Advertisement

Access to health care services among people with chronic or disabling conditions: patterns and predictors 1

      Abstract

      Beatty PW, Hagglund KJ, Neri MT, Dhont KR, Clark MJ, Hilton SA. Access to health care services among people with chronic or disabling conditions: patterns and predictors. Arch Phys Med Rehabil 2003;84:1417–25.

      Objective

      To examine patterns of access to a variety of specific health care services among people with chronic or disabling conditions, focusing on factors that predict access to services.

      Design

      National survey of 800 adults with cerebral palsy (CP), multiple sclerosis (MS), spinal cord injury (SCI), or arthritis.

      Setting

      Respondents were surveyed in the general community.

      Participants

      National convenience sample of adults with CP, MS, SCI, or arthritis.

      Interventions

      Not applicable.

      Main outcomes measures

      Access to services from primary care doctors, services from specialists, rehabilitative services, assistive equipment, and prescription medications. Cross tabulations and logistic regression analyses were performed on survey data to examine patterns and predictors of access to health care services.

      Results

      Only half of all respondents received needed rehabilitative services. Respondents covered by fee-for-service health plans were more likely than those covered by managed care organizations to receive needed services from specialists. Respondents with the poorest health and with the lowest incomes were the least likely to receive all health services examined.

      Conclusions

      People with chronic or disabling conditions often require a comprehensive array of health care services. Reform of the current health care payment and delivery structures is needed so that health care is more responsive to those with the greatest service needs.

      Keywords

      ACCESS TO TIMELY AND APPROPRIATE health care is vital to the health and well-being of people with chronic or disabling conditions. Both individuals with disabilities and those with chronic conditions use a disproportionate amount of the total American health care dollar
      • DeJong G.
      • Basnett I.
      Disability and health policy the role of markets in the delivery of health services.
      and have been classified as “vulnerable populations.”
      • Richardson M.
      The Health Security Act of 1993 and health care for persons with disabilities.
      Both of these groups are more likely to have problems accessing health services than those without a chronic illness or disability.
      • Gold M.
      • Nelson L.
      • Brown R.
      • et al.
      Disabled Medicare beneficiaries in HMOs.
      Lack of timely access to health care services can be particularly problematic for people with chronic and/or disabling conditions, who are often at risk of deteriorating health, secondary conditions, secondary functional losses, and decreased independence.
      • Sutton J.
      • DeJong G.
      Managed care and people with disabilities framing the issues.
      ,
      • DeJong G.
      • Palsbo S.
      • Beatty P.
      • Jones G.
      • Kroll T.
      • Neri M.
      The organization and financing of health services for persons with disabilities.
      ,
      • Scheer J.
      • Kroll T.
      • Neri M.
      • Beatty P.
      Access barriers for persons with disabilities the consumers’ perspective.
      ,
      • Neri M.T.
      • Kroll T.
      Understanding the consequences of access barriers to health care experiences of adults with disabilities.
      This study examined access to health care services that are important to people with chronic or disabling conditions. Particular attention was paid to whether access to specific services is associated with health plan type.
      With the increased prevalence of managed care organizations (MCOs), there have been changes in the cost, availability, and organization of health care, all of which have an effect on access to services for vulnerable populations.
      • Miller R.H.
      Healthcare organizational change implications for access to care and its measurement.
      These changes and the increasingly complex system of rules and regulations in managed care plans have further complicated health care service delivery and access to health care services,
      • Gold M.
      Beyond coverage and supply measuring access to healthcare in today’s market.
      especially for people with chronic or disabling conditions. Although access to health care has always been a research and policy priority, the growing number of enrollees in public and private managed care plans is placing a renewed focus on access issues. People with chronic or disabling conditions are a bellwether group, providing a good test of the ability of all health plans to meet the present and potential health care needs of the population of those with greater than average health care needs.
      Regardless of health plan type, navigating the health care system can be problematic when consumers are experiencing severe or disabling pain or are unable to perform basic activities of daily living (ADLs). The complexity of referral processes in MCOs, and claims and reimbursement processes in fee-for-service (FFS) plans, could pose additional barriers for people with chronic or disabling conditions as they seek timely access to specialty care and other “downstream” health services, such as rehabilitation and durable medical equipment (DME).
      • Sutton J.
      • DeJong G.
      Managed care and people with disabilities framing the issues.
      A number of studies highlight the related difficulties of health plan coverage and affordability of services faced by people with disabilities.
      • Beatty P.W.
      • Dhont K.R.
      Medicare health maintenance organizations and traditional coverage perceptions of health care among beneficiaries with disabilities.
      ,
      • DeJong G.
      • Jones G.C.
      • Beatty P.W.
      ,
      The White House National Economic Council and the Domestic Policy Council
      ,
      • Reschovsky J.D.
      Do HMOs make a difference? Access to health care.
      ,
      • Marge M.
      Health promotion for people with disabilities moving beyond rehabilitation.
      ,
      • Sisk J.E.
      • Gorman S.A.
      • Reisinger A.L.
      • et al.
      Evaluation of Medicaid managed care satisfaction, access, and use.
      ,
      • Mark T.
      • Mueller C.
      Access to care in HMOs and traditional insurance plans.
      ,
      • Kassirer J.P.
      Access to specialty care.
      ,
      • O’Day B.L.
      • Corcoran P.J.
      Assistive technology problems and policy alternatives.
      Working-age beneficiaries with disabilities in Medicare managed care plans have also experienced greater difficulty obtaining access to care relative to similarly covered elderly beneficiaries.
      • Gold M.
      • Nelson L.
      • Brown R.
      • et al.
      Disabled Medicare beneficiaries in HMOs.
      Few studies have examined differences in access to services by health plan type among people with chronic or disabling conditions, and none addresses the full spectrum of health care services.
      The scant research on people with chronic or disabling conditions covered by managed care and FFS plans indicates either that access to primary care physicians (PCPs) does not differ by plan type
      • Beinecke R.
      • Pfeifer R.
      • Pfeiffer D.
      • Soussou N.
      The evaluation of fee for service and managed care from the viewpoint of people with disabilities in the USA.
      or that access to PCPs is greater for those in managed care plans.
      • Wholey D.R.
      • Burns L.R.
      • Lavizzo-Mourey R.
      Managed care and the delivery of primary care to the elderly and the chronically ill.
      Access to specialists has been more difficult for people with chronic conditions enrolled in MCOs relative to those in FFS plans.
      • Wholey D.R.
      • Burns L.R.
      • Lavizzo-Mourey R.
      Managed care and the delivery of primary care to the elderly and the chronically ill.
      In terms of DME and technology, anecdotal evidence and accounts of personal experiences with managed care denials abound.
      • Batavia A.I.
      Of wheelchairs and managed care.
      Beyond anecdotal evidence, little or no systematic research has been conducted to examine the extent to which access to obtaining assistive equipment, prescription drugs, and rehabilitative services may differ between people with chronic or disabling conditions in managed care and FFS plans.
      Our study examined access to health care services for people with disabilities or chronic conditions in managed care and traditional FFS plans by answering the following research questions: (1) What percentage of people with chronic and/or disabling conditions report a need for common health care services, such as visits to PCPs and specialists, physical rehabilitation, assistive equipment, and prescription medications? (2) Of those reporting a need for each health service, what percentage receive it? (3) Does health plan type (managed care vs FFS) affect access to a variety of health care services among people with chronic and/or disabling conditions? and (4) What sociodemographic, health, and disability-related variables are associated with access to a variety of health care services among people with chronic and/or disabling conditions?

      Methods

      Data source

      Sampling

      Data for this research came from the first year of a 3-year longitudinal national survey of adults with the following conditions: cerebral palsy (CP), multiple sclerosis (MS), spinal cord injury (SCI), or arthritis (rheumatoid or osteoarthritis). This survey was performed with funding from the National Institute on Disability and Rehabilitation Research as a joint project of the Rehabilitation Research and Training Center on Managed Care and Disability, the Missouri Arthritis Rehabilitation Research and Training Center, and the Rehabilitation Research and Training Center on Health and Wellness.
      Random sampling from the mailing lists of national membership organizations for individuals with CP, MS, SCI, or arthritis was our primary means of recruitment. Survey recruitment packets were mailed to 12,000 individuals from these mailing lists. Through this recruitment method, 571 individuals consented to participate in our survey—for a recruitment rate of less than 5%. In the end, more than half (51%) of the sample for this survey was recruited through direct mailings to randomly selected individuals from the mailing lists of national organizations for individuals with CP, MS, SCI, or arthritis.
      Because our random sampling techniques generated a relatively small number of survey recruits, additional means of drawing a convenience sample were used. In addition to the 51% of the sample that was recruited through national membership organizations, 15% responded to notices in national or local publications for people with these conditions. Fourteen percent of the study sample responded to e-mail notices on disability-related listserv groups or notices on the authors’ institutional Web sites. Referrals from clinical centers or physicians specializing in the target conditions accounted for 7% of study recruits, and an additional 4% were recruited from national and local educational groups associated with the target conditions. Fliers posted in libraries and independent living centers across the United States accounted for 3% of study recruits. The remaining 6% of the sample recruits were introduced to the study via word of mouth from a variety of the sources detailed previously.
      Because both random sampling and convenience sampling methods were used, this survey is not representative of all people with CP, MS, SCI, or arthritis. Population-based surveys of people with low-prevalence conditions are expensive and difficult to conduct. There are no viable national sampling frames for the populations of people with CP, MS, SCI, or arthritis. Randomly calling and screening members of US households until a sufficient sample of people with these conditions is identified is cost prohibitive. The recruitment methods we used may have generated a sample of individuals with more intense interests or difficulties in the US health care system. The sample is heterogeneous in medical symptoms, disability severity, and health care needs. It is homogenous in that people with the targeted conditions are all at increased risk for the development of costly and potentially life-threatening secondary conditions.
      This survey provides a more thorough and detailed picture of the health care experience of people with chronic or disabling conditions than is available in the more generally focused nationally representative data sources. Recruitment methods were intended to capture a wide variety of experiences by sampling from 4 impairment groups with different etiologies, onset ages, and symptoms. The survey was designed to capture the geographic diversity of the United States by recruiting respondents from all 4 US census regions (table 1).
      Table 1Sociodemographic, Health Status, and Disability Severity
      Disability severity is measured by a 3-category variable indicating the total number of ADLs for which respondents require assistance (0, 1–3, 4–6).
      Variables, by Health Plan Type
      VariablenHealth Plan Type
      FFSManaged CareP
      Total800372 (46.5%)428 (53.5%)
      Average age (y)52.02, σ=14.1648.35, σ=13.06.046
      Health status
      Excellent/very good26432.2%33.6%.844
      Good31240.1%38.1%
      Fair/poor22427.7%28.3%
      Disability severity
      0 ADLs47954.8%64.3%.025
      1–3 ADLs14920.7%16.8%
      4–6 ADLs17224.5%18.9%
      Condition type
      Arthritis35743.5%45.6%.174
      CP11013.2%14.3%
      MS16418.8%22.0%
      SCI16924.5%18.1%
      Gender
      Female55166.4%71.0%.091
      Male24933.6%29.0%
      Payer source
      Private45741.9%70.3%<.001
      Medicare27749.7%21.5%
      Medicaid668.4%8.2%
      Income
      <$20,00024935.8%27.1%.027
      $20,000–$40,00021827.7%26.9%
      $40,001–$60,00016918.2%23.6%
      >$60,00016418.3%22.4%
      Region
      Northeast15620.7%18.3%.000
      South22331.7%24.6%
      Midwest22530.4%26.2%
      West19617.2%30.9%
      Disability severity is measured by a 3-category variable indicating the total number of ADLs for which respondents require assistance (0, 1–3, 4–6).

      Data collection

      Data for this research were collected between January and September of 1999 via paper-and-pencil questionnaires delivered through the mail. The 20-page questionnaire consisted of 80 questions about respondents’ access to, use of, and satisfaction with health care services. A total of 1114 surveys were mailed during this time period; replies were received from 942 individuals (84.6%). The base sample size for this research was 800. One hundred forty-two individuals were excluded from analysis (1) because they were younger than 18 years of age, (2) because they reported being uninsured at the time they filled out the questionnaire, or (3) because their surveys contained missing data on any of the key variables examined here. The project team administered the questionnaire over the telephone with the 5% of study participants who requested assistance filling out the questionnaire.

      Variables

      Dependent variables

      A series of indicators, based on Aday’s conceptualization of access as the use of services relative to need,
      • Aday L.
      Economic and noneconomic barriers to the use of needed medical services.
      were used as dependent variables. This study focused on the need for and receipt of (1) services from one’s PCP in the last 6 months, (2) services from one’s specialist(s) in the last 6 months, (3) assistive equipment (eg, mobility aids, orthotic devices, prosthetics, communication devices, home equipment or modifications) in the last 12 months, (4) physical rehabilitation services (physical therapy, occupational therapy, speech therapy) in the last 3 months, and (5) prescription medications in the last 12 months. For each of these 5 health service delivery areas, survey respondents were asked about their need for and use of that service during the indicated time span.
      For each health service delivery area, respondents were asked to check from among the following responses: (1) I did not need X service, (2) I received X service every time I needed it, and (3) I did not receive X service every time I needed it. For respondents reporting that they did not need a particular health service (response 1), access to that particular service was not relevant. Respondents reporting that they received a service every time they needed it in the given time frame were categorized as having access to that service. Respondents reporting that they did not receive the service every time they needed it were categorized as having a lack of complete access to that service.

      Primary independent variable

      A 2-category indicator of health plan type was used as the primary independent variable in all bivariate and multivariate analyses. Given the following broad descriptions of health plan type in our survey, respondents indicated the type of coverage that best described their primary health insurance plan.

      Fee-for-service

      Traditional health insurance that allows you to choose your provider and the services you receive without limitation. Claims are filed by you or your doctor for health services received, and the insurance company covers the services after you pay a deductible amount.

      Managed care

      An insurance plan in which all or most of the health care you receive is from doctors who are associated with the plan. Some managed care plans allow you to visit doctors not associated with the plan, but this involves an extra cost to you. Generally, you do not have to submit claims for the costs of medical services, and you may be required to visit your PCP before visiting a specialist.
      The health plan type distinction is not straightforward. All health plans exist somewhere on a continuum between traditional FFS plans and health maintenance organizations. In recent years, FFS plans have begun to incorporate managed care characteristics, and many managed care plans have incorporated FFS characteristics by loosening their networks and utilization rules.
      • Kongstvedt P.R.
      Because an average sample of health care consumers would not be able to make such nuanced self-report distinctions, our categorization of health plan types was condensed into the descriptions described previously, and respondents were asked to identify the definition that best matched their plan. There were almost no missing or ambiguous responses to this item, indicating that survey respondents were able to place themselves comfortably into 1 category or the other.

      Secondary independent variables

      In addition to an indicator of health plan type, a general measure of primary payer source was used (1, private; 2, Medicare; 3, Medicaid). Categorical measures were used to measure gender (1, female; 2, male), annual household income level (1, <$20,000; 2, $20,000–$40,000; 3, $40,001–$60,000; 4, >$60,000) and census region (1, West; 2, Midwest; 3, Northeast; 4, South).
      Individuals in managed care plans are likely to be younger
      • Shi L.
      Vulnerable populations and health insurance.
      and in better health, with fewer functional limitations than their counterparts in FFS plans.
      • Riley G.
      • Lubitz J.
      • Rabey E.
      Enrollee health status under Medicare risk contracts an analysis of mortality rates.
      ,
      • Riley G.
      • Tudor C.
      • Chiang Y.P.
      • Ingber M.
      Health status of Medicare enrollees in HMOs and fee-for-service in 1994.
      Because these variables are also likely to affect health care access, their effect was measured and controlled for with a continuous age variable and with categorical measures of health status (1, very good/excellent; 2, good; 3, fair/poor) and disability severity (1, require assistance with 0 ADLs; 2, require assistance with 1–3 ADLs; 3, require assistance with 4–6 ADLs; ADLs include bathing, dressing, eating, toileting, transferring, and getting around inside the home).
      • Katz S.
      • Akpom C.
      A measure of primary sociobiological functions.
      A categorical measure of condition type was also used in our analyses (1, arthritis; 2, CP; 3, MS; 4, SCI).

      Statistical analyses

      To describe the sample and to compare characteristics across health plan type, cross tabulations were performed for each of the secondary independent variables with the primary independent variable. Chi-square tests were completed for each cross tabulation to examine possible differences on study variables between those covered by managed care and FFS plans (table 1). Cross tabulations of the independent variables with the access measure for each service delivery area were performed to examine (1) the extent to which a need for each service existed and (2) the extent to which that need was associated with sociodemographic, health status, and disability severity variables. Chi-square tests were completed to examine the extent to which the need for each service was significantly associated with each of these variables (table 2).
      Table 2Percentage of Respondents Who Needed Service, by Sociodemographic, Health Status, and Disability Severity Variables
      VariablePCP (n=800)Specialist (n=800)Physical Rehabilitation (n=800)Assistive Equipment (n=800)Prescription Medications (n=800)
      Total71.3%71.2%36.3%42.5%94.6%
      Health plan typeP=.24P=.04P=.16P=.14P=.02
      Managed care72.4%73.3%34.6%40.6%92.9%
      FFS69.9%68.6%38.2%44.7%96.5%
      Condition typeP=.004P<.001P=.20P<.001P<.001
      Arthritis77.3%82.3%32.2%25.4%97.4%
      CP70.0%52.7%40.9%60.0%83.7%
      MS67.7%73.2%39.0%40.2%96.3%
      SCI62.7%57.4%39.1%69.2%94.1%
      Health statusP<.001P<.001P<.001P=.003P<.001
      Excellent/very good60.2%58.7%25.0%47.4%88.3%
      Good71.8%71.5%35.3%35.0%97.5%
      Fair/poor83.5%85.2%50.9%47.3%98.2%
      Disability severityP=.10P=.11P<.001P<.001P=.24
      0 ADLs71.2%72.0%36.3%29.8%94.8%
      1–3 ADLs72.7%75.2%47.0%53.7%96.6%
      4–6 ADLs66.3%65.1%54.6%68.0%92.4%
      GenderP=.02P=.001P<.276P<.001P<.001
      Female73.5%74.8%37.0%38.5%96.9%
      Male66.3%63.0%34.6%51.4%89.5%
      Payer sourceP=.95P=.06P<.001P<.001P=.14
      Private70.9%74.2%29.5%37.4%94.6%
      Medicare71.5%68.3%42.3%46.2%96.0%
      Medicaid72.7%62.1%57.5%62.1%89.4%
      IncomeP=.008P=.29P<.001P<.001P=.14
      <$20,00076.9%67.1%51.0%54.6%93.2%
      $20,000–$40,00068.3%71.6%30.7%39.9%93.1%
      $40,001–$60,00071.0%72.2%28.4%35.0%97.6%
      >$60,00064.0%75.6%29.3%35.4%95.7%
      RegionP=.91P=.88P=.07P=.09P=.08
      Northeast69.7%71.0%45.2%38.7%92.3%
      South71.3%72.2%33.1%43.0%97.8%
      Midwest72.9%72.0%35.6%48.9%93.4%
      West70.4%68.9%33.6%37.8%94.9%
      NOTE. Percentages in boldface indicate statistical significance of the χ2 test at the P<.05 level.
      On a service-by-service basis, participants were excluded from both bivariate and multivariate analysis if they reported that they did not need that specific service. The number of valid cases ranged from 290 to 757 for the 5 models (Table 3, Table 4). For respondents reporting a need for each service, that measure of access was cross-tabulated with each of the sociodemographic, health status, and disability severity variables. Chi-square tests were completed to examine the extent to which access to services was associated with the independent variables (table 3).
      Table 3Percentage of Respondents Needing Service Who Received It, by Sociodemographic, Health Status, and Disability Severity Variables
      VariablePCP (n=570)Specialist (n=569)Physical Rehabilitation (n=290)Assistive Equipment (n=340)Prescription Medications (n=757)
      Total76.7%82.1%51.4%69.7%90.4%
      Health plan typeP=.15P=.02P=.45P=.76P=.04
      Managed care74.8%79.0%50.7%69.0%88.4%
      FFS78.8%85.9%52.1%70.5%92.5%
      Condition typeP=.03P=.02P=.12P=.61P=.39
      Arthritis81.2%86.7%58.3%74.7%90.5%
      CP76.6%81.0%55.6%68.2%90.2%
      MS74.8%76.7%46.9%65.2%87.3%
      SCI67.0%75.3%40.9%69.2%93.1%
      Health statusP=.03P=.02P=.002P=.003P=.019
      Excellent/very good76.7%85.8%65.2%80.8%93.6%
      Good81.7%84.8%55.5%65.1%91.1%
      Fair/poor70.6%75.9%39.5%61.3%85.9%
      Disability severityP=.22P=.01P=.02P=.18P=.006
      0 ADLs79.2%85.5%61.1%74.8%92.3%
      1–3 ADLs73.0%80.4%44.3%63.8%83.3%
      4–6 ADLs72.8%73.2%43.6%67.5%91.2%
      GenderP=.26P=.26P=.33P=.50P=.09
      Female77.5%81.3%52.5%68.4%89.3%
      Male74.5%84.1%48.8%71.9%92.8%
      Payer sourceP=.41P=.77P=.45P=.10P=.14
      Private75.3%82.6%51.1%74.9%90.7%
      Medicare79.8%82.0%48.7%63.3%91.4%
      Medicaid72.9%78.0%60.5%68.3%83.1%
      IncomeP=.50P=.01P=.18P=.008P=.02
      <$20,00073.5%75.4%46.5%60.3%85.3%
      $20,000–$40,00080.5%84.6%49.3%75.9%92.1%
      $40,001–$60,00076.7%80.3%54.2%69.5%92.1%
      >$60,00077.1%89.5%64.6%82.8%93.6%
      RegionP=.61P=.01P=.96P=.23P=.56
      Northeast81.5%83.6%52.9%65.0%93.0%
      South74.8%78.3%48.6%69.8%90.8%
      Midwest76.2%89.5%52.5%76.4%88.6%
      West75.4%76.3%51.5%63.5%89.8%
      NOTE. Percentages in boldface indicate statistical significance of the χ2 test at the P<.05 level.
      Table 4Odds Ratios From the Logistic Regression of Health Service Access Measures, on Health Plan Type, Health Status, Disability Severity, and Sociodemographic Variables
      VariablePCP (n=570)Specialist (n=569)Physical Rehabilitation (n=290)Assistive Equipment (n=340)Prescription Medications (n=757)
      Health plan type
      FFS1.0001.0001.0001.0001.000
      Managed care.866.568.898.844.644
      Condition type
      Arthritis1.0001.0001.0001.0001.000
      CP.511.388.430.5161.564
      MS.609.624.550.9151.018
      SCI.703.701.560.8182.017
      Health status
      Very good/excellent1.0001.0001.0001.0001.000
      Good.650.491.298.362.609
      Fair/poor.538.591.545.791.774
      Disability severity
      0 ADLs1.0001.0001.0001.0001.000
      1–3 ADLs.940.574.486.859.990
      4–6 ADLs1.079.683.8421.2721.977
      Gender
      Female1.0001.0001.0001.0001.000
      Male1.0161.5811.0141.2541.279
      Payer source
      Private insurance1.0001.0001.0001.0001.000
      Medicare1.3191.3362.6801.319.977
      Medicaid.9691.0031.6531.528.899
      Income
      <$20,0001.0001.0001.0001.0001.000
      $20,000–$40,0001.2922.7462.6073.0932.816
      $40,001–$60,000.8841.5782.2531.6321.402
      >$60,0001.0491.9851.6182.1071.376
      Age.9871.0041.0111.003.972
      Region
      West1.0001.0001.0001.0001.000
      Midwest1.0241.0221.1771.1691.103
      Northeast.940.391.905.5881.143
      South.727.648.922.987.675
      Hosmer-Lemeshow goodness of fit statistic (χ2, df, P)(6.2, 8, .62)(6.8, 8, .56)(5.2, 8, .74)(7.6, 8, .47)(13.5, 8, .09)
      NOTE. Odds ratios in boldface indicate statistical significance at the P<.05 level.
      A series of full-entry logistic regression analyses were then conducted to measure and control for the potential effects of health plan type, payer source, age, gender, health status, disability severity, condition type, and region on access to specific services. Multivariate logistic regression analysis is the proper method to use when examining the relationship between a 2-category dependent variable and any number of categorical or continuous independent variables.
      • Stokes M.
      • Davis C.
      • Koch G.
      When more than 1 independent variable is entered into a logistic regression model, the effect of each independent variable is provided, net of the effects of the other independent variables in the model. Relationships between independent and dependent variables are characterized in terms of odds ratios.

      Results

      Bivariate results

      Differences and similarities between managed care and FFS enrollees

      There were several demographic differences between participants who were enrolled in FFS and managed care health plans (table 1). Participants enrolled in managed care plans tended to be younger, less likely to have activity limitations, more likely to be covered by a private insurance plan, more likely to be in a higher income bracket, and more likely to live in the western United States than were those covered by FFS plans. Health plan type was not significantly related to health status, condition type, or gender.

      Relationship between individual characteristics and need for services

      Results of bivariate cross tabulations and chi-square tests are depicted in table 2. In these analyses, only need for health care services was examined.
      There was wide variation in the extent of need reported for each of the 5 services examined here. Exactly 71.3% of the sample reported a need for primary care. Similarly, 71.2% of the sample reported a need for specialty care. Just over a third (36.3%) reported a need for physical rehabilitation services. Exactly 42.5% reported a need for assistive equipment, and 94.6% reported a need for prescription medications.
      Health plan type was significantly associated with the need for specialty services and for prescription medications. Study respondents covered by managed care plans were more likely to report a need to see a specialist and were less likely to report a need for prescription medications, relative to those covered by FFS plans.
      Condition type was significantly related to need for 4 of the 5 service areas examined. Study respondents with arthritis were the most likely to report a need for services from PCPs and specialists and for prescription medications. Respondents with SCI were the most likely to report a need for assistive equipment.
      Health status and disability severity were significantly related to the need for many of the health services examined. Not surprisingly, respondents reporting fair or poor health status were the most likely to report a need for all 5 health services. Individuals with 4 or more ADL limitations were the most likely to report a need for rehabilitative services and assistive equipment.
      Women were significantly more likely than men to report a need for services from PCPs and specialists and for prescription medications. Men were more likely to report a need for assistive equipment. Medicaid recipients were the most likely—and people with private insurance were the least likely—to report a need for rehabilitation and assistive equipment.
      Respondents in the lowest income category were the most likely to report a need for PCP, rehabilitation, and assistive equipment services. Respondents in the highest income categories were the most likely to report a need for prescription medications. Region was unrelated to need for all 5 services.

      Multivariate analyses: determinants of access to services

      The results outlined in this section are the products of bivariate cross tabulations and chi-square tests, as well as multivariate logistic regression analyses. Table 3 displays the bivariate results, and table 4 displays the results of the multivariate analyses. The Hosmer-Lemeshow statistics
      • Stokes M.
      • Davis C.
      • Koch G.
      (table 4) indicate that all 5 of the logistic regression models adequately fit the survey data.
      Among those who reported a need, the percentage of respondents who reported consistent access to each specific health care service varied considerably (table 3). Just over three fourths of respondents who reported a need for services from their PCP received those services every time they were needed. Exactly 82.1% of those who reported a need for specialty care received it every time it was needed. Only about half (51.4%) of those who reported a need for physical rehabilitation services received them every time they were needed. Exactly 69.7% of those who needed assistive equipment received it every time it was needed, and 90.4% of those who needed prescription drugs received them every time they were needed.
      Statistically significant predictors of access to a spectrum of health care services are highlighted below.

      Health plan type

      Results of the cross-tabulation analyses indicate that respondents covered by FFS plans were more likely than those in managed care plans to receive needed specialty care and prescription medications. After controlling for condition type, health status, and sociodemographic variables in the model, the relationship between health plan type and access to needed prescription medications disappeared, whereas the relationship between plan type and access to specialty services persisted. The odds of accessing needed specialty care in managed care plans were about half those for FFS plans (table 4).

      Condition type, severity, and health status variables

      According to bivariate analyses, respondents with arthritis were more likely than those with CP, MS, or SCI to receive needed services from their PCPs and specialists. These relationships persisted after controlling for the other variables in the model. In addition, a relationship between condition type and access to rehabilitation emerged in the multivariate model. As shown in table 4, the odds of participants with CP getting needed rehabilitative services were less than those of participants with arthritis.
      According to bivariate analyses, health status was significantly related to access to the full spectrum of services examined here. After controlling for all other variables in the model, this relationship persisted for all access areas except for prescription medications. The odds of respondents in “excellent or very good” health receiving services from PCPs and specialists, as well as rehabilitation and assistive equipment, were more than twice those of respondents reporting poorer health.
      Bivariate analyses indicated that individuals with no ADL limitations were more likely than those with ADL limitations to receive needed specialty services, rehabilitative services, and prescription medications (table 3). This relationship persisted only for access to rehabilitative services in our multivariate models. The odds of respondents with 1 to 3 ADL limitations receiving needed rehabilitative services were less than half those of respondents with no ADL limitations (table 4).

      Sociodemographic variables

      Rates of access to health services examined in this study did not differ significantly by gender. According to the multivariate analyses, the odds of Medicare beneficiaries receiving needed rehabilitative services were more than twice those of beneficiaries covered by private insurance (table 4).
      Survey respondents in the lowest income categories were the least likely to report receipt of needed specialty services, assistive equipment, and prescription medications (table 3). These relationships persisted after controlling for all other independent variables, and a relationship between income level and rehabilitation emerged. The odds of respondents with incomes less than $20,000 receiving needed specialty services, rehabilitation, assistive equipment, and prescription medications were roughly one third those for respondents with higher incomes (table 4).
      Increasing age was associated with a decreased likelihood of receiving prescription medications among our respondents (table 4). The bivariate relationship between region and access to specialty services disappeared after controlling for the other variables in the model. Multivariate analyses indicated that region was related to access to specialists. After controlling for other variables in the model, the odds of respondents in the Northeast receiving specialty services every time they were needed were one third those of respondents in the West (table 4).

      Discussion

      Perhaps the most challenging health care policy issue in the United States is the provision of timely, appropriate, and high-quality health care to its most vulnerable citizens. This study examined variables that may influence access to health care among adults with disabilities or chronic conditions. The results suggest that people with chronic or disabling conditions often require a comprehensive array of health care services but that such care is not consistently available.
      The analyses uncovered the following major findings, which are discussed in the following paragraphs: (1) people with disabilities or chronic conditions covered by MCOs face restricted access to services from specialists, relative to those covered by FFS plans; (2) only half of those who report a need for rehabilitation services receive those services every time they are needed, regardless of health plan type; and (3) people with disabilities or chronic conditions who report being in the poorest health and those reporting the lowest income categories are the least likely to receive a wide variety of needed health services.

      Access to specialists

      The results of this study are consistent with previous work that has shown that people with chronic conditions in MCOs have greater difficulty obtaining specialty care compared with those in FFS plans.
      • Wholey D.R.
      • Burns L.R.
      • Lavizzo-Mourey R.
      Managed care and the delivery of primary care to the elderly and the chronically ill.
      This particular result is disconcerting, because people with chronic conditions or disabilities often require specialty care to maintain health and functional abilities and to avoid costly secondary conditions and hospitalizations.
      • Bockneck W.
      • Mann N.
      • Lanig I.
      • et al.
      Primary care for persons with disability.
      The individuals who participated in this study have conditions that often involve multiple organ systems and require complex treatment. Some patients and physicians have called for specialists to be considered PCPs for people with serious chronic illnesses or disabilities because of the complexities in rendering care.
      • Gabriel S.
      Primary care specialists or generalists.
      Poor access to needed specialty care is linked to the development of preventable secondary conditions that are ultimately more costly to treat.
      • Bockneck W.
      • Mann N.
      • Lanig I.
      • et al.
      Primary care for persons with disability.
      The finding regarding access to specialty care suggests that MCOs may not have a sufficient network of physicians who have experience providing services to people with chronic conditions or disabilities.
      • Huntt D.C.
      • Growick B.S.
      Managed care for people with disabilities.

      Access to rehabilitation

      Rehabilitative services were obtained by only half of the respondents reporting a need. All other services examined in the study were considerably more likely to be obtained. This is a troubling finding given that rehabilitation is often central to the ongoing treatment of the functional sequelae of the conditions examined in this study. Coverage limitations for “maintenance therapy” may partially explain the relatively poor access to rehabilitative services uncovered here.
      • Ireys H.
      • Wehr E.
      • Cooke R.
      Most insurance policies and federal and state regulations require that health care services be restorative, that is, designed to restore an individual to a previous level of health or functioning. This coverage restriction often works well in acute health care settings to hold down costs by restricting payment for unnecessary services. For people with chronic or disabling conditions, however, this practice is unjustified. Maintenance therapy is often critical to preserving overall health and functioning,
      • Marge M.
      Health promotion for people with disabilities moving beyond rehabilitation.
      maintaining independence, avoiding institutionalization, and preventing the development of secondary conditions and the associated need for costly surgeries or hospitalizations.
      • Hjeltnes N.
      • Jensen T.
      Physical endurance capacity, functional status and medical complications in spinal cord injured subjects with long-standing lesions.
      ,
      • Lockette K.
      • Keyes A.
      ,
      • Rimmer J.

      Access inequities

      Income level was consistently and substantially related to difficulty receiving needed care across all services except primary care. The results showed that those individuals whose annual incomes were below $20,000 were the least likely to obtain care every time it was needed. The positive relationship between income and access to services is consistent with previous research involving individuals with and without chronic or disabling conditions.
      • Berk M.L.
      • Schur C.L.
      • Cantor J.C.
      Ability to obtain health care recent estimates from the Robert Wood Johnson Foundation National Access to Care Survey.
      People who have chronic conditions or disabilities, on average, require health care services more frequently, and these services are often costly. Those people with incomes below $20,000 per year are unlikely to be able to pay for needed health care, regardless of health plan type.
      Current health care delivery systems require reform to prioritize access to care for people with the greatest needs. The results of our study consistently showed that there are gaps in the health care delivery system for people with chronic or disabling conditions. In general, a substantial percentage of the sample needed health services but did not receive them. Health status was consistently and strongly related to receipt of needed health care services. People in the poorest health status, who arguably have the greatest health care needs, were the least likely to report receiving services every time they were needed. Of course, those individuals with poor health status may need health services more frequently than those reporting more positive levels of health—increasing the potential for unmet need. For example, a person with MS who needs monthly checkups with her/his neurologist will be more likely than a person needing annual neurologic consultations to report that he/she did not receive services from specialists every time they were needed. Although this increased level of need may help explain the strength of the relationship between health status and access to services, it does not reduce the clinical significance of the finding or the risks associated with reduced access to care among people with relatively poor health status and a thinner margin of health. Unmet health care needs among people with disabilities or chronic conditions result in long-term negative consequences for individuals, including declining physical and mental health and reduced social participation.
      • Neri M.T.
      • Kroll T.
      Understanding the consequences of access barriers to health care experiences of adults with disabilities.
      Access inequities were also found across the 4 condition types examined here. Research into reimbursement patterns for maintenance therapies, as well as the disability knowledge of providers, may prove fruitful in explaining these differences. For example, arthritis treatment options exist that provide for potential improvement in function. Treatments aimed at functional improvement fit well into our health care system because it is oriented strongly toward the treatment of acute conditions. Effective medical interventions for treating the often severe and chronic symptoms of CP, MS, or SCI, however, do not yet exist. Maintenance of current health and function is therefore of primary importance for these groups.
      • Huntt D.C.
      • Growick B.S.
      Managed care for people with disabilities.
      As discussed earlier, these maintenance treatments are often not covered in our delivery systems, which are oriented toward cures and improvement.
      Higher rates of access among people with arthritis may also be due to the fact that arthritis is much more prevalent in the adult population than CP, MS, or SCI. The relatively low prevalence of CP, MS, and SCI is likely associated with a lack of knowledge and experience among health care providers regarding the health and health care needs of people who have them.

      Limitations of the study

      The results of this study should be considered in light of its limitations. As noted, the sample is not a random sample of all people with CP, MS, SCI, or arthritis, and the results should not be generalized to these populations. Heterogeneity of health status among respondents was obtained, but recruitment methods may have biased our sample toward those who have experienced more difficulties in obtaining health care than the broader population of people with disabilities. In addition, the uninsured were not included in this study. It is reasonable to conclude, however, that the uninsured who have disabilities or a chronic condition are likely to have even poorer access to health care services than the people represented in this study.
      Additionally, the complexity of health plan types could not be fully assessed with our 2-category independent variable. There may be differences in access across different types of health plans that were obscured by our managed care-FFS dichotomy. In recent years, health insurance companies have offered an increasing variety of health plan types that combine managed care and FFS characteristics. Future research should address the relationship between coverage by these hybrid health plan types and access to health care services for individuals with disabilities or chronic conditions.

      Conclusion

      Health care reform policies aimed at increasing access to services for people with chronic or disabling conditions will not resolve access problems if they focus solely on tightening the reins on MCOs. Relatively large proportions of people with chronic or disabling conditions are still covered by traditional FFS plans. However, neither FFS nor MCO plans have made meaningful changes to increase access to services for members of this population. The unmet needs identified in this study suggest that proposed legislative remedies—such as a “patients’ bill of rights,” increasing the number of people insured, or regulatory reforms—are, on their own, unlikely to improve access for the populations examined here. Providers, researchers, and policymakers must also look beyond the current health care payment and delivery structures to serve people with chronic or disabling conditions better. Further research into health care delivery and financing for people with disabilities or chronic conditions is needed to provide a foundation on which to build more responsive service delivery models for those with the most substantial health care needs.

      References

        • DeJong G.
        • Basnett I.
        Disability and health policy.
        in: Albrecht G. Seelman K. Bury M. Handbook of disability studies. Sage, Thousand Oaks (CA)2001: 610-632
        • Richardson M.
        The Health Security Act of 1993 and health care for persons with disabilities.
        Health Care Manag. 1994; 1: 145-155
        • Gold M.
        • Nelson L.
        • Brown R.
        • et al.
        Disabled Medicare beneficiaries in HMOs.
        Health Aff (Millwood). 1997; 16: 149-162
        • Sutton J.
        • DeJong G.
        Managed care and people with disabilities.
        Arch Phys Med Rehabil. 1998; 79: 1312-1316
        • DeJong G.
        • Palsbo S.
        • Beatty P.
        • Jones G.
        • Kroll T.
        • Neri M.
        The organization and financing of health services for persons with disabilities.
        Milbank Q. 2002; 80: 261-301
        • Scheer J.
        • Kroll T.
        • Neri M.
        • Beatty P.
        Access barriers for persons with disabilities.
        J Disabil Policy Stud. 2003; 13: 221-230
        • Neri M.T.
        • Kroll T.
        Understanding the consequences of access barriers to health care.
        Disabil Rehabil. 2003; 25: 85-96
        • Miller R.H.
        Healthcare organizational change.
        Health Serv Res. 1998; 33: 653-680
        • Gold M.
        Beyond coverage and supply.
        Health Serv Res. 1998; 33: 625-652
        • Beatty P.W.
        • Dhont K.R.
        Medicare health maintenance organizations and traditional coverage.
        Arch Phys Med Rehabil. 2001; 82: 1009-1017
        • DeJong G.
        • Jones G.C.
        • Beatty P.W.
        Prescription drugs for working-age Medicare beneficiaries with disabilities. Health and disability policy brief. National Rehabilitation Hospital Center for Health and Disability Research, Washington (DC)2000
        • The White House National Economic Council and the Domestic Policy Council
        Executive summary: disability, Medicare, and prescription drugs, July 31, 2000. 2003 (Available at: http://clinton4.nara.gov/textonly/WH/EOP/nec/html/ParkMedicareReport000730.html. Accessed January 8,)
        • Reschovsky J.D.
        Do HMOs make a difference? Access to health care.
        Inquiry. 1999; 36: 390-399
        • Marge M.
        Health promotion for people with disabilities.
        Am J Health Promot. 1998; 2: 29-35
        • Sisk J.E.
        • Gorman S.A.
        • Reisinger A.L.
        • et al.
        Evaluation of Medicaid managed care.
        JAMA. 1996; 276: 50-55
        • Mark T.
        • Mueller C.
        Access to care in HMOs and traditional insurance plans.
        Health Aff (Millwood). 1996; 15: 81-87
        • Kassirer J.P.
        Access to specialty care.
        N Engl J Med. 1994; 331: 1151-1153
        • O’Day B.L.
        • Corcoran P.J.
        Assistive technology.
        Arch Phys Med Rehabil. 1994; 75: 1165-1169
        • Beinecke R.
        • Pfeifer R.
        • Pfeiffer D.
        • Soussou N.
        The evaluation of fee for service and managed care from the viewpoint of people with disabilities in the USA.
        Disabil Rehabil. 1997; 19: 513-522
        • Wholey D.R.
        • Burns L.R.
        • Lavizzo-Mourey R.
        Managed care and the delivery of primary care to the elderly and the chronically ill.
        Health Serv Res. 1998; 33: 322-353
        • Batavia A.I.
        Of wheelchairs and managed care.
        Health Aff (Millwood). 1999; 18: 177-182
        • Aday L.
        Economic and noneconomic barriers to the use of needed medical services.
        Med Care. 1975; 13: 447-456
        • Kongstvedt P.R.
        Essentials of managed health care. 4th ed. Aspen, Gaithersburg (MD)2001
        • Shi L.
        Vulnerable populations and health insurance.
        Med Care Res Rev. 2000; 57: 110-134
        • Riley G.
        • Lubitz J.
        • Rabey E.
        Enrollee health status under Medicare risk contracts.
        Health Serv Res. 1991; 26: 137-163
        • Riley G.
        • Tudor C.
        • Chiang Y.P.
        • Ingber M.
        Health status of Medicare enrollees in HMOs and fee-for-service in 1994.
        Health Care Financ Rev. 1996; 17: 65-76
        • Katz S.
        • Akpom C.
        A measure of primary sociobiological functions.
        Int J Health Sci. 1976; 6: 493-508
        • Stokes M.
        • Davis C.
        • Koch G.
        Categorical data analysis using the SAS system. SAS Institute Inc, Cary (NC)2000
        • Bockneck W.
        • Mann N.
        • Lanig I.
        • et al.
        Primary care for persons with disability.
        in: DeLisa J.A. Gans B.M. Bockenek W.L. Rehabilitation medicine: principles and practice. 3rd ed. Lippincott-Raven, Philadelphia1998: 905-928
        • Gabriel S.
        Primary care.
        Mayo Clin Proc. 1996; 71: 415-419
        • Huntt D.C.
        • Growick B.S.
        Managed care for people with disabilities.
        J Rehabil. 1997; 63: 10-14
        • Ireys H.
        • Wehr E.
        • Cooke R.
        Defining medical necessity: strategies for promoting access to quality care for persons with developmental disabilities, mental retardation, and other special health care needs. National Center for Education in Maternal and Child Health, Washington (DC)1999
        • Hjeltnes N.
        • Jensen T.
        Physical endurance capacity, functional status and medical complications in spinal cord injured subjects with long-standing lesions.
        Paraplegia. 1990; 28: 428-432
        • Lockette K.
        • Keyes A.
        Conditioning with physical disabilities. Human Kinetics, Champaign (IL)1994
        • Rimmer J.
        Fitness and rehabilitation programs for special populations. Human Kinetics, Champaign (IL)1994
        • Berk M.L.
        • Schur C.L.
        • Cantor J.C.
        Ability to obtain health care.
        Health Aff (Millwood). 1995; 14: 139-146