar access to private vehicles. Ó 2016 Elsevier Inc. All rights reserved. Keywords: Disability; Transportation; Access to health care

In the United States, persons with disabilities experience multiple barriers to obtaining necessary medical care. Identifying appropriate health care providers who are familiar with disability, offer services in an accessible way, demonstrate clear communication and are covered by available insurance are common concerns.1,2 Such obstacles might limit provider choice to such an extent that persons with disabilities must consider a larger geographic area when choosing providers. This can be problematic as transportation has frequently been identified as a barrier for persons with disabilities who are seeking necessary health care.3e6 Legislative and programmatic responses have made public transportation more accessible to persons with disabilities in recent years, yet transportation barriers remain.7,8,9 Disclosures: This project was funded by the U.S. Department of Health and Human Services (DHHS), National Institute on Disability, Independent Living, and Rehabilitation Research under cooperative agreement H133B130015. The findings and conclusions are those of the authors and do not represent the policy of the DHHS. The authors retain sole responsibility for any errors or omissions. The authors have no conflict of interest to disclose. * Corresponding author. Tel.: þ1 603 862 1643. E-mail address: [email protected] (D.L. Brucker). 1936-6574/$ - see front matter Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.dhjo.2016.01.001

Access to transportation, including public transit and other options, can be influenced by factors that often co-occur for persons with disabilities, such as poverty10 and rural geography.11 As the interplay of these factors makes it difficult to tease out the specific effect that disability has on travel patterns and access to transportation, this paper will apply multivariate techniques to national level household survey data to examine travel patterns and access to transportation for persons with disabilities who are accessing medical care. Persons with disabilities struggle to find physicians who are knowledgeable about their disabilities, particularly in rural areas, and the high rates of public health insurance coverage common among persons with disabilities may limit choice of providers.2 All of the aforementioned factors may necessitate traveling further distances to receive care. In addition, disability is more common in rural areas, areas where it is common to travel greater distances to obtain care.12 This leads to our first hypothesis: 1) Persons with disabilities have to travel further distances to receive necessary care, holding other variables constant. On a related note, barriers to transportation may also lengthen the amount of travel time required to access

ARTICLE IN PRESS 2

D.L. Brucker and N.G. Rollins / Disability and Health Journal

medical care. Other research has confirmed that even though overall health outcomes are poorer for persons with disabilities, persons with disabilities spend more time on health-related activities than persons without disabilities.13 For the general population, travel time to providers is a more meaningful measure of health care access than traditional proxies such as rural-urban residence or providers per capita.14 We will thus test a second hypothesis: 2) Travel to medical appointments takes a longer amount of time for persons with disabilities, controlling for distance and other factors. Access to certain modes of transportation is likely to influence travel patterns. Public transit only accounts for a small piece of the transportation puzzle, as the U.S. is a highly automobile dependent country.15 Furthermore, for the general population, walking or using public transportation to receive medical care is associated with not having a usual source of care and with delays in receiving care.16 We will thus test a final hypothesis: 3) Disability is the key factor influencing access to transportation options, holding economic and other variables constant.

Methods Data The National Household Travel Survey (NHTS), a nationally representative household survey that collects great detail about travel patterns among U.S. households, was most recently administered in 2009.17 The survey tracks detailed information about all trips that occur during an assigned day of the week, including weekends. While the full 2009 sample included information about 1,167,321 trips that were taken by 270,760 people ages 16 and older, we restricted the sample to only include trips that occurred to receive medical care, among persons age 18 and older. We also removed trips that had no documented mode of travel as well as any outliers. Our final unweighted analytic sample included 20,941 trips taken by 18,539 people. Measures Medical trips The NHTS captures information about trip purpose for each trip. Respondents may indicate that travel occurred for ‘‘medical or dental care.’’ We considered such trips as those occurring within a more broadly defined category of ‘‘medical care.’’ Travel limiting medical condition The NHTS currently does not gather detailed data on disability per se but does identify persons who have a medical condition which limits their ability to travel. The survey includes a question that asks whether the respondent has a temporary or permanent condition or handicap that

-

(2016)

-

makes it difficult to travel outside the home. Approximately 11 percent of respondents answered ‘‘Yes’’ to this question in 2009. For purposes of assessing our hypotheses, we consider this answer as positive disability status. Long trip distance Mileage was estimated based on self-report, although the NHTS makes certain adjustments to clean up self-reported mileage information. Overall, the mean distance for trips to medical care was slightly over nine miles. Miles was a highly skewed variable, however, with most trips shorter than 20 miles (32.2 km). Other researchers who have used the NHTS have addressed this skewed data by creating a categorical variable.18e20 For our analyses, we recoded this variable into a long trip distance variable, assigning a value of one for trips that ranged roughly in the top 25th percentile of all trips. This included trips of 11 or more miles (17.7 or more km). A value of zero was assigned to trips of shorter lengths.a Long travel time In the data set, starting and ending times of trips were recorded using military time. These values were used to derive the length of the trip in minutes. Similar to miles traveled, length of travel time was skewed. Average travel time to medical care was approximately 20 min. Whereas others have used a cut-off of 30 min as a measure of high travel burden,21 we recoded this variable into a long travel time variable by assigning a value of one to trips that had travel times falling roughly in the top 25th percentile in terms of travel time.b This included trips of 26 or more minutes. Transportation As we were interested in exploring differences among specific modes of transportation, we included variables for each type of transportation in our first two sets of regressions with private vehicles as the reference group. Separate variables indicated bus/train, special transit for persons with disabilities, bike/walk, taxi, and other. We created a binary private motorized vehicle variable to identify trips that occurred by car, van, sport utility vehicle, pick-up truck, other trucks, recreational vehicle, motorcycle, or light electric vehicle.c a

We tested the sensitivity of this long trip distance cut-off when conducting our regression analyses. Results obtained using a cut-off of the top ten percent of all trips did not change the significance level of our focal variable for any of the regressions. Results are available from the authors. b We tested the sensitivity of this long travel time cut-off when conducting our regression analyses. Results obtained using a cut-off of the top ten percent of all trips changed the significance level of our focal variable for one of the three regressions. Results are available from the authors. c Bicycle and walking are not considered transit or private motor vehicle.

ARTICLE IN PRESS D.L. Brucker and N.G. Rollins / Disability and Health Journal

-

(2016)

3

regression included all trips. The second and third regressions included those trips taken by persons living in urban and rural areas, respectively, to account for travel differences in these areas. For all regressions, long distance trips were defined as those greater than or equal to 11 miles (17.7 km), the top 25 percent of all medical trips in the sample. Our first hypothesis is not confirmed, as the presence of a medical condition was not associated with an increased probability of taking a longer trip in any of the regressions. In fact, the only significant difference related to the presence of medical condition was found in the urban-specific regression, where trips taken by persons with a medical condition were less likely to be long trips. Overall, trips taken by older persons and by persons living in urban areas were significantly less likely to be long trips. In addition, trips taken by bus or train, as well as trips that were walked or taken by bicycle, as opposed to trips taken by private vehicle, were significantly less likely to be long trips. Poverty was not significantly associated with trip distance to medical care. Table 2 includes results from the logistic regressions used to predict trips that took a long time. Again, results are first shown for all trips and then for trips occurring by travelers who were living in urban and rural areas, respectively. Long duration trips for medical care were defined as greater than 25 min, or the top 25 percent of all medical trips in the sample. Our second hypothesis was confirmed, as the presence of a medical condition which limits travel was significantly associated with an increased probability of taking a longer trip (OR 1.43, p ! .01), even when controlling for distance traveled, mode of transportation and other factors. This association was significant for the subset of trips taken in urban areas (OR 1.41, p ! .05), but was not significant for the subset of trips taken in rural areas. Overall, the odds ratios for

Demographic variables For our regressions, demographic variables relate to the traveler, the person that took a specific trip. Age was measured as a continuous variable. Gender, race, and ethnicity were measured as binary variables, with values of one indicating female, non-white, and Hispanic. A traveler was considered to live in an urban area if the household was located in an urban area, an urban cluster, or an area surrounded by urban areas. Given that the NHTS only provides income information in categorical terms, we measured poverty following guidelines used by the U.S. Department of Transportation in a recent brief.22

Results Our descriptive findings suggest that adults having a medical condition limiting travel were significantly more likely to be female, older, and non-white. Whereas 70 percent of persons without a limiting medical condition reported being employed, only 21 percent of persons with a medical condition were employed. Thirty-one percent of persons with a medical condition were in poverty, compared to 12 percent of persons without a medical condition. From a travel standpoint, adults with a medically limiting condition took significantly fewer trips per day than persons without a medically limiting condition. In addition, persons with a limiting medical condition had significantly lower mean numbers of household vehicles and lower percentages of household adults who were drivers. Trips by private motorized vehicle were significantly less likely to be taken by persons with a medical condition. This result held for all trips and for trips to medical care. Table 1 shows results from logistic regressions used to predict long distance trips for medical care. The first

Table 1 Logistic regression of long distance traveled (in miles) for routine medical care (N(trips) 5 20,941) All (N(trips) 5 20,941) Urban (N(trips) 5 14,976)

Intercept Female Non-White Hispanic Age (years) Medical condition Number of HH adults In poverty Urban Bus/train Special transit Bike/walk Taxi Other transit Private vehicle (ref group)

-

Rural (N(trips) 5 5965)

OR

se

p

OR

se

p

OR

se

p

1.06 0.84 1.06 0.94 0.99 0.82 1.09 1.29 0.32 0.39 1.17 0.01 0.51 1.37 e

0.28 0.08 0.13 0.15 0.00 0.09 0.06 0.17 0.03 0.12 0.37 0.01 0.49 0.46 e

NS NS NS NS * NS NS NS *** ** NS *** NS NS e

0.44 0.80 1.07 0.93 0.99 0.73 1.07 0.90 e 0.38 1.22 0.01 0.62 1.39 e

0.13 0.09 0.15 0.17 0.00 0.10 0.07 0.15 e 0.11 0.47 0.01 0.64 0.52 e

*** * NS NS * * NS NS e ** NS *** NS NS e

0.64 0.93 1.08 1.20 1.00 1.18 1.11 2.52 e 0.62 0.80 0.00 0.15 3.45 e

0.31 0.15 0.24 0.31 0.01 0.21 0.11 0.62 e 0.47 0.64 0.00 0.11 2.23 e

NS NS NS NS NS NS NS *** e NS NS *** * NS e

Trips taken by persons age 18 and over. Long trip: miles > 11. *p ! .05; **p ! .01; ***p ! .001. Source: Authors’ analysis of 2009 NHTS data.

ARTICLE IN PRESS 4

D.L. Brucker and N.G. Rollins / Disability and Health Journal

-

(2016)

Table 2 Logistic regression of long travel time (in minutes) to routine medical care (N(trips) 5 20,941) All (N(trips) 5 20,941) Urban (N(trips) 5 14,976)

Intercept Female Non-White Hispanic Age (years) Medical condition Number of HH adults In poverty Urban Distance (miles) Bus/train Special transit Non-motorized vehicle Taxi Other transit Private vehicle (ref group)

-

Rural (N(trips) 5 5965)

OR

se

p

OR

se

p

OR

se

p

0.00 0.90 1.04 1.33 1.02 1.43 0.98 1.35 1.26 1.33 14.92 15.46 4.93 2.05 2.52 e

0.00 0.10 0.15 0.26 0.00 0.19 0.07 0.25 0.17 0.01 6.24 7.84 1.38 1.25 1.33 e

*** NS NS NS *** ** NS NS NS *** *** *** *** NS NS e

0.00 0.78 1.11 1.23 1.02 1.41 1.03 1.28 e 1.33 15.72 22.01 5.32 2.41 2.83 e

0.00 0.10 0.17 0.26 0.00 0.21 0.07 0.27 e 0.01 6.78 10.55 1.55 1.48 1.52 e

*** * NS NS *** * NS NS e *** *** *** *** NS NS e

0.01 1.46 0.71 2.64 1.02 1.59 0.78 1.61 e 1.33 14.01 0.47 0.94 0.02 0.53 e

0.00 0.30 0.24 1.06 0.01 0.41 0.12 0.53 e 0.02 10.16 0.43 0.74 0.03 0.37 e

*** NS NS * ** NS NS NS e *** *** NS NS * NS e

Trips taken by persons aged 18 and over. Long travel time (minutes > 26). *p ! .05; **p ! .01; ***p ! .001. Source: Authors’ analysis of 2009 NHTS data.

trips taken by a bus or train and special transit for persons with disabilities were high and close to one another (14.92 and 15.46, respectively), suggesting that both modes of transportation are much more likely to be associated with long trip times. Similar results were found in the regression limited to urban areas but differed in the regression for rural areas. In rural areas, only trips occurring by bus/train were significantly associated with an increased likelihood of long travel times. The results from the logistic regression used to test our third hypothesis are presented in Table 3. The presence of a medical condition which limits travel was not associated with whether or not a trip occurred by private motorized vehicle. Trips taken by persons who were non-white, living in poverty, or living in urban areas were significantly less likely to occur by private motorized vehicle. The likelihood of trips occurring via private motorized vehicle increased as the number of household adults increased.

Table 3 Logistic regression of private motorized vehicle (N(trips) 5 20,977) OR se p Female Non-White Hispanic Age in years Medical condition Number of HH adults In poverty Urban

1.01 0.54 0.62 1.01 0.76 1.31 0.36 0.27

Civilian, non-institutionalized, aged 18 and over. *p ! .05; **p ! .01; ***p ! .001. Source: Authors’ analysis of 2009 NHTS data.

0.19 0.11 0.15 0.01 0.14 0.14 0.08 0.06

NS ** * NS NS ** *** ***

Discussion In sum, our findings suggest that persons with a disability experience longer travel times to access medical care. The actual distance traveled does not vary from the distance traveled by others, however. On a positive note, access to private modes of transportation is similar between persons with and without disabilities. We discuss each of these findings in turn below. First, we find no evidence that trips taken by persons with a travel limitation are of greater distance than trips taken by persons without a travel limitation. This result could imply either that the geographic availability of providers is similar between these two groups or that the amount of miles that people are willing to travel is similar between the two groups. Prior research has established that differences in distance to care is associated with fewer regular care check-ups, but is not significantly related to numbers of chronic care and acute care visits.23 As the NHTS does not include such detailed data about types of care, we can only state that the appointments that were attended by persons with disabilities were within the same geographic catchment area as appointments attended by persons without travel limitations. Within the urban landscape, however, trips taken by persons with disabilities were more likely to be of shorter distance. This could imply that persons with a medical condition have closer access to needed services or that they are not willing to travel further distances for care. Future research, using data on travel as well as measures of need for care and health insurance coverage, can provide more detail about the relationship between distance and access to care. Our second set of results suggest that trips for medical care taken by persons with disabilities are likely to take

ARTICLE IN PRESS D.L. Brucker and N.G. Rollins / Disability and Health Journal

a longer amount of time than trips taken by persons without disabilities even when controlling for distance, mode of travel, and other factors. Qualitative research could provide a more detailed understanding of why trips are taking longer for persons with travel limitations, uncovering differences across specific disability types and other possible barriers that are undocumented in the NHTS. Not surprisingly, use of public transit was associated with increased odds of taking a trip of longer duration, when compared to trips completed via private vehicles. Future research can explore exactly what components of public transportation are contributing to this increase in travel time. On- and off-boarding impediments, difficult or dangerous street crossings, inaccessible stops or ticket purchase procedures, and increased need for transfers can all increase travel time for persons with disabilities. Similarly, trips taken by special transit for persons with disabilities were about 15 times more likely to result in a long trip than trips taken by private vehicles. These results suggest that special transit may not be the most expedient means of transportation for persons with travel limitations. Lastly, our results show that having a disability is not associated with decreased access to private transportation. Poverty, race and urban living are, however, associated with decreased access to private modes of transportation. Where increasing access to private modes of transportation is not feasible, land use planners can work to ensure that medical facilities are located near public transit hubs, improving overall access. The data we used for this analysis, while having strengths in the level of detail about travel, has some limitations which warrant discussion. Omitted variables, including information about health insurance coverage and actual need for care, limits our ability to make conclusions about whether transportation concerns are limiting access to services. In addition, a more detailed method of capturing information about disability would help to frame the research results, as transportation issues likely vary by disability type. Additional research on the relationships between travel distance, travel time, mode of transportation and access to health care services is required to further understand how the transportation needs of persons with disabilities can best be addressed.

Acknowledgments The authors would like to thank John Renne and Adella Santos for helpful comments and suggestions.

References 1. McDoom MM, Koppelman E, Drainoni M-L. Barriers to accessible health care for Medicaid eligible people with disabilities: a comparative analysis. J Disabil Policy Stud. 2014;25(2):154e162.

-

(2016)

-

5

2. Iezzoni LI, Killeen MB, O’Day BL. Rural residents with disabilities confront substantial barriers to obtaining primary care. Health Serv Res. 2006;41(4 Pt 1):1258e1275. 3. Lia-Hoagberg B, Rode P, Skovholt CJ, et al. Barriers and motivators to prenatal-care among low-income women. Soc Sci Med. 1990;30(4): 487e495. 4. McCray T. Delivering healthy babies: transportation and healthcare access. Plann Pract Res. 2000;15(1/2):17e29. 5. Melnikow J, Alemagno S. Adequacy of prenatal care among inner-city women. J Fam Pract. 1993;37(6):575e582. 6. Waller M. High Cost or High Opportunity Costs? Transportation and Family Economic Success (Center on Children & Families #35). Washington, DC: The Brookings Institution; 2005. 7. National Council on Disability (NCD). Transportation Update: Where We’ve Gone and what We’ve Learned. Washington, DC: NCD; 2015. 8. U.S. Government Accountability Office (GAO). GAO-15e110: Transportation Disadvantaged Populations: Nonemergency Medical Transportation Not Well Coordinated, and Additional Federal Leadership Needed. Washington DC: GAO, http://gao.gov/assets/670/667362.pdf; 2014. 9. U.S. Department of Transportation. Grants and Programs. Washington, DC: US Department of Transportation, http://www.transportation.gov/ livability/grants-programs#ResearchandMiscellaneous; 2015. 10. Brucker DL, Mitra S, Chaitoo N, Mauro J. More likely to be poor whatever the measure: working-age persons with disabilities in the United States. Soc Sci Q. 2015;96(1):273e296. http://dx.doi.org/10.1111/ssqu.12098. 11. Seekins T, Greiman L. Map Facts: Disability in Rural America (Rural Facts). Working Paper. University of Montana Rural Institute. Accessed 26.07.14, http://rtc.ruralinstitute.umt.edu/_rtcBlog/wp-content/ uploads/MapFacts.pdf; 2014. 12. Probst JC, Laditka SB, Wang JY, Johnson AO. Mode of Travel and Actual Distance Traveled for Medical and Dental Care by Rural and Urban Residents. Columbia, SC: South Carolina Rural Health Research Center; 2006. 13. Anand P, Ben-Shalom Y. How do working-age people with disabilities spend their time? New evidence from the American Time Use Survey. Demography. 2014;51:1977e1998. 14. Fortney J, Rost K, Warren J. Comparing alternative methods of measuring geographic access to health services. Health Serv Outcomes Res Methodol. 2000;1(2):173e184. 15. Renne J, Bennett P. Socioeconomics of urban travel: evidence from the 2009 National Household Travel Survey with implications for sustainability. World Transp Policy Pract. 2014;20(4):7e27. 16. Rask KJ, Williams MV, Parker RM, McNagny SE. Obstacles predicting lack of a regular provider and delays in seeking care for patients at an urban public hospital. J Am Med Assoc. 1994;271(24):1931e1933. 17. U.S. Department of Transportation, Federal Highway Administration. 2009 National Household Travel Survey. URL, http://nhts.ornl.gov; 2010. 18. Trowbridge MJ, McDonald NC. Urban sprawl and miles driven daily by teenagers in the United States. Am J Prev Med. 2008;34(3): 202e206. http://dx.doi.org/10.1016/j.amepre.2007.11.013. 19. Jacoby I. Geographic distribution of physician manpower: the GMENAC (Graduate Medical Education National Advisory Committee) legacy. J Rural Health. 1991;7:427e436. 20. Ricketts TC, Savitz LA, Gesler WM, Osborne DN. Using Geographic Methods to Understand Health Issues. Washington DC: U.S. Department of Health and Human Services; 1997. 21. Pathman DE, Ricketts TC, Konrad TR. How adults’ access to outpatient physician services relates to the supply of primary care physicians in the rural Southeast. Health Serv Res. 2006;41(1):79e102. http://dx.doi.org/10.1111/j.1475-6773.2005.00454.x. 22. U.S. Department of Transportation. FHWA NHTS Brief: Mobility Challenges for Households in Poverty. Washington, DC: US Department of Transportation, FHWA; 2014. 23. Arcury TA, Gesler WM, Presisser JS, Sherman J, Spencer J, Perin J. The effects of geography and spatial behavior on health care utilization among the residents of a rural region. Health Serv Res. 2005;40(1): 135e156. http://dx.doi.org/10.1111/j.1475-6773.2005.00346.x.

Trips to medical care among persons with disabilities: Evidence from the 2009 National Household Travel Survey.

Persons with disabilities experience multiple barriers to obtaining necessary medical care. Problems with access to transportation and provider choice...
562B Sizes 0 Downloads 7 Views