SEARCH

SEARCH BY CITATION

Keywords:

  • Arthritis;
  • Health care utilization;
  • Cost;
  • Longitudinal data

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  9. Appendix

Objective

To evaluate the effect of arthritis on subsequent 2-year use of health care services and out-of-pocket costs among older adults and determine if comorbidities or economic resources mitigate that effect.

Methods

Data were analyzed from 6,230 participants interviewed in 1993 and 1995 in the Asset and Health Dynamic Survey Among the Oldest Old (AHEAD), a national probability sample of community-dwelling adults. Baseline arthritis status was ascertained from the report of an arthritis-related physician's visit or a joint replacement not associated with a hip fracture. The effect of baseline arthritis on the odds of subsequent 2-year health care utilization and high out-of-pocket expenses were estimated from multiple logistic regression controlling for demographic factors, comorbidity, and economic resources.

Results

Older adults with arthritis are significantly more likely to have a physician visit (odds ratio [OR] 3.0), hospital admission (OR 1.6), outpatient surgery (OR 1.3), receive home health care (OR 1.6), and have out-of-pocket cost >$5,000 (OR 1.6) compared with contemporaries having similar demographics (age, sex, racial/ethnic group, marital status), comorbid conditions, and economic resources (education, income, wealth, health insurance), but not reporting arthritis.

Conclusions

Older adults with symptomatic arthritis reported greater medical utilization and cost compared with people not reporting arthritis. These disparities persisted after accounting for differences in demographics, comorbidities, and economic factors. These findings document greater economic burdens on a personal and societal level among people with arthritis. As individuals, older adults with arthritis spend more out-of-pocket dollars for health care than their contemporaries without arthritis. On a societal level, these findings of greater health care utilization among people with arthritis point to increasing future demands on the US health care system due to demographic increases in the numbers of older adults with arthritis and support policies aimed at improving arthritis prevention and treatment as well as reducing the economic disparities between those with and without arthritis.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  9. Appendix

This study presents the effect of arthritis and rheumatic conditions, referred to simply as “arthritis” hereafter, on the subsequent use of medical services among older adults based on the Asset and Health Dynamic Survey Among the Oldest Old (AHEAD) study, a national probability sample of older Americans living in the community.

National survey data document that people with arthritis are substantial users of health care services. The Centers for Disease Control and Prevention (CDC) reported that arthritis is the first listed hospital discharge diagnosis annually for 744,000 hospitalizations, lasting a total of 3.8 million days, based on 1997 National Hospital Discharge Survey data (1). The same CDC report also estimates that 44 million ambulatory care visits are attributed to arthritis, of which almost 39 million were to physician offices, based on the National Hospital Ambulatory Medical Care Survey (1).

Older adults are major consumers of medical services, thus their health care utilization is important from a public policy perspective. Estimates of medical utilization based on the 1996 Medical Expenditure Survey show that adults with musculoskeletal conditions were more likely to use ambulatory (physician and non-physician visits), home health care days, and hospital services (2). Similarly, cross-sectional findings from the 1990 Ontario Health Survey showed that musculoskeletal conditions ranked first among chronic conditions for consultations with a health professional (3). However, it is not known if other factors such as demographics (e.g., age, sex), comorbid conditions, or economic access factors may explain these disparities related to musculoskeletal conditions.

This study adds to the literature by addressing 2 questions. First, “If older Americans have arthritis, how does that affect their subsequent use of medical care?” Second, “Do comorbidity and economic access explain greater health care utilization among people with arthritis?” Evaluating the importance of arthritis as a predictor of medical utilization and whether or not comorbidity and economic factors mitigate this role are issues not addressed by earlier literature.

The AHEAD longitudinal data provide self-reported information on the subsequent use of medical services among older adults who reported arthritis at the baseline (1993) interview. Arthritis is ascertained at the baseline interview based on the report of an arthritis-related physician visit in the prior year, or a joint replacement not related to a hip fracture. This study describes the subsequent 2-year (1993–1995) use of medical services and out-of-pocket costs among older adults with and without self-reported arthritis by demographic factors, health needs (comorbidities and functional limitations), and economic characteristics. Understanding the relationship of arthritis with future medical care and costs may help policy makers to better plan for the health care needs of older adults.

SUBJECTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  9. Appendix

Data source.

Public use 1993 and 1995 data from AHEAD (4), a prospective study of people aged 70 and older (born in 1923 or earlier) were analyzed. The AHEAD is a probability sample of community dwelling elderly persons in the US sponsored by the National Institute on Aging and conducted by the University of Michigan (5). Respondents were sampled from 2 sources to guard against under-representation of the extremely disabled; about half the respondents born in 1923 or earlier were recruited from a national area sample and the remaining half were obtained from a stratified sampling of Medicare enrollees (6). The cohort includes an oversample of the very old (85 years or more), African Americans, Latinos, and residents of Florida. The oldest group was largely interviewed face-to-face while those ages 70–79 primarily had telephone interviews. Specific details of the sampling design of AHEAD are described elsewhere (5, 7, 8). All analyses are based on the weighted sample to reflect this population. AHEAD monitored transitions in physical and functional health, economic resources, and health care demand. Of those selected in the multi-stage sample, 7,447 people from the 1923 or earlier birth cohort participated in 1993 interviews (an 80% response rate).

This study focuses on 6,230 adults aged 70 or older who participated in the baseline (1993) interview and provided information in 1995 regarding health care utilization over the intervening 2 years. The 1995 interview status of the initial AHEAD cohort includes 6,230 interviewed, 794 decedents, and 423 non-respondents. We restrict analyses to the 6,230 individuals who were alive and interviewed in 1995, due to dependence on self-reported utilization information. Mortality over 2 years was similar among people who did and did not report baseline arthritis (10.9% versus 10.6%).

Medical utilization outcomes.

AHEAD provides self-reported information on the 2-year use medical services following the baseline interview that includes physician visits, hospital stays, outpatient surgery, home health care, and nursing home stays. Respondents were asked in 1995 if in the last 2 years they saw or talked to a medical doctor including emergency room and clinic visits (any physician visit); if they were an overnight hospital patient (any hospital admission); if they had outpatient surgery; if a medically trained person came to the home (among people not residing in a nursing home); and if they had been a nursing home patient. Respondents were also asked about out-of-pocket costs (costs not covered by third parties) for each of these services.

Baseline arthritis.

The baseline report of arthritis is ascertained from either an affirmative response to the 1993 interview question, “During the last 12 months, have you seen a doctor specifically for arthritis or rheumatism?” and/or the report of joint replacement surgery for reasons other than a hip fracture. For the purpose of analysis, 38 (0.6%) people reporting arthritis at baseline who corrected their baseline status to no arthritis at the 1995 interview are classified as not having baseline arthritis.

Baseline demographic data.

Demographic information includes marital status (married, divorced, separated, widowed, never married), age, sex, and ethnicity (white/Caucasian, black/African American, Latino, other). Ethnicity was used to categorize race into non-Latino white, non-Latino black, Latino, and other.

Baseline comorbid conditions.

Information was obtained in seven categories of chronic conditions in addition to arthritis: diabetes, cancer, hypertension, heart disease, lung disease, stroke, and psychiatric problems. Participants were asked if they currently had diabetes or if a physician ever told them that they had cancer, hypertension, lung disease (chronic bronchitis, emphysema), heart disease (heart attack, coronary artery disease, congestive heart failure, angina, other), or stroke and if they ever had or had ever been told by a physician that they had psychiatric problems (emotional, nervous, or psychiatric). Obesity was determined from a body mass index (BMI) over 30 calculated from self-reported height and weight (weight [kg])/ (height [m]2). Imputed height or weight (age/sex/race specific sample median) was used to estimate BMI for 77 (1.03%) people, for whom that information was missing. Analytic results with and without imputed BMI values were identical.

Baseline economic access measures.

Baseline (1993) economic access is measured by an individual's level of human capital and financial ability to pay for medical care, consistent with Andersen's widely-used health services utilization model (9, 10). Human capital is measured by education; financial ability to pay is measured by income, wealth, and health insurance. Education was determined from completed years of education reported at the initial interview. Family income was determined from the income received by the respondent and spouse/partner during the preceding year from all sources. (AHEAD asks respondents about all income received the previous month from Social Security, Supplemental Security Income, private pensions, labor income, annuities, IRAs, stocks and bonds, veteran's benefits, food stamps, and other sources) (11). For respondents who were unable or unwilling to provide exact amounts, the interviewer gave the opportunity to provide an amount within a bracketed category. Imputed estimates of income utilize bracketed responses (12). Net worth represents household net worth, which summarizes the household's tangible wealth in terms of housing equity and non-housing equity (e.g., savings). Similar procedures were used to minimize item non-response. Imputed values provided on the public use files were used for cases of missing income and wealth data. Health insurance information related to Medicare, Medicaid, other government insurance, and private insurance coverage.

Analysis.

The AHEAD is a probability sample of the national population. All analyses are weighted and adjust for the complex sampling design to reflect the national population. Logistic regression was used to estimate the odds of using each type of health care service in the 2 years following the baseline interview. These models were estimated using SUDAAN software (13), by applying Taylor series methods with between-cluster robust estimation to adjust for the complex sampling design (14).

Adjusted odds ratios are calculated to account for the potential influence of other factors on differences in utilization and cost related to arthritis. These adjusted analyses are done hierarchically, based on Andersen's model of health services utilization (9, 10). This model evaluates the effect of 1) demographic or predisposing variables (e.g., age, sex, ethnicity, marital status) that influence the propensity of individuals to use health services, 2) health conditions or needs, the most immediate cause of health services use, and 3) economic access or enabling variables, that describe the economic means and human capital people have to access medical services. Multiple logistic analyses were developed in hierarchical stages entering 1) demographics, 2) comorbid conditions, and 3) economic access factors.

Analyses are restricted to people who participated in both the 1993 and 1995 interviews. Of those alive, 94% participated in the 1995 interview. To make utilization statements about people alive in 1995, we adjusted for potential bias due to non-response by handling respondents as an additional sampling stage to obtain sampling weights for 1995 respondents, using standard sampling methodology (15). Respondents were compared with non-respondents to identify differential baseline characteristics, which included an incomplete baseline interview, hearing problems, withholding permission to obtain additional records, non-response to sensitive baseline questions, and geographic region. The sampling weight for 1995 respondents equaled the 1993 AHEAD sampling weight multiplied by the inverted probability of having a 1995 interview given these characteristics; that probability was estimated using logistic regression.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  9. Appendix

Data were analyzed from 6,230 respondents to the baseline 1993 AHEAD survey who provided subsequent health care information at the 1995 interview. This allowed baseline risk factors to be evaluated as potential predictors of subsequent use of medical services.

Table 1 shows the distribution of demographic, health needs, and economic access factors for the entire sample and by baseline arthritis status. The cohort represented a population with a mean age of 77 years, a median family income of $15,000, which was predominantly non-Latino white (86%), female (63%), with at least a high school education (59%), and almost all had Medicare coverage (97%). Older adults with baseline arthritis compared with people not reporting arthritis were demographically more likely to be female, older, minority members (African American or Latino), and were less likely to be married. Comorbid conditions were more frequently reported by older adults with arthritis, particularly diabetes, hypertension, heart disease, and obesity. Older adults with baseline arthritis reported fewer economic access resources including less education, income, and wealth; in addition, they were more likely to hold Medicaid health insurance coverage to supplement Medicare coverage compared with those without arthritis. These demographic, health need, and economic access differences among people with and without arthritis are consistent with what is reported in other literature (16–18).

Table 1. Asset of health dynamics among the oldest old study
Baseline (1993) characteristicsSample n = 6,230 Population %Arthritis baseline n = 1,771 Population %No arthritis baseline n = 4,459 Population %
Demographic factors   
 Race/ethnicity   
  Non-Latino white85.578.188.4
  African American9.514.47.6
  Latino3.86.12.9
  Other1.21.41.2
 Sex   
  Male37.230.039.9
  Female62.870.060.1
 Age in years   
  70–7969.565.770.9
  ≥8030.534.329.1
 Marital status   
  Currently married or live with partner51.745.254.1
  Widowed/divorced45.251.942.6
  Never married3.12.93.2
Chronic conditions   
 Cancer13.312.313.7
 Diabetes11.914.111.1
 Hypertension49.357.846.0
 Heart disease30.135.228.1
 Lung disease10.312.39.6
 Obesity27.536.324.5
 Psychiatric10.813.39.8
 Stroke7.37.47.3
Economic access factors   
 Education in years   
  <1240.848.737.8
  =1231.328.132.5
  >1227.923.129.8
 Annual family income   
  <$7,50017.321.415.7
  $7,500–$14,99927.229.526.3
  ≥$15,00055.649.058.1
 Net assets   
  <$1,0009.613.58.1
  $1,000–$49,99924.027.722.6
  ≥$50,00066.458.869.3
 Health insurance   
  Medicare + Private/Government76.171.178.0
  Medicare + Medicaid8.213.06.3
  Medicare only12.913.412.7
  Other health insurance2.32.32.3
  No health insurance0.60.20.7

Figure 1 shows the frequency with which older adults used health care services and the proportion with high (>$5,000) reported out-of-pocket costs by baseline arthritis status. Older adults with arthritis at the baseline interview consistently used more health care services and had greater costs in the subsequent 2 years compared with people not reporting arthritis. People with arthritis were more likely to use physician (98.0% versus 93.6%, OR 3.4, 95% confidence interval [95%CI] 2.4–4.8), hospital (44.3% versus 31.0%, OR 1.8, 95% CI 1.6–2.0), outpatient surgery (22.2% versus 18.8%, OR 1.2, 95% CI 1.0–1.5), home health care (21.7% versus 12.3%, OR 2.0, 95% CI 1.7– 2.3), and nursing home services (8.2% versus 6.0%, OR 1.4, 95% CI 1.1–1.7) in the subsequent 2 years compared with people not reporting arthritis.

thumbnail image

Figure 1. Two-year medical utilization and cost (1993–1995) by baseline arthritis status. OR = odds ratio.

Download figure to PowerPoint

In addition, Figure 1 shows that 53.6% of older adults with baseline arthritis reported high out-of-pocket costs compared with 40.5% of people not reporting arthritis (OR 1.7, 95% CI 1.5–2.0). Prescription drugs not covered by health insurance contribute to these out-of-pocket costs. In this older cohort, people with arthritis are more likely to use prescription drugs (88.3% versus 76.0%). However, the added costs of prescription drugs are not absorbed by health insurance since similar proportions of people with and without baseline arthritis using prescription medication lack full coverage for prescription costs (79.3% versus 83.7% report none or partial coverage for prescription drugs). As a result, people with arthritis more frequently pay prescription drug costs not fully covered by health insurance, adding to their out-of-pocket medical costs.

It is recognized that other factors influence the use of medical services, in addition to arthritis. People who are older, with more comorbid conditions and better economic resources are more likely to use medical care. For this reason, the impact of arthritis was examined on utilization and expense of medical services accounting for other potentially influential factors.

Table 2 summarizes the odds ratios of 2-year utilization and out-of-pocket costs among older adults with arthritis compared with people not reporting arthritis from hierarchically staged multiple logistic regression analyses that sequentially entered: demographics (age, sex, marital status, racial/ethnic group); chronic conditions (cancer, diabetes, heart condition, hypertension, obesity, psychiatric problems, stroke); and finally economic access factors (education, income, wealth, health insurance). Controlling for demographics had little influence on the magnitude of the effect from arthritis on greater health care utilization or high out-of-pocket costs, although its effect on nursing home use became marginally nonsignificant. While additional control for other chronic conditions modified the greater odds ratios due to arthritis, the effect remained statistically significant for physician visits (OR 3.0, 95% CI 2.1–4.3), hospital admissions (OR 1.6, 95% CI 1.5–1.8), outpatient surgery (OR 1.3, 95% CI 1.1–1.5), home health care (OR 1.7, 95% CI 1.4–2.0), and out-of-pocket costs (OR 1.6, 95% CI 1.4–1.8). These effects remained significant even after further adjustment for economic access factors, which did little to moderate the OR related to arthritis on the increased use of medical care and high out-of pocket costs. It should be noted that the estimates in Table 2 of the effect of arthritis are conservative in the following sense. By 1995 an additional 1,736 (28%) people reported arthritis who had not at baseline; their 1993–1995 medical services and costs may be partially influenced by the development of arthritis. If the utilization of this incident cohort were added to the baseline arthritis group, the odds ratios related to baseline arthritis would be even larger than those shown.

Table 2. Odds of 2-year medical utilization (1993–1995) and out-of-pocket costs due to arthritis*
Medical utilizationnArthritis OR adjustment factors (referent is no arthritis)
Unadjusted OR (95% CI)Demographics OR (95% CI)Plus chronic conditions OR (95% CI)Plus economic access§ OR (95% CI)
  • *

    OR = odds ratio; 95% CI = 95% confidence interval.

  • Adjusted for demographics (age: 70–79, 80 or older years; gender: male, female; marital status: married or live with partner, widowed/divorced, never married, race: non-Latino white, Latino, non-Latino African-American, other) and reported for reference categories in bold.

  • Adjusted for demographics + chronic conditions (none, cancer, diabetes, hypertension, heart, lung, obesity, psychiatric, stroke) and reported for reference categories in bold.

  • §

    Adjusted for demographics + chronic conditions + access (education: <9, 9–11, 12, >12 years; income: <$7,500, $7,500–14,999, $15,000 or above; wealth: <$1, $1–49,999, $50,000 or above, health insurance: Medicare only, Medicare plus Medicaid, Medicare plus private/government, other) and reported for reference categories in bold.

  • Excludes people residing in nursing home at 1995 interview.

Physician visit6,0183.4 (2.4–4.8)3.6 (2.5–5.1)3.0 (2.1–4.3)3.0 (2.1–4.3)
Hospital admission6,2241.8 (1.6–2.0)1.8 (1.6–2.0)1.6 (1.5–1.8)1.6 (1.4–1.8)
Outpatient surgery6,2251.2 (1.1–1.5)1.3 (1.1–1.6)1.3 (1.1–1.5)1.3 (1.1–1.6)
Home health care5,9802.0 (1.7–2.3)1.8 (1.5–2.1)1.7 (1.4–2.0)1.6 (1.4–1.9)
Nursing home stay6,2301.4 (1.1–1.7)1.3 (1.0–1.6)1.3 (1.0–1.6)1.3 (1.0–1.6)
Out of pocket >$5,0006,0141.7 (1.5–2.0)1.8 (1.5–2.0)1.6 (1.4–1.8)1.6 (1.4–1.8)

The appendix summarizes the odds ratios from the final health services and cost models that control for demographics, comorbid conditions, and economic access factors. As expected, comorbid conditions are significantly associated with increased use of medical services and cost, controlling for demographics. All comorbid conditions (cancer, diabetes, heart disease, hypertension, lung disease, psychiatric problems, and stroke) predicted higher out-of pocket costs. Substantial subsets of these conditions were significant predictors for the use of physician, hospital, outpatient, and home health care services. Only stroke significantly predicted nursing home utilization.

In contrast, for this cohort composed primarily of Medicare beneficiaries, economic access factors had few significant effects on the use of medical services or cost. Economic access factors did not significantly predict the use of hospital or home health care services. Nor did they predict high out-of-pocket expenses. Although some coefficients were statistically significant in models for physician contacts, outpatient surgery, and nursing home services, there were no clear patterns suggesting how economic access factors affected these variables.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  9. Appendix

This study documents the effects of arthritis “downstream” in terms of future use of medical care and out-of-pocket costs, based on longitudinal data from a national probability sample of older adults. The experience from this cohort documents that older adults with arthritis have subsequently greater 2-year use of medical services compared with those without this condition, including increased 2-year utilization of physician, hospital, outpatient, home health and nursing home care (whether or not related to arthritis). In addition, people with baseline arthritis were more likely to report high subsequent out-of-pocket medical expenditures. With the exception of nursing home services, arthritis remained a significant predictor of higher utilization and cost even after controlling for demographics, comorbid conditions, and economic factors.

Several limitations of this study should be noted. First, AHEAD relies on self-reported as opposed to physician-confirmed diagnoses of arthritis and rheumatism. Arthritis was ascertained from self-reported arthritis-related physician contacts. This approach assumes that individuals had symptoms (i.e., pain, swelling, stiffness, limitation, etc.) of arthritis severe enough to seek care for this problem from a physician. Prior studies indicate that the most reliable self-reports of arthritis are associated with symptomatic conditions (19).

Second, AHEAD collects utilization information via self-report, as is common for large, nationally representative studies. When self-reported data are compared with administrative data, there is evidence that self-reports underestimate actual utilization (20). One study of older adults who self-reported utilization within the past year found significant underreporting for both physician use and hospitalizations (21). However, numbers of visits were subject to the greatest underreporting. To limit this problem, we emphasize measures of whether any utilization occurred rather than the quantity of utilization. In addition there was weak, if any correlation of underreporting with specific predictor variables, so this effect should be similar among people with and without baseline arthritis.

Third, these analyses are adjusted for potential non-reporting bias, but are limited to those alive at 2-year follow-up. There is no differential loss of information due to deaths, since people with and without arthritis had similar 2-year mortality (10.9% versus 10.6%, respectively). However, these results cannot be extended to make inferences about utilization and cost of decedents, since those who died may be expected to have higher utilization rates, on average, but a shorter exposure period. Finally, the reader is reminded that this study evaluates medical utilization prior to the Balanced Budget Act (BBA) of 1997. The BBA reduced support for indigent hospital care through reductions in Medicare disproportionate share hospitals (DSH) payments, outlier supplements, and indirect medical education payments. The BBA of 1997 also constrained the growth of state Medicaid DSH and repealed the Boren Amendment, which may negatively affect state Medicaid payment formulae that traditionally favored providers with high indigent costs. Later, the Balanced Budget Refinement Act of 1999 reduced many of these adverse payment effects. Although these changes were not in effect during the study period, they could increase access related effects beyond those observed in this study.

Despite these limitations, this study extends our understanding of the impact of arthritis on medical utilization and cost among older adults. First, this study had notable methodologic strengths. The longitudinal design of this analysis and the use of a recent national probability sample of older adults provides an assessment of the impact of arthritis on 2-year health care utilization and costs that applies to the national population as a whole. Second, to our knowledge, this is the first study using recent population-based data to report higher out-of-pocket costs among older US adults with arthritis. Third, these findings document that arthritis is related to greater future use of health services.

Finally, this study explores whether greater utilization and cost related to arthritis are explained by greater burdens from other chronic conditions or differences in economic access. Although comorbid conditions are strong predictors of higher out-of-pocket costs and almost all types of medical utilization, as expected, it is notable that arthritis remains an independently strong predictor. In contrast, measured economic factors are not related to higher out-of-pocket costs. Furthermore, there is no informative pattern from utilization analyses among significant economic factors suggesting how these factors affect health care utilization in this cohort primarily insured through Medicare. The effect of comorbid conditions and economic access factors on use of medical services for this cohort covered largely under Medicare are described elsewhere (22). More importantly, these findings show that older adults with arthritis have greater subsequent 2-year use of medical services and higher out-of-pocket costs compared with their contemporaries with similar demographic, comorbid condition, and economic access profiles but not reporting arthritis at baseline.

These results from longitudinal AHEAD data show that older adults with arthritis have greater use of health services (physician, hospital, outpatient surgery, home health care) before adjustment for other risk factors compared with people without arthritis. These findings in older adults are consistent with other recent studies based on the general population that show greater unadjusted medical utilization among people with arthritis related to physician, home health, and nursing home services (3, 16). However, unique to this report are additional findings that show this relationship persists after adjusting for demographics, comorbid conditions, and economic access. Findings of persistent differences in adjusted utilization due to arthritis based on analyses using the 1993–1995 AHEAD data are in contrast to published results from earlier longitudinal studies. Analyses using 1984–1990 Longitudinal Study on Aging (LSOA) data did not show greater physician (23) or hospital utilization (24) related to arthritis after adjusting for comorbid conditions and limited measures of economic access. However, those LSOA analyses used data that largely predated the effects of managed care and evaluated 6-year rather than 2-year utilization.

In addition, this study documents that older adults with arthritis are more likely to report high out-of-pocket medical costs than people without arthritis but having similar profiles of demographics, chronic conditions, and economic factors. Furthermore, since people with arthritis more frequently report prescription drug use not fully covered by health insurance, the higher out-of-pocket costs reflect disparities in prescription drug costs borne by people with arthritis.

Recognizing that annual physician visits are recommended for people over 65 years for preventive care (25), the greater likelihood for people with arthritis to see a physician has potential benefit. However, the bad news is that people with arthritis are also more likely to use hospital, outpatient, and home health services, which generally represents reactive health care. Even worse news is the finding that people with arthritis incur greater out-of-pocket costs compared with similar contemporaries without arthritis. Given demographic increases in the number of older adults with arthritis, which will accompany the aging of the baby-boomer generation, these findings point to heavier demands on the US health care system. These findings are important for policy makers whose interest is containing rising medical costs. Certainly, better arthritis treatment and prevention strategies are needed to reduce the increases in health care utilization borne by persons with arthritis. Policy efforts to increase public funding for research pertaining to arthritis prevention and treatment are supported by our findings. Our findings also support increased public funding for arthritis public health initiatives aimed at the prevention of arthritis and its associated disability. Finally, if our findings that persons with arthritis have more out-of-pocket medical care costs are replicated, then future health care policy, (e.g., the establishment of a Medicare prescription drug benefit), would need to address this unjustified adverse economic burden, especially if this is due to an increased reliance on non-covered drug prescription costs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  9. Appendix

The authors gratefully acknowledge the comments of Mr. Jason Bredle.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  9. Appendix
  • 1
    Centers for Disease Control and Prevention. Impact of arthritis and other rheumatic conditions on the health-care system. MMWR Morb MortalWkly Rep 1999; 48: 34953.
  • 2
    Yelin E, Herrndorf A, Trupin L, Sonneborn D. A national study of medical care expenditures for musculoskeletal conditions: the impact of health insurance and managed care. Arthritis Rheum 2001; 44: 11609.
  • 3
    Badley EM, Rasooly I, Webster GK. Relative importance of musculoskeletal disorders as a cause of chronic health problems, disability, and health care utilization: findings from the 1990 Ontario Health Survey. J Rheumatol 1994; 21: 50514.
  • 4
    National Institute on Aging. Asset and Health Dynamics Among the Oldest Old Study, Wave 1 and 2 public use dataset. University of Michigan; 1999.
  • 5
    Heeringa S. Technical description of the Asset and Health Dynamics Among the Oldest Old (AHEAD) Study Sample Design. Ann Arbor (MI): University of Michigan Population Studies Center; 1995.
  • 6
    Rodgers W. Comparison of two sampling frames for surveys of the oldest old. In: WarneckerRB, ed. Health Survey Research Methods. Washington (DC): US Department of Health and Human Services; 1996. p. 119122.
  • 7
    Soldo BJ, Hurd MD, Rodgers WL, Wallace RB. Asset and Health Dynamics Among the Oldest Old: an overview of the AHEAD Study. J Gerontol B Psychol Sci Soc Sci 1997; 52: 120.
  • 8
    National Institute on Aging. Health and Retirement Study: http://www.umich.edu/∼hrswww/data/index.html, 1998. Accessed January 15, 2002.
  • 9
    Andersen RM. A behavioral model of families' use of health services. Chicago (IL): University of Chicago Center for Health Administration Studies; 1968.
  • 10
    Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav 1995; 36: 110.
  • 11
    Schoenbaum M, Waidmann T. Race, socioeconomic status, and health: accounting for race differences in health. J Gerontol B Psychol Sci Soc Sci 1997; 52 Spec: 6173.
  • 12
    Smith JP. Wealth inequality among older Americans. J Gerontol B Psychol Sci Soc Sci 1997; 52 Spec: 7481.
  • 13
    Shah BV, Barnwell BG, Bieler GS. SUDAAN User's Manual Release 7.5. Research Triangle Park (NC): Research Triangle Institute; 1997.
  • 14
    Williams R. A note on variance estimation for cluster-correlated data. Biometrics 2000; 56: 2189.
  • 15
    United States Bureau of the Census. The Current Populations Survey -A report on methodology. Washington (DC): US Government Printing Office; 1963.
  • 16
    Yelin EH, Trupin LS, Sebesta DS. Transitions in employment, morbidity, and disability among persons ages 51-61 with musculoskeletal and non-musculoskeletal conditions in the US, 1992-1994. Arthritis Rheum 1999; 42: 76979.
  • 17
    Yelin EH. Musculoskeletal conditions and employment. Arthritis Care Res 1995; 8: 3117.
  • 18
    Yelin E. The earnings, income, and assets of persons aged 51-61 with and without musculoskeletal conditions. J Rheumatol 1997; 24: 202430.
  • 19
    Hughes SL, Edelman P, Naughton B, Singer RH, Schuette P, Liang G, et al. Estimates and determinants of valid self-reports of musculoskeletal disease in the elderly. J Aging Health 1993; 5: 24463.
  • 20
    Cleary PD, Jette AM. The validity of self-reported physician utilization measures. Med Care 1984; 22: 796803.
  • 21
    Wallihan DB, Stump TE, Callahan CM. Accuracy of self-reported health services use and patterns of care among urban older adults. Med Care 1999; 37: 66270.
  • 22
    Dunlop DD, Manheim LM, Song J, Chang RW. Gender and ethnic/racial disparities in health care utilization among older adults. J Gerontol B Psychol Sci Soc Sci 2002; 57: 522133.
  • 23
    Stump TE, Johnson RJ, Wolinsky FD. Changes in physician utilization over time among older adults. J Gerontol B Psychol Sci Soc Sci 1995; 50: S45S58.
  • 24
    Verbrugge L. Disability transitions for older persons with arthritis. J Aging Health 1992; 4: 21243.
  • 25
    United States Preventive Services Tasks Force. Guide to clinical preventive services. Baltimore: Williams & Wilkins; 1996.

Appendix

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  9. Appendix
Table  . Appendix Odds of 2-year medical utilization (1993–1995) and out-of-pocket cost final models*
Baseline independent variablesAny physician contact (n = 6,018)Any hospital admission (n = 6,224)Outpatient surgery (n = 6,225)Home health care (n = 5,980)Nursing home stay (n = 6,230)Out of pocket >$5,000 (n = 6,014)
  • *

    Values are odds ratios.

  • Excludes people residing in nursing home at 1995 interview.

  • P < 0.01.

  • §

    P < 0.05.

Arthritis (yes)3.001.601.321.641.28§1.56
Predisposing      
 Age in years      
  70–79 (Reference)
  ≥801.131.301.111.953.201.22
 Sex      
  Female (Reference)
  Male0.831.291.210.750.911.17
 Marital status      
  Married (Reference)
  Widowed/divorced0.991.141.091.151.681.02
  Never married0.640.991.190.842.470.87
 Ethnic group      
  Non-Latino White (Reference)
  Non-Latino African-American0.950.880.811.010.720.93
  Latino1.140.700.731.000.29§0.37
  Other3.450.660.33§1.170.600.53
Health needs      
 Cancer2.09§1.261.261.350.971.55
 Diabetes2.241.541.112.141.331.57
 Hypertension2.991.211.201.110.831.32
 Heart Condition2.801.651.071.301.051.75
 Lung Disease1.961.551.24§1.491.061.62
 Obesity1.241.070.991.151.041.24
 Psychiatric1.511.411.101.391.371.27§
 Stroke0.971.680.881.952.121.66
Economic access      
 Education in years      
  <90.881.060.711.011.010.99
  9–110.931.001.000.960.861.13
  12 (Reference)
  ≥131.69§0.931.000.881.040.97
 Income      
  <$7,5000.690.951.101.090.791.01
  $7,500–$14,999 (reference)
  ≥$15,0001.090.911.280.880.66§0.96
 Assets      
  <$1,0000.791.080.901.001.231.11
  $1,000–$49,999 (reference)
  $50,000 or more1.230.891.160.840.660.98
 Health insurance      
  Medicare only0.570.900.950.841.240.91
  Medicare + Medicaid1.411.280.921.160.801.26
  Other health insurance0.570.941.430.632.681.01
  No health insurance0.591.121.371.272.100.94
  Medicare + private/Government (reference)