- Email: [email protected]
Disability, Insurance Coverage, Area Deprivation and Health Care: Using Spatial Analysis to Inform Policy Decisions
Available online at www.sciencedirect.com
ScienceDirect Procedia Environmental Sciences 36 (2016) 20 – 25
International Conference on Geographies of Health and Living in Cities: Making Cities Healthy for All, Healthy Cities 2016
Disability, Insurance Coverage, Area Deprivation and Health Care: Using Spatial Analysis to Inform Policy Decisions Henan Lia,* a
Brandeis University, 415 South Street MS 035, Waltham, MA 02453, USA
Abstract In the United States, people with disabilities with public insurance are often unable to find suitable health care providers nearby, resulting in needing to travel long distances to large health centres to access necessary health care. This barrier is even more pronounced to those with disabilities living in deprived areas. Using Census data and Area Deprivation Index (ADI) developed by University of Wisconsin-Madison’s ADI project, OLS and Geographically Weighted Regression (GWR) regression analyses showed that high area deprivation index (ADI) and high percentage of public-only health insurance coverage predict disability prevalence. While ADI, public-only insurance coverage percentage and disability prevalence do not significantly predict the number of hospitals, they slightly improved the fit of the GWR model, which was significantly predictive. Future research should continue exploring related factors, and address the challenges of having limited access to hospitals in poorer suburban and rural areas faced by people with disabilities, who are also more likely to have public-only health insurance. © by Elsevier B.V. This is an openB.V. access article under the CC BY-NC-ND license ©2016 2016Published The Authors. Published by Elsevier (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of Healthy Cities 2016. Peer-review under responsibility of the organizing committee of Healthy Cities 2016 Keywords: Disability; Health insurance; Area deprivation, Health care, Geographic information system (GIS)
1. Introduction Access to health care is a topic of great importance to many socially-disadvantaged groups in the United States. The Affordable Care Act (ACA) mandates that all Americans have health insurance and that health care be accessible to all, regardless of geographic location, socioeconomic status, or disability1. However, for those with disabilities
* Corresponding author. Tel.: +1-781-736-3858; fax: +1-781-736-3773. E-mail address: [email protected]
1878-0296 © 2016 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of Healthy Cities 2016 doi:10.1016/j.proenv.2016.09.004
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
living in deprived areas, considerable barriers continue to exist2, 3. It was hypothesized that being in deprived areas and having only public health insurance (Medicare, Medicaid etc.) coverage contributed significantly to the access barriers. Economically deprived areas typically have fewer hospitals, resulting in poorer health care access for local residents4. Similarly, people with disabilities with public insurance are often unable to find suitable health care providers nearby, resulting in needing to periodically travel long distances to large hospitals to access necessary routine or specialty care5, 6. The aim of this present study is to explore the predicting effects of disability prevalence, public-only insurance coverage, and area deprivation index (ADI) on the access to hospitals in the New England region of the U.S. This study examines the New England region, which is an area with a total population of 11,477,279 (2014 data), spanning six Northeastern U.S. states: Connecticut (CT), Maine (ME), Massachusetts (MA), New Hampshire (NH), Rhode Island (RI) and Vermont (VT)7. Despite having several highly-ranked health care systems in the area, many residents with disabilities still experience a great deal of difficulties accessing care due to factors such as lack of reliable transportation and traffic congestions8. For instance, a recent report showed that about 54% of Massachusetts residents with disabilities and their family members or caregivers indicated that transportation to health care was very problematic 9. Respondents reported they cannot access needed medical care when affordable, reliable transportation is unavailable or inaccessible9. Being far away from care providers and subsequently lacking timely access to health care can translate into delayed or foregone primary, preventive and specialist care10, 11. Delayed or foregone care are associated with multiple negative outcomes for people with disabilities, such as development of preventable secondary conditions12, prolonged untreated active diseases13 and higher risks of physical and mental health problems6. Studying the geographical aspects of existing systems is of critical significance for the New England disability community9. 2. Data The data analyzed in this study were 2010-2014 U.S. Census/American Community Survey (ACS) 5-year estimate data7, as well as Area Deprivation Index data, from University of Wisconsin-Madison’s HIPxChange program14, 15. Disability is defined as having limitations in vision, hearing, cognitive, ambulatory, self-care, or independent living7. Area deprivation is defined as the socioeconomic deprivation experienced by a neighborhood, which takes into accounts factors such as median family income, home/rent value, education/profession/employment, and percent of households with access to motor vehicles. Public insurance coverage is defined as only having Medicare or Medicaid. Names and locations of hospitals located in the six states were provided through the Geographic Names Information System (GNIS)16. The analytical framework is presented below (Fig.1).
Fig. 1. Analytical Framework
21
22
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
3. Methods Maps were generated to present a descriptive overview of disability prevalence, ADI, public-only insurance coverage, number of hospitals on the county level. Ordinary Least Squares (OLS) regressions and Geographically weighted regressions (GWR) were conducted to examine the relationship between independent and dependent variables. The fixed kernel type and Akaike Information Criterion (AICc) was used in the GWR analyses. Esri ArcGIS Pro version 1.2 was used to manage and analyse maps and data17. 4. Results
Fig. 2. Disability Prevalence, Area Deprivation Index, Percent of population with Public Insurance only and Number of Hospitals
Descrptive results were presented in Fig. 2. The value categories of Maps 1, 3 and 4 were generated by quantiles while ADI values (Map 2) were broken into 10 categories by natural breaks. Natural break was chosen because it would group similar values together and maximize the differences between categories.
23
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
In the top half of Fig. 2, geographical patterns of associations can be observed between high disability prevalance and high ADI. The bottom half of Fig. 2 showed that relatively fewer hospitals are available where high percentages of the population only have public insurance. Overall, Fig.2 demonstrated the unique circumstances faced by people with disabilities living in deprived areas seeking health care, who only have public health insurance. Table 1. OLS Regression: Disability Prevalence as the Dependent Variable Independent Variables
Coefficients (ȕ)
t
P value
ADI
0.067
6.69
<.001
Public coverage
0.221
4.21
<.001
Adjusted R2 .641
Table 1 presents the OLS regression results with disability prevalence as the dependent variable. Area deprivation index and percentage of public-only health insurance coverage predict disability prevalence in the same direction (e.g. more deprived areas also have higher prevalance of disability). This outcome supports the findings of previous studies18 that demonstrated the connections between poverty, public insurance and disability. Table 2. OLS Regression: Number of Hospitals as the Dependent Variable Independent Variables
Coefficients (ȕ)
t
P value
Disability Prevalence
0.143
0.53
0.597
ADI
0.005
0.19
0.851
Public coverage
0.103
0.84
0.408
Population (in thousands)
0.048
19.56
<.001
Adjusted R2 .799
Table 2 presents the OLS regression results with “number of hospitals” as the dependent variable. The model had an adjusted R2 of .799, or about 80%, indicating that using these independent variables, the model is explaining roughly 4/5 of the variations in hospital availability. Disability prevalence, ADI and percentage of public-only health insurance coverage did not significantly predict the number of hospitals in the county. This was likely due to the limitation of county-level data. On the other hand, population was highly predictive of number of hospitals, which can be explained as New England has a wide range of urbanity: some of the most concentrated hospital zones in the country are located in densely populated areas (e.g. Boston, Massachusetts), whereas fewer hospitals are available in sparsely populated, rural areas (e.g. North Vermont). Table 3. GWR Regression: Number of Hospitals as the Dependent Variable* Value R2
.9820
Adjusted R2
.9800
Bandwidth
528292
Residual Squares
890014
Effective Number
7.498
Sigma
125.51
AICc
809.29
Number of Neighbours
30
*Independent Variables: Disability Prevalence, ADI, Public Coverage, and County Population
Table 3 presents the GWR regression results, again, with “number of hospitals” as the dependent variable. In this model, an adjusted R2 of .982 was achieved, greatly superior to the OLS model (R2= .799). This suggests that various determinants of hospital availability determinants and resources may have varied across geographical and administrative borders and the assumption that random effects are spatially stationary should be rejected. The GWR
24
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
analysis found, using the AICc method, that 30 neighbors were sufficient to calibrate each local regression equation for optimal results. The AICc, residual squares, effective number and sigma values are also reported (Table 3). Overall, the GWR model showed a better goodness-of-fit than the OLS model. 5. Conclusions In order to build healthy cities, an understanding of the determinants of health is required; by the same token, including and understanding geographic factors is also a critical part of today’s public health research19. In order to be effective, researchers and policy makers must recognize that health determinants and resources likely vary across geographical and administrative borders and take actions to address this variation. Modeling the factors that contribute to health-related factors based on geography enables policy makers to predict their potential impact in the most affected locations and allocate resources accordingly. This report highlighted the unique regional contexts governing health care resources in the New England region of United States, and underscored the unique circumstances faced by people with disabilities living in deprived areas seeking health care, with many of whom only having public health insurance. To address the lack of hospitals in needed areas, it is important to advocate for new policies that increase the number of community health centers and provide means of transportation such as paratransit, public transit and non-emergency medical transportation to and from hospitals. Public health studies using geographical information hold great promise as they will allow stakeholders not only understand current demand, but also anticipate future demand and equip tools to assess the effectiveness of implemented remediation. Acknowledgements The author acknowledges funding support from National Institute on Disability, Independent Living, and Rehabilitation Research grant 90AR5024-01-00, Dr. Susan Parish (PI); and 90RT5020, Dr. Tamar Heller (PI). References 1. Koh HK, Sebelius KG. Promoting prevention through the affordable care act. New England Journal of Medicine. 2010;363(14):1296-9. 2. Drainoni M-L, Lee-Hood E, Tobias C, Bachman SS, Andrew J, Maisels L. Cross-disability experiences of barriers to health-care access consumer perspectives. Journal of Disability Policy Studies. 2006;17(2):101-15. 3. Ali A, Scior K, Ratti V, Strydom A, King M, Hassiotis A. Discrimination and Other Barriers to Accessing Health Care: Perspectives of Patients with Mild and Moderate Intellectual Disability and Their Carers. PLoS ONE. 2013;8(8):e70855. 4. Said F, Musaddiq T, Mahmud M. Macro level Determinants of Poverty: Investigation Through Poverty Mapping of Districts of Pakistan2011. 895-911 p. 5. Ngui EM, Flores G. Satisfaction with care and ease of using health care services among parents of children with special health care needs: the roles of race/ethnicity, insurance, language, and adequacy of family-centered care. Pediatrics. 2006;117(4):1184-96. 6. Ford CA, Bearman PS, Moody J. Foregone health care among adolescents. JAMA. 1999;282(23):2227-34. 7. 2010-2014 American Community Survey 5-Year Estimates [Internet]. [cited 3/6/2016]. Available from: http://factfinder2.census.gov. 8. Bump SM. Review of Non-Emergency Ambulance Transportation Boston, MA: Office of the State Auditor, 2015 2014-1374-3M1 9. Mitra M, Lifford CJ, Smith LD, Landers B, Tanenhaus R, May GS. Health Needs Assessment of People with Disabilities in Massachusetts, 2013. Worcester, MA: University of Massachusetts Medical School, 2013. 10. Gresenz CR. Dimensions of the Local Health Care Environment and Use of Care by Uninsured Children in Rural and Urban Areas. Pediatrics. 2006;117(3):e509-e17. 11. Jhanjee I, Saxeena D, Arora J, Gjerdingen DK. Parents' Health and Demographic Characteristics Predict Noncompliance with Well-Child Visits. The Journal of the American Board of Family Medicine. 2004;17(5):324-31. 12. Simeonsson RJ, McMillen JS, Huntington GS. Secondary conditions in children with disabilities: spina bifida as a case example. Mental retardation and developmental disabilities research reviews. 2002;8(3):198-205. 13. Foster HE, Eltringham MS, Kay LJ, Friswell M, Abinun M, Myers A. Delay in access to appropriate care for children presenting with musculoskeletal symptoms and ultimately diagnosed with juvenile idiopathic arthritis. Arthritis Care & Research. 2007;57(6):921-7. 14. Kind AJ, Jencks S, Brock J, Yu M, Bartels C, Ehlenbach W, et al. Neighborhood socioeconomic disadvantage and 30-day rehospitalization: a retrospective cohort study. Annals of internal medicine. 2014;161(11):765-74. 15. Singh GK. Area deprivation and widening inequalities in US mortality, 1969-1998. Am J Public Health. 2003;93(7):1137-43. 16. U.S. Geological Survey. Geographic Names Information System. Reston, VA: U.S. Department of the Interior U.S. Geological Survey, 2016. 17. ESRI. ArcGIS Pro: Release 1.2. Redlands, CA: Environmental Systems Research Institute; 2016. 18. Compton WM, Thomas YF, Stinson FS, Grant BF. Prevalence, correlates, disability, and comorbidity of DSM-IV drug abuse and
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
dependence in the United States: results from the national epidemiologic survey on alcohol and related conditions. Archives of general psychiatry. 2007;64(5):566-76. 19. Cromley EK, McLafferty SL. GIS and public health. New York, NY: Guilford Press; 2011.
25
ScienceDirect Procedia Environmental Sciences 36 (2016) 20 – 25
International Conference on Geographies of Health and Living in Cities: Making Cities Healthy for All, Healthy Cities 2016
Disability, Insurance Coverage, Area Deprivation and Health Care: Using Spatial Analysis to Inform Policy Decisions Henan Lia,* a
Brandeis University, 415 South Street MS 035, Waltham, MA 02453, USA
Abstract In the United States, people with disabilities with public insurance are often unable to find suitable health care providers nearby, resulting in needing to travel long distances to large health centres to access necessary health care. This barrier is even more pronounced to those with disabilities living in deprived areas. Using Census data and Area Deprivation Index (ADI) developed by University of Wisconsin-Madison’s ADI project, OLS and Geographically Weighted Regression (GWR) regression analyses showed that high area deprivation index (ADI) and high percentage of public-only health insurance coverage predict disability prevalence. While ADI, public-only insurance coverage percentage and disability prevalence do not significantly predict the number of hospitals, they slightly improved the fit of the GWR model, which was significantly predictive. Future research should continue exploring related factors, and address the challenges of having limited access to hospitals in poorer suburban and rural areas faced by people with disabilities, who are also more likely to have public-only health insurance. © by Elsevier B.V. This is an openB.V. access article under the CC BY-NC-ND license ©2016 2016Published The Authors. Published by Elsevier (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of Healthy Cities 2016. Peer-review under responsibility of the organizing committee of Healthy Cities 2016 Keywords: Disability; Health insurance; Area deprivation, Health care, Geographic information system (GIS)
1. Introduction Access to health care is a topic of great importance to many socially-disadvantaged groups in the United States. The Affordable Care Act (ACA) mandates that all Americans have health insurance and that health care be accessible to all, regardless of geographic location, socioeconomic status, or disability1. However, for those with disabilities
* Corresponding author. Tel.: +1-781-736-3858; fax: +1-781-736-3773. E-mail address: [email protected]
1878-0296 © 2016 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of Healthy Cities 2016 doi:10.1016/j.proenv.2016.09.004
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
living in deprived areas, considerable barriers continue to exist2, 3. It was hypothesized that being in deprived areas and having only public health insurance (Medicare, Medicaid etc.) coverage contributed significantly to the access barriers. Economically deprived areas typically have fewer hospitals, resulting in poorer health care access for local residents4. Similarly, people with disabilities with public insurance are often unable to find suitable health care providers nearby, resulting in needing to periodically travel long distances to large hospitals to access necessary routine or specialty care5, 6. The aim of this present study is to explore the predicting effects of disability prevalence, public-only insurance coverage, and area deprivation index (ADI) on the access to hospitals in the New England region of the U.S. This study examines the New England region, which is an area with a total population of 11,477,279 (2014 data), spanning six Northeastern U.S. states: Connecticut (CT), Maine (ME), Massachusetts (MA), New Hampshire (NH), Rhode Island (RI) and Vermont (VT)7. Despite having several highly-ranked health care systems in the area, many residents with disabilities still experience a great deal of difficulties accessing care due to factors such as lack of reliable transportation and traffic congestions8. For instance, a recent report showed that about 54% of Massachusetts residents with disabilities and their family members or caregivers indicated that transportation to health care was very problematic 9. Respondents reported they cannot access needed medical care when affordable, reliable transportation is unavailable or inaccessible9. Being far away from care providers and subsequently lacking timely access to health care can translate into delayed or foregone primary, preventive and specialist care10, 11. Delayed or foregone care are associated with multiple negative outcomes for people with disabilities, such as development of preventable secondary conditions12, prolonged untreated active diseases13 and higher risks of physical and mental health problems6. Studying the geographical aspects of existing systems is of critical significance for the New England disability community9. 2. Data The data analyzed in this study were 2010-2014 U.S. Census/American Community Survey (ACS) 5-year estimate data7, as well as Area Deprivation Index data, from University of Wisconsin-Madison’s HIPxChange program14, 15. Disability is defined as having limitations in vision, hearing, cognitive, ambulatory, self-care, or independent living7. Area deprivation is defined as the socioeconomic deprivation experienced by a neighborhood, which takes into accounts factors such as median family income, home/rent value, education/profession/employment, and percent of households with access to motor vehicles. Public insurance coverage is defined as only having Medicare or Medicaid. Names and locations of hospitals located in the six states were provided through the Geographic Names Information System (GNIS)16. The analytical framework is presented below (Fig.1).
Fig. 1. Analytical Framework
21
22
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
3. Methods Maps were generated to present a descriptive overview of disability prevalence, ADI, public-only insurance coverage, number of hospitals on the county level. Ordinary Least Squares (OLS) regressions and Geographically weighted regressions (GWR) were conducted to examine the relationship between independent and dependent variables. The fixed kernel type and Akaike Information Criterion (AICc) was used in the GWR analyses. Esri ArcGIS Pro version 1.2 was used to manage and analyse maps and data17. 4. Results
Fig. 2. Disability Prevalence, Area Deprivation Index, Percent of population with Public Insurance only and Number of Hospitals
Descrptive results were presented in Fig. 2. The value categories of Maps 1, 3 and 4 were generated by quantiles while ADI values (Map 2) were broken into 10 categories by natural breaks. Natural break was chosen because it would group similar values together and maximize the differences between categories.
23
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
In the top half of Fig. 2, geographical patterns of associations can be observed between high disability prevalance and high ADI. The bottom half of Fig. 2 showed that relatively fewer hospitals are available where high percentages of the population only have public insurance. Overall, Fig.2 demonstrated the unique circumstances faced by people with disabilities living in deprived areas seeking health care, who only have public health insurance. Table 1. OLS Regression: Disability Prevalence as the Dependent Variable Independent Variables
Coefficients (ȕ)
t
P value
ADI
0.067
6.69
<.001
Public coverage
0.221
4.21
<.001
Adjusted R2 .641
Table 1 presents the OLS regression results with disability prevalence as the dependent variable. Area deprivation index and percentage of public-only health insurance coverage predict disability prevalence in the same direction (e.g. more deprived areas also have higher prevalance of disability). This outcome supports the findings of previous studies18 that demonstrated the connections between poverty, public insurance and disability. Table 2. OLS Regression: Number of Hospitals as the Dependent Variable Independent Variables
Coefficients (ȕ)
t
P value
Disability Prevalence
0.143
0.53
0.597
ADI
0.005
0.19
0.851
Public coverage
0.103
0.84
0.408
Population (in thousands)
0.048
19.56
<.001
Adjusted R2 .799
Table 2 presents the OLS regression results with “number of hospitals” as the dependent variable. The model had an adjusted R2 of .799, or about 80%, indicating that using these independent variables, the model is explaining roughly 4/5 of the variations in hospital availability. Disability prevalence, ADI and percentage of public-only health insurance coverage did not significantly predict the number of hospitals in the county. This was likely due to the limitation of county-level data. On the other hand, population was highly predictive of number of hospitals, which can be explained as New England has a wide range of urbanity: some of the most concentrated hospital zones in the country are located in densely populated areas (e.g. Boston, Massachusetts), whereas fewer hospitals are available in sparsely populated, rural areas (e.g. North Vermont). Table 3. GWR Regression: Number of Hospitals as the Dependent Variable* Value R2
.9820
Adjusted R2
.9800
Bandwidth
528292
Residual Squares
890014
Effective Number
7.498
Sigma
125.51
AICc
809.29
Number of Neighbours
30
*Independent Variables: Disability Prevalence, ADI, Public Coverage, and County Population
Table 3 presents the GWR regression results, again, with “number of hospitals” as the dependent variable. In this model, an adjusted R2 of .982 was achieved, greatly superior to the OLS model (R2= .799). This suggests that various determinants of hospital availability determinants and resources may have varied across geographical and administrative borders and the assumption that random effects are spatially stationary should be rejected. The GWR
24
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
analysis found, using the AICc method, that 30 neighbors were sufficient to calibrate each local regression equation for optimal results. The AICc, residual squares, effective number and sigma values are also reported (Table 3). Overall, the GWR model showed a better goodness-of-fit than the OLS model. 5. Conclusions In order to build healthy cities, an understanding of the determinants of health is required; by the same token, including and understanding geographic factors is also a critical part of today’s public health research19. In order to be effective, researchers and policy makers must recognize that health determinants and resources likely vary across geographical and administrative borders and take actions to address this variation. Modeling the factors that contribute to health-related factors based on geography enables policy makers to predict their potential impact in the most affected locations and allocate resources accordingly. This report highlighted the unique regional contexts governing health care resources in the New England region of United States, and underscored the unique circumstances faced by people with disabilities living in deprived areas seeking health care, with many of whom only having public health insurance. To address the lack of hospitals in needed areas, it is important to advocate for new policies that increase the number of community health centers and provide means of transportation such as paratransit, public transit and non-emergency medical transportation to and from hospitals. Public health studies using geographical information hold great promise as they will allow stakeholders not only understand current demand, but also anticipate future demand and equip tools to assess the effectiveness of implemented remediation. Acknowledgements The author acknowledges funding support from National Institute on Disability, Independent Living, and Rehabilitation Research grant 90AR5024-01-00, Dr. Susan Parish (PI); and 90RT5020, Dr. Tamar Heller (PI). References 1. Koh HK, Sebelius KG. Promoting prevention through the affordable care act. New England Journal of Medicine. 2010;363(14):1296-9. 2. Drainoni M-L, Lee-Hood E, Tobias C, Bachman SS, Andrew J, Maisels L. Cross-disability experiences of barriers to health-care access consumer perspectives. Journal of Disability Policy Studies. 2006;17(2):101-15. 3. Ali A, Scior K, Ratti V, Strydom A, King M, Hassiotis A. Discrimination and Other Barriers to Accessing Health Care: Perspectives of Patients with Mild and Moderate Intellectual Disability and Their Carers. PLoS ONE. 2013;8(8):e70855. 4. Said F, Musaddiq T, Mahmud M. Macro level Determinants of Poverty: Investigation Through Poverty Mapping of Districts of Pakistan2011. 895-911 p. 5. Ngui EM, Flores G. Satisfaction with care and ease of using health care services among parents of children with special health care needs: the roles of race/ethnicity, insurance, language, and adequacy of family-centered care. Pediatrics. 2006;117(4):1184-96. 6. Ford CA, Bearman PS, Moody J. Foregone health care among adolescents. JAMA. 1999;282(23):2227-34. 7. 2010-2014 American Community Survey 5-Year Estimates [Internet]. [cited 3/6/2016]. Available from: http://factfinder2.census.gov. 8. Bump SM. Review of Non-Emergency Ambulance Transportation Boston, MA: Office of the State Auditor, 2015 2014-1374-3M1 9. Mitra M, Lifford CJ, Smith LD, Landers B, Tanenhaus R, May GS. Health Needs Assessment of People with Disabilities in Massachusetts, 2013. Worcester, MA: University of Massachusetts Medical School, 2013. 10. Gresenz CR. Dimensions of the Local Health Care Environment and Use of Care by Uninsured Children in Rural and Urban Areas. Pediatrics. 2006;117(3):e509-e17. 11. Jhanjee I, Saxeena D, Arora J, Gjerdingen DK. Parents' Health and Demographic Characteristics Predict Noncompliance with Well-Child Visits. The Journal of the American Board of Family Medicine. 2004;17(5):324-31. 12. Simeonsson RJ, McMillen JS, Huntington GS. Secondary conditions in children with disabilities: spina bifida as a case example. Mental retardation and developmental disabilities research reviews. 2002;8(3):198-205. 13. Foster HE, Eltringham MS, Kay LJ, Friswell M, Abinun M, Myers A. Delay in access to appropriate care for children presenting with musculoskeletal symptoms and ultimately diagnosed with juvenile idiopathic arthritis. Arthritis Care & Research. 2007;57(6):921-7. 14. Kind AJ, Jencks S, Brock J, Yu M, Bartels C, Ehlenbach W, et al. Neighborhood socioeconomic disadvantage and 30-day rehospitalization: a retrospective cohort study. Annals of internal medicine. 2014;161(11):765-74. 15. Singh GK. Area deprivation and widening inequalities in US mortality, 1969-1998. Am J Public Health. 2003;93(7):1137-43. 16. U.S. Geological Survey. Geographic Names Information System. Reston, VA: U.S. Department of the Interior U.S. Geological Survey, 2016. 17. ESRI. ArcGIS Pro: Release 1.2. Redlands, CA: Environmental Systems Research Institute; 2016. 18. Compton WM, Thomas YF, Stinson FS, Grant BF. Prevalence, correlates, disability, and comorbidity of DSM-IV drug abuse and
Henan Li / Procedia Environmental Sciences 36 (2016) 20 – 25
dependence in the United States: results from the national epidemiologic survey on alcohol and related conditions. Archives of general psychiatry. 2007;64(5):566-76. 19. Cromley EK, McLafferty SL. GIS and public health. New York, NY: Guilford Press; 2011.
25