Access to physician care and medications among stroke survivors by age according to the National Health Interview Survey from years 1998 to 2002. The P values are associated with χ2 test for trend.
Levine DA, Kiefe CI, Houston TK, Allison JJ, McCarthy EP, Ayanian JZ. Younger Stroke Survivors Have Reduced Access to Physician Care and MedicationsNational Health Interview Survey From Years 1998 to 2002. Arch Neurol. 2007;64(1):37-42. doi:10.1001/archneur.64.1.noc60002
Copyright 2007 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2007
More than 5 million US stroke survivors require comprehensive care for risk factor modification and secondary prevention. Younger stroke survivors may have reduced access to physician care and medications because they are more frequently uninsured.
To assess age-related differences in access to physician care and medications among stroke survivors (aged 45-64 years vs ≥65 years).
National Health Interview Survey from years 1998 to 2002.
A US population-based survey.
Stroke survivors (n = 3681) aged 45 years and older among 159 985 survey respondents.
Main Outcome Measures
General doctor visit, medical specialist visit, and inability to afford medications within the last 12 months.
Compared with older stroke survivors, younger stroke survivors more frequently reported no general doctor visit (10% vs 14%, respectively; P = .002), no general doctor or medical specialist visit (5% vs 8%, respectively; P = .003), and the inability to afford medications (6% vs 15%, respectively; P<.001). Younger age was independently associated with no general doctor visit (odds ratio, 1.40; 95% confidence interval, 1.04-1.88), no general doctor or medical specialist visit (odds ratio, 1.69; 95% confidence interval, 1.14-2.52), and the inability to afford medications (odds ratio, 2.94; 95% confidence interval, 2.19-3.94) after adjusting for sex, race, income, neurological disability, health status, and comorbidity. With further adjustment for health insurance, younger age remained independently associated with the inability to afford medications but not the lack of physician visits.
Stroke survivors younger than 65 years reported worse access to physician care and medication affordability than older stroke survivors. Inadequate access among younger stroke survivors may lead to inadequate risk factor modification and recurrent cardiovascular events.Published online November 13, 2006 (doi:10.1001/archneur.64.1.noc60002).
The approximately 5.4 million stroke survivors in the United States require comprehensive care for risk factor modification and secondary prevention.1Patients with ischemic stroke have an approximately 5% to 15% annual risk of recurrent stroke.1- 3 Poststroke cardiovascular events increase the morbidity, mortality, and health care costs of this population.2 Although specialist care is important, secondary stroke prevention, particularly hypertension and hyperlipidemia treatment, is usually performed by primary care providers.4 Despite available evidence-based guidelines for secondary stroke prevention, risk factor modification among ischemic stroke survivors is suboptimal.5- 10
Few studies have assessed access to care among stroke survivors. In the United States, stroke survivors younger than 65 years may have reduced access to care because they do not qualify for Medicare health insurance unless they have been receiving Social Security or Railroad Retirement Board disability benefits for 2 or more years,11 and they are more likely to be uninsured.12 Little is known regarding access to care for the US national population of stroke survivors younger than 65 years. This study examines age-related differences in access to care among community-dwelling stroke survivors using data from a nationally representative, population-based survey.
The National Health Interview Survey (NHIS) is a continuing, in-person household survey of the civilian, noninstitutionalized US population conducted annually by the National Center for Health Statistics using face-to-face interviews.13 Data from the NHIS from years 1998 to 2002, which use similar survey designs and data collection methods,13 were combined to maximize statistical power. The final response rates for the sample adult survey from years 1998 to 2002 ranged from 70% to 74%.13 Specific information regarding the NHIS is available elsewhere.13 Stroke survivors were identified as respondents who answered yes to the question, “Have you ever been told by a doctor or other health professional that you had a stroke?” We defined stroke survivors as younger (aged 45-64 years) and older (aged ≥65 years).
Report of a general doctor visit (“During the past 12 months, have you seen or talked to a general doctor who treats a variety of illnesses, ie, a doctor in general practice, family medicine, or internal medicine?”) was the primary outcome measure. Report of a medical specialist visit (“During the past 12 months, have you seen or talked to a medical doctor who specializes in a particular medical disease or problem, ie, other than obstetrician or gynecologist, psychiatrist, or ophthalmologist, about your own health?”) and the inability to afford medications (“During the past 12 months, was there any time when you needed prescription medicines but didn't get [them] because you couldn't afford [them]?”) were secondary outcome measures.
Health care access variables and outcome measures were compared between age groups (ages 45-64 years vs ≥65 years) using χ2 or t test as appropriate, and US population-based estimates were calculated. Covariates were selected using the Andersen Behavioral Model framework,14,15 literature review, and clinical observation. Covariates included race (white, black, other), sex, education, annual household income, health insurance status, lack of transportation delaying care, no usual place of care, neurological disability due to stroke, self-reported health status, and comorbidity using a composite score based on 5 major health conditions (hypertension, coronary heart disease, diabetes mellitus, emphysema, and heart condition or disease). The entire analysis was repeated in the subgroup of stroke survivors with disability. The frequencies of the main outcome measures in the age groups (ages 45-54, 55-64, 65-74, and ≥75 years) were compared using χ2 test for trend.
Multivariable logistic regression analyses were performed to examine the adjusted associations between age group and the 3 outcome measures. Model A was adjusted for all covariates except health insurance. Three covariates, including usual place of care, education, and lack of transportation, were excluded due to a lack of statistical contribution to the multivariable models. Model B was adjusted for the covariates in model A plus health insurance status. We repeated all of the multivariable analyses stratified by age group to examine predictors of access separately for each age group. All of the analyses used SAS-callable SUDAAN software version 7.5 (Research Triangle Institute, Research Triangle Park, NC) to obtain proper variance estimations that accounted for the complex sampling design of the NHIS and results that were weighted to reflect national population estimates.
Appropriate institutional review board approval was obtained from the University of Alabama at Birmingham.
We identified 3681 stroke survivors aged 45 years and older, representing an estimated 4.1 million US stroke survivors, 1.3 million of whom are aged 45 to 64 years. Compared with older stroke survivors (mean ± SD age, 76 ± 0.2 years; n = 2509), younger stroke survivors (mean ± SD age, 56 ± 0.2 years; n = 1172) were more likely to be male (47% vs 52%, respectively; P = .01), be black (10% vs 19%; P<.001), and lack health insurance (0.4% vs 11%, respectively; P<.001) (Table 1). Compared with older stroke survivors, younger stroke survivors more often reported no general doctor visit, no general doctor or medical specialist visit, and the inability to afford medications (Table 2). Both groups similarly reported no medical specialist visit (44%).
Lack of health insurance was associated with reduced access to care on all 3 outcome measures (Table 2). Compared with insured stroke survivors, uninsured stroke survivors more frequently reported no medical specialist visit (42% vs 67%, respectively; P<.001). The percentage of stroke survivors reporting no general doctor visit, no general doctor or medical specialist visit, and the inability to afford medications decreased with increasing age category (Figure).
Among those stroke survivors with disability, younger age as compared with older age was associated with the inability to afford medications (16% vs 5%, respectively; P<.001) and no general doctor visit (12% vs 8%, respectively; P = .08); no age difference was seen in the reporting of no general doctor or medical specialist visit (5%). Stroke survivors with disability had rates of no health insurance by age group that were similar to the rates of the entire stroke cohort (11% for the younger group vs 0.6% for the older group; P<.001).
In unadjusted analyses, younger age was associated with no general doctor visit (odds ratio, 1.50; 95% confidence interval, 1.17-1.91; R2, 0.004), no general doctor or medical specialist visit (odds ratio, 1.69; 95% confidence interval, 1.23-2.33; R2, 0.003), and the inability to afford medications (odds ratio, 2.95; 95% confidence interval, 2.28-3.82; R2, 0.022). After adjustment, younger age remained independently associated with the 3 outcome variables at approximately the same magnitudes (model A, Tables 3, 4, and 5). With further adjustment for health insurance, these odds ratios (95% confidence intervals) became 1.16 (0.83-1.61), 1.21 (0.79-1.86), and 2.56 (1.85-3.55), respectively. Lack of health insurance was associated with reduced access for all 3 outcome measures in the younger age group but was less so in the older group (model B, Tables 3, 4, and 5).
In this nationally representative sample, younger stroke survivors (aged 45-64 years) have reduced access to physician care and medications compared with older stroke survivors (aged ≥65 years). Lack of health insurance explained some of the reduced access to physician care among younger stroke survivors but not their more frequent problems with medication affordability. Clinicians may assume that stroke survivors have health insurance owing to misunderstandings about qualifications for insurance based on disability. Our data show that these assumptions may well be unwarranted for younger stroke survivors. Moreover, most of the uninsured stroke survivors in this study were not born outside the United States; only 16% and 43% of uninsured younger and older stroke survivors, respectively, were foreign born.
Our findings provide population-based estimates of rates of uninsured status, physician visits, and medication affordability for middle-aged and elderly community-dwelling US stroke survivors. Given that secondary stroke prevention is usually provided by primary care providers and requires pharmacological therapy to modify risk factors such as hypertension, a condition reported by 70% of stroke survivors in our study, these results suggest that younger stroke survivors are more likely to have inadequate access to physicians, medications, and possibly secondary stroke prevention. Indeed, lack of physician contact after stroke has been associated with reduced rates of antihypertensive and antiplatelet therapy.16
Our findings are consistent with earlier studies showing inadequate health care access for stroke survivors in the United Kingdom16 and Canada.17 In the Brain Attack Surveillance in Corpus Christi project, younger Mexican-American stroke survivors (mean age, 70 years) were less likely to report having health insurance and a primary care physician compared with non-Hispanic white stroke survivors (mean age, 75 years); however, after adjusting for age and sex, these disparities were attenuated.18 Black race has been associated with less access to carotid endarterectomy and neurologists in patients aged 65 years and older who receive Medicare insurance and have been hospitalized with transient ischemic attack.19 We found that black race was associated only with no medical specialist visit for the total stroke survivor group (data not shown).
Prior studies20- 22 have found an association between medication nonadherence for other conditions, such as hypertension and diabetes, attributable to the lack of prescription coverage. In our study of stroke survivors, younger age was associated with the lack of medication affordability even after adjusting for health insurance status. The cause of a more significant problem affording medication in the younger stroke survivors may be related to competing household costs or a lack of prescription drug coverage, which we could not assess directly in our analysis.
Our study design is observational and we cannot infer causation from the associations that we observed. Residual confounding by the subset of variables in the regression models, uncontrolled confounding, and persisting unadjusted risk between groups may exist. Several factors, such as provider characteristics, stroke type (although most strokes [>88%] are ischemic rather than hemorrhagic),1 stroke severity, and stroke acuity, could not be assessed adequately. Data are self-reported and subject to recall bias and reporting error. Respondents may fail to remember physician visits23,24 or may inaccurately report a history of stroke.25,26 More recent data suggest that self-report of stroke is accurate,27 even in elderly persons with disability.28 Because the NHIS does not sample persons who are institutionalized, this analysis includes community-dwelling stroke survivors and does not include stroke survivors living in long-term care or skilled nursing facilities. Selection bias may occur if stroke survivors with poor access are less likely than stroke survivors without poor access to be sampled given the NHIS study design. However, the calculated number of stroke survivors approximates expected population-based stroke estimates.
Reduced access to care and medications for stroke survivors may lead to inadequate risk factor modification and recurrent cardiovascular events. Cardiovascular risk reduction is reduced in uninsured adults, particularly in long-term uninsured adults.29,30 Uninsured persons also have reduced access to ambulatory medical care31 and, for near-elderly persons, substantially increased adjusted mortality.32 Given the increasing number of uninsured nonelderly Americans12,33 and the high costs associated with recurrent stroke (which are increased compared with first stroke34) and other cardiovascular events, the potential costs of reduced access to health care among younger stroke survivors are substantial. Expanding health insurance by providing affordable Medicare insurance or, more radically, immediate Medicare insurance to uninsured patients with stroke would be expected to improve access to care and increase the use of basic clinical services like physician visits.35 Affordable prescription coverage would be necessary to increase access to medications.22,35
Further research is needed to determine whether this younger high-risk population has adverse outcomes, such as death and cardiovascular events, or has increased long-term health care utilization due to reduced access to physician care and medications. Further work addressing access gaps, linking to related health outcomes and costs, and demonstrating the effectiveness and cost-effectiveness of possible improvement strategies is warranted.
Correspondence: Deborah A. Levine, MD, MPH, Division of General Internal Medicine, University of Alabama at Birmingham, 1530 Third Ave S, FOT 720, Birmingham, AL 35294-3407 (email@example.com).
Accepted for Publication: January 10, 2006.
Published Online: November 13, 2006 (doi:10.1001/archneur.64.1.noc60002).
Author Contributions:Study concept and design: Levine, Kiefe, Houston, Allison, and Ayanian. Acquisition of data: Levine and Ayanian. Analysis and interpretation of data: Levine, Kiefe, Houston, Allison, McCarthy, and Ayanian. Drafting of the manuscript: Levine, Kiefe, Houston, Allison, and Ayanian. Critical revision of the manuscript for important intellectual content: Levine, Kiefe, Houston, Allison, McCarthy, and Ayanian. Statistical analysis: Levine, Kiefe, Houston, Allison, McCarthy, and Ayanian. Administrative, technical, and material support: Levine and McCarthy. Study supervision: Kiefe, Houston, Allison, and Ayanian.
Financial Disclosure: None reported.
Disclaimer: All analyses, interpretations, and conclusions reached are attributed to the authors (recipients of the data file) and not to the National Center for Health Statistics, which is responsible only for the initial data.