Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner

Health Care Costs and Financing

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Medicare and privately insured patients are likely to pay more than others for similar home health services

The home health industry has grown tremendously over the past decade, with expenditures for these services increasing from $5.6 billion in 1985 to almost $25 billion in 1993. Nearly two-thirds of home health patients have Medicare as their primary source of payment; about 10 percent of patients have either private insurance or pay out of pocket. According to a recent study by the Agency for Health Care Policy and Research, Medicare and private insurers may pay more than other payors for such services.

Former AHCPR staff members Vicki A. Freedman, Ph.D., (currently with RAND) and James D. Reschovsky, Ph.D., (currently with the Center for Studying Health System Change) used the 1992 National Home and Hospice Care Survey to compare the average charge for a Medicare home health visit to the average charge for patients with other sources of payment. They found that on average, home health patients were charged $75 per visit by home health care agencies in 1992. Those covered by Medicare were charged an average of $80 per visit, about $10 more than Medicaid patients, $19 more than patients paying out-of-pocket, and twice as much as patients with other primary sources of payment. Patients with private insurance were charged an average of $88. Differences between Medicare beneficiaries and the privately insured were small (about 6 percent) and not significant. These differences persisted even after controlling for disparities across payors in mix of home health services and agency characteristics.

Charges do not generally correspond to actual reimbursement, especially for Medicare and Medicaid. Medicare pays the lower of cost ceilings or charges. For Medicaid, incentives for charge-setting vary by State depending on the reimbursement system.

In 1992, charges by home health agencies for Medicare beneficiaries exceeded actual payments to those agencies by 25 percent. Although this research analyzes charges, not payments, the findings suggest that Medicare and private pay insurers may be paying more than other payors for the same services.

For details, see "Differences across payors in charges for agency-based home health services: Evidence from the national home and hospice care survey," by Drs. Freedman and Reschovsky, in the October 1997 Health Services Research 32(4), pp. 433-452. Reprints (AHCPR Publication No. 98-R021) are available from the AHCPR Publications Clearinghouse.

Rural residents are more likely than city-dwellers to be without health insurance for longer periods

About 40 million Americans are without health insurance coverage. Spells of uninsurance have increased in length for all citizens, particularly rural residents. Clearly, uninsured individuals do visit their physicians less often than those with insurance. However, when urban residents finally obtain health insurance, the number of physician visits actually declines. Rural residents, on the other hand, visit the doctor as often or as little as when they had no insurance, according to a study supported by the Agency for Health Care Policy and Research (HS05760).

Keith J. Mueller, Ph.D., of the University of Nebraska Medical Center, and his colleagues examined the relationship between insurance status and number of physician visits from 1989 to 1993. They also looked at spells of uninsurance among a sample of 1,235 Nebraska residents who were interviewed in 1992 and were asked about their insurance status for the previous 5 years. The researchers found that the median length of uninsurance spells was 14 months, with three-fourths of the spells lasting at least 6 months and one-fourth lasting longer than 34 months. Median uninsurance spells were about three times longer among rural (16 months) and frontier (22 months) residents compared with their urban neighbors (6 months), who are more apt to have employer-based insurance.

Urban residents actually recorded more physician visits at the end of a spell of uninsurance than at the beginning of the next spell of insurance. For rural and frontier residents, the pattern was reversed. Frontier county residents changed from an average of 0.16 physician visits in the 3 months ending a spell of uninsurance to an average of 0.43 visits in the first 3 months of having insurance. The difference was less dramatic among non-frontier rural residents, but it was in the same direction. Urban residents may have increased physician visits at the end of an uninsured spell because of availability of free clinics and the prospect of copayments or a deductible with insurance, or it may be in anticipation of preexisting condition clauses in the insurance plans. Rural residents had longer spells of uninsurance and hence a greater likelihood of pent-up demand. Also, they are more likely not to seek physician services when they cannot pay, both because of pride and because of limited opportunities for free clinic care, notes Dr. Mueller.

More details are in "Lengthening spells of uninsurance and their consequences," by Dr. Mueller, Kashinath Patil, Ph.D., and Fred Ullrich, in the Winter 1997 issue of The Journal of Rural Health 13(1), pp. 29-37.

Return to Contents

Elderly/Long-term Care

Low-risk elderly patients may be safely discharged early from the hospital following bypass surgery

In the wake of cost-cutting efforts, more patients are being discharged from the hospital within 5 days of coronary artery bypass graft (CABG) surgery. This has raised concerns that such early discharges may compromise quality of care, particularly for elderly patients. However, a recent study supported by the Agency for Health Care Policy and Research (HS06503 and HS08805) suggests that low-risk elderly CABG patients may be able to be safely discharged within 5 days of surgery. In particular, it showed that early discharge did not raise the risk of short-term (60-day) mortality or readmission among elderly CABG patients.

The Ischemic Heart Disease Patient Outcomes Research Team (PORT), supported by AHCPR and led by Elizabeth R. DeLong, Ph.D., of Duke University, used Medicare discharge data from the 1992 National Claims History File to examine the prevalence of early discharge (postsurgical hospital stay of 5 days or less) among 83,347 non-HMO Medicare patients who underwent CABG in the United States in 1992. Most patients were white (94 percent) and male (68 percent), with an average age of 72 years. The researchers identified patient characteristics associated with early discharge and obtained rates of readmission or death for postoperative stays between 4 and 14 days.

In 1992, 6 percent of Medicare CABG patients were discharged within 5 days of the operation. But the prevalence of early discharge varied considerably among States, ranging from 1 percent to 21 percent. Patients discharged early tended to be younger and male and have fewer coexisting illnesses. They also had the lowest rates of death and hospital readmission and lower odds of experiencing an adverse outcome relative to patients with average postoperative stays. This suggests that physicians were able to identify low-risk candidates for early discharge and that higher rates of early discharge might be safely achieved, conclude the researchers.

For more information, see "Impact of early discharge after coronary artery bypass graft surgery on rates of hospital readmission and death," by Patricia A. Cowper, Ph.D., Eric D. Peterson, M.D., M.P.H., Dr. DeLong, and others, in the October 1997 Journal of the American College of Cardiology 30, pp. 908- 913.

Occupational therapy may reduce health declines in older adults

Individualized occupational therapy (OT) that focuses on development of needed skills to cope with everyday activities—such as healthful shopping and mastering the local public transportation system—may be the answer for many older, independent-living adults who want to improve their quality of life, psychological well-being, and overall health. According to a recent study, preventive health programs based on individualized OT may mitigate against the health risks of older adults. The study was supported by a joint agreement between the National Institute on Aging and the Agency for Health Care Policy and Research (AG11810).

The study's findings show that older adults need more than social activity or arts and crafts, notes Florence Clark, Ph.D., of the University of Southern California. Dr. Clark and her colleagues randomly assigned 361 culturally diverse men and women 60 years of age and older to one of three groups: occupational therapy, social activities, and no intervention. The subjects included residents living in one of two Government-subsidized apartment complexes for independent-living older adults and residents living in nearby private homes who used the complexes' senior centers. The OT sessions included 2 hours of group OT a week and 9 total hours of individual OT during the 9-month study period. The sessions showed adults with limited income and resources how to select and perform activities, such as grooming and exercising, to achieve a healthy and satisfying lifestyle. The social activities ranged from community outings and dances to craft projects.

Persons who received OT ranked higher on a life satisfaction index, health perception survey, and seven of eight health status survey scales that measured items ranging from bodily pain and social functioning to general mental health. Also, individuals in the OT group that had certain medical problems tended to improve or decline less severely over time than those in the social activity and nontreatment control groups, who tended to decline over time. The finding that only 5 of 15 outcome measures failed to demonstrate a significant gain for the OT group relative to the control groups provides solid evidence of the positive effects of OT, according to the researchers. They attribute the success of the OT program to its focus on meaningful and health-promoting activities and to its individual tailoring.

See "Occupational therapy for independent-living older adults," by Dr. Clark, Stanley P. Azen, Ph.D., Ruth Zemke, Ph.D., O.T.R., and others, in the October 22, 1997, Journal of the American Medical Association 278(16), pp. 1321-1326.

Researchers examine use of health care networks by older adults

The long-term care networks used by the elderly vary by factors such as race, residence, and age. Two recent studies, led by University of Florida researchers Julie K. Netzer, Ph.D., and Chuck W. Peek, Ph.D., and supported in part by the Agency for Health Care Policy and Research (National Research Service Award fellowships F32 HS00086 and F32 HS00088), identify some of these factors.

The first study (HS00086 and HS00088) shows that race has a significant effect on the use of community-based services but not use of in-home services. The second study (HS00086) demonstrates that once elderly persons activate the caregiving network, it undergoes many changes in type and source of care over time.

Netzer, J.K., Coward, R.T., Peek, C.W., and others (1997, September). "Race and residence differences in the use of formal services by older adults." Research on Aging 19(3), pp. 300-332.

Black elders, especially rural blacks, are more likely to use community-based services than other elderly adults, concludes this study. It shows, for instance, that black rural elders had the highest rate of using senior centers; nearly one-third of these elders (30 percent) reported attending a senior center. Also, older blacks were nearly four times as likely as older whites to report using special transportation services (12.4 percent vs. 3.5 percent). In urban settings, the odds of older blacks using senior centers were more than twice as high as for their white counterparts. And in rural areas, older blacks were five times more apt to use these centers than their white neighbors.

Black rural elders in this study were more likely than whites to emphasize the value of formal community-based services in improving the quality of their lives than white elders living in the same communities. Location of the community-based services also played a role. The only senior center in several of the rural sites studied was located in the black community and staffed predominantly by blacks. Blacks in these communities expressed strong positive attitudes about the senior center and special transportation services. Whites, on the other hand, reported feeling uncomfortable about attending mainly black senior centers or using special transportation used primarily by blacks. Instead, elderly whites may rely more extensively on church groups, grandmothers' clubs, relatives, and friends for the help they need, explain the researchers.

They used data from a longitudinal study of race and residence differences in the use of long-term care among 1,200 community-dwelling adults 65 years of age and older and living in one of four North Florida counties that include a range of residential contexts (i.e., a large city, several small towns, and rural farming areas). They examined use of two community-based services (senior centers and special transportation services for senior citizens) and two home-based services (homemaker and nursing care). Analysis revealed no racial differences in use of homemaker services or home health services. However, rural elders were more apt than urban elders to use both services (9.3 percent and 8.6 percent vs. 3.7 percent and 5.4 percent, respectively).

Peek, C.W., Zsembik, B.A., and Coward, R.T. (1997, September). "The changing caregiving networks of older adults." Research on Aging 19(3), pp. 333-361.

More than 70 percent of elderly persons who begin a 2-year period with no help needed for daily living tasks, ranging from eating and dressing to shopping and housework, remain independent of the care of others at the end of that time. Once they activate caregiving networks, they tend to continue to use them for substantial periods of time, concludes this study. It shows, for example, that among the elderly who began an interval receiving exclusively informal care (unpaid care, usually from family members), 44 percent still had only informal caregivers, and 19 percent had no help from others 2 years later. Conversely, 15 percent of elderly individuals who began a period receiving only informal care needed more care within 2 years: 11 percent added a formal (paid) care provider, and nearly 5 percent were receiving institutional care by the end of 2 years.

Of elderly persons who began a period with only formal help, 38 percent continued to receive only formal help at the end of 2 years, and 19 percent ended the time period with no helpers. Nearly 15 percent added an informal helper, while 3 percent received institutional care. About one-third (34 percent) of those who entered a period receiving mixed care (formal and informal help) remained in that state 2 years later. This suggests that mixed care is not simply a transitory response to a health crisis, and that a significant number of elders receive intensive community-based care from both formal and informal sources on a long-term basis.

Older adults whose caregiving network contained both formal and informal care appear to be among the most frail elderly in this study. For instance, 30 percent had died and 11 percent were residing in a nursing home 2 years later. Similarly, 93 percent of those who began an interval in a nursing home either continued to receive institutional care or had died by end of the interval. As the sample aged, the proportions receiving formal help and mixed help rose steadily, as did the proportion of those residing in a nursing home. These findings are based on examination of data from the nationally representative Longitudinal Study of Aging (LSOA) from 1984 to 1990 to describe the changes adults 70 years of age and older experience in their caregiving arrangements over time.

More than half of nursing home residents have behavior problems

Some nursing home residents have behavior problems. They cry for long spells, get lost or wander the halls, yell at or physically hurt others, steal things, or have hallucinations or delusions. In 1987, over half (54 percent) of U.S. nursing home residents had some type of behavior problem. Certain mental and physical disabilities, as well as personal characteristics, increase the likelihood that a nursing home resident will have behavioral problems, concludes William D. Spector, Ph.D., of the Center for Organization and Delivery Studies, Agency for Health Care Policy and Research.

Dr. Spector and his colleagues Mary E. Jackson, Ph.D., of the MEDSTAT Group, and Peter V. Rabins, M.D., of Johns Hopkins University, used the Institutional Population Component of the 1987 National Medical Expenditure Survey (NMES)—a national sample of the institutionalized elderly—or the study. They analyzed the data to identify which factors correlated with four types of behaviors: delusions/hallucinations, aggressive behaviors, collecting behaviors (hoarding and stealing), and wandering or inability to avoid dangers.

Cognitive impairment, dependencies in activities of daily living (ADLs; e.g., eating, dressing, bathing, using the toilet, etc.), incontinence, a history of psychiatric problems, difficulty understanding oral communications, and vision problems were associated with a higher likelihood of exhibiting disruptive behaviors. Being unable to walk decreased the likelihood of behavior problems, presumably because of increased isolation. The patterns of association varied by type of disruptive behavior. Men were found to be more at risk for aggressive behaviors than women, who were more at risk for delusions and hallucinations. Generally, cognitive impairment, schizophrenia, and other psychoses had a greater impact for almost all behavior types. Impairment in receptive communication (understanding what is said) was a risk factor for all behaviors but had a more moderate impact. Incontinence and ADL dependencies, in general, had less impact and affected fewer behaviors.

For details, see "Risk of behavior problems among nursing home residents in the United States," by Drs. Jackson, Spector, and Rabins in the November 1997 Journal of Aging and Health 9(4), pp. 451-472. Reprints (AHCPR Publication No. 98-R011) are available from the AHCPR Publications Clearinghouse.

Return to Contents


Interactive computer game encourages adolescents to delay parenthood

An interactive computer program can encourage adolescents to delay parenthood, according to a study supported in part by the Agency for Health Care Policy and Research (HS07399). It shows that teenagers who used the program knew more about the costs of having a child and were significantly more apt to value contraception, view a child as adding problems to a teen's life, and want to delay parenthood. Also, 91 percent of teens said they would play the game again and would recommend it to a friend. The program was developed as part of the Harlem Hospital Adolescent Pregnancy Prevention Project.

In this paper, researchers at Columbia University School of Public Health and Harlem Hospital Center use the computer program as an example of four phases of a community-based public health program model for research: preresearch, research, transition phases, and full institutionalization. They point out that coalition-building was critical in the preresearch phase to share responsibility for conceptualizing and implementing the project. In this case, the team included physicians, social workers, labor and school representatives, librarians, and emergency medical services personnel. Workshops covered themes selected by the adolescents: education and careers, sexuality, AIDS, and teen advocacy. Over a 2-year period, 24 different groups participated in interactive presentations engaging 250 adolescents. The researchers distributed a flier and offered the program in the outpatient clinic and on the adolescent ward of Harlem Hospital.

Acceptance by teens was crucial and set the stage for the transition phase to making the program a long-term community resource. The researchers sought continued funding and staffing from institutions to embed the project within existing programs, and they maintained community linkages. Also critical to the transition phase was the extensive outreach effort conducted as part of the project, including two major conferences held at a school and another at the hospital, which featured the computer game but also addressed a whole range of adolescent life issues. As a result of the success of these events, an alternative high school for pregnant teens became an additional research site.

Details are in "A community research model: A challenge to public health," by Pamela Brown-Peterside, Ph.D., and Danielle Laraque, M.D., in the September 1997 American Journal of Public Health 87(9), pp. 1563-1564.

Return to Contents

Clinical Decisionmaking

Use of hormone replacement therapy among women without menopausal symptoms has doubled, but still remains low

Although menopausal symptoms such as hot flashes and headaches are the most common clinical indication for hormone replacement therapy (HRT), there is growing awareness of the ability of HRT to reduce the risk of osteoporosis and heart disease in postmenopausal women. The number of postmenopausal women without menopausal symptoms who reported use of HRT doubled from 3 percent in 1989 to nearly 7 percent in 1993 and 1994. HRT use rates for women with symptoms stayed about the same (40 percent to 38 percent).

Use rates varied by physician specialty, patient characteristics, and region. Primary care physicians other than obstetrician/gynecologists (Ob-Gyns) continued to prescribe HRT at a low rate (4 to 7 percent of visits by postmenopausal women, compared with 23 percent of similar visits to Ob-Gyns), according to a study supported by the Agency for Health Care Policy and Research (HS07892). These low use rates and substantial HRT practice variations suggest missed opportunities to improve the health of patients, conclude the Harvard Medical School researchers who conducted the study.

The investigators analyzed a nationally representative sample of 6,341 office visits by women 40 years of age and older to primary care physicians using the 1993 and 1994 National Ambulatory Medical Care Surveys to identify independent predictors of estrogen use and evaluate time trends from 1989 through 1994. Nearly 5 percent of women visiting the doctor in 1989 and 1990 reported use of HRT; this figure grew to 8 percent in 1993 and 1994. Although this increase in the use of HRT was confined to asymptomatic women, menopausal symptoms continued to be the primary determinant of HRT use. In 1993-1994, women with menopausal symptoms were six times more likely than asymptomatic women to receive HRT.

Physicians did not seem inclined to prescribe HRT to prevent heart disease, but did seem inclined to prescribe it to prevent complications from osteoporosis or hyperlipidemia (high blood fat levels). Fewer or a similar number of women with risk factors for heart disease such as hypertension, a smoking habit, obesity, and diabetes mellitus, or with known coronary heart disease used HRT compared with women who did not have these risk factors. However, more women with osteoporosis or hyperlipidemia used HRT than women without these conditions (10 percent vs. 8 percent and 11 percent vs. 8 percent, respectively).

Details are in "Low rates of hormone replacement in visits to United States primary care physicians," by Randall S. Stafford, M.D., Ph.D., Demet Saglam, M.A., Nancyanne Causino, Ed.D., and David Blumenthal, M.D., M.P.P., in the American Journal of Obstetrics and Gynecology 177(2), pp. 381-387, 1997.

Head injury patients who develop late seizures are candidates for anticonvulsant medicine

People who suffer severe head injury commonly have posttraumatic seizures that can occur shortly after the injury to months or even years later. When late seizures develop (i.e., 7 days after the injury), the probability of having another seizure is high, finds a recent study supported in part by the Agency for Health Care Policy and Research (HS06497). Patients should be treated aggressively with anticonvulsant medication after an initial unprovoked late seizure, recommends Sureyya S. Dikmen, Ph.D.

Dr. Dikmen and her University of Washington colleagues conducted a longitudinal study of 63 moderately to severely head-injured adults who developed late posttraumatic seizures. The subjects suffered the seizures while participating in a randomized, placebo-controlled study of the effectiveness of phenytoin to prevent posttraumatic seizures. The researchers measured the time from the first unprovoked late seizure to time of seizure recurrence and found that 86 percent of patients experienced more late seizures within 2 years. Of the patients studied, there was a 47 percent chance of having another unprovoked seizure within 1 month of the first late seizure. The incidence of recurrence was 69 percent by 6 months, 82 percent by 1 year, and 86 percent by 2 years. The frequency of recurrent seizures varied considerably: 52 percent of patients experienced at least 5 late seizures and 37 percent had 10 or more late seizures within the 2-year period.

The risk of having another seizure after the initial late seizure was significantly greater for persons whose head injuries were characterized by a depressed skull fracture or an acute subdural hematoma (accumulation of blood next to the brain). The risk of having five or more late seizures was greatest for those having prolonged coma (longer than 7 days). Most patients in this study suffered from closed head injuries and were treated with anticonvulsant medication after their first late seizure. Without medication, the seizures might have recurred earlier and been more frequent and severe, note the researchers. They point out the need to study newer anticonvulsant drugs for efficacy in controlling recurrent seizures in this high-risk population.

For more information, see "Risk of seizure recurrence after the first late posttraumatic seizure," by Alan M. Haltiner, Ph.D., Nancy R. Temkin, Ph.D., and Dr. Dikmen, in the Archives of Physical Medicine and Rehabilitation 78, pp. 835-840, 1997.

Sociodemographic and clinical factors affect use of diagnostic imaging tests for patients with back pain

One of the major costs associated with the $24 billion in health care expenditures in 1990 for low back problems is use of expensive diagnostic imaging procedures such as magnetic resonance imaging (MRI), noncontrast computed tomography (CT), and CT-myelography. It has been suggested that these imaging tests are often used inappropriately for patients who do not have clinical symptoms to warrant their use.

A recent study compared the actual use of diagnostic imaging tests for persistent low back pain with the recommendations set forth in the clinical practice guideline on low back pain, which was developed with support from the Agency for Health Care Policy and Research and released by AHCPR in 1994. The study, conducted by Stacey J. Ackerman, M.S.E., Ph.D., and colleagues at Johns Hopkins University, analyzed use of diagnostic imaging examinations in 2,374 patients with persistent low back problems who were enrolled in the National Low Back Pain Study from 1986 to 1991.

Patients who underwent MRI were distinguished from those who received only lumbosacral spine x-rays by more physician visits in the preceding 12 months, more functional impairment, presence of sciatica, and presence of neurologic signs/symptoms suggestive of nerve root compromise, as well as higher socioeconomic status. Those who underwent MRI had suspected soft tissue involvement; suspected structural involvement characterized those who received noncontrast CT. These findings represent the first indication that actual practice may be reasonably close to the guideline recommendations, according to the researchers.

Persons who underwent both MRI and CT-myelography were distinguished from those who received only CT-myelography only by nonclinical factors, such as higher annual household income, disability compensation, and male sex. Patients eligible for disability compensation may have complained of more severe or persistent symptoms, simply had coverage for two advanced imaging tests, or needed both tests to obtain disability compensation. These findings suggest that nonclinical factors, such as socioeconomic characteristics and patient preferences, play a substantial role in the diagnostic imaging decision. Nevertheless, the use of tests in this study closely followed clinical indications for their use recommended by the AHCPR-supported guideline.

See "Patient characteristics associated with diagnostic imaging evaluation of persistent low back problems," by Dr. Ackerman, Earl P. Steinberg, M.D., M.P.P., R. Nick Bryan, M.D., Ph.D., and others, in Spine 22(14), pp. 1634-1641, 1997.

Return to Contents
Proceed to Next Article

The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care