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

Feature Story

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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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.

Adjusting capitation rates may not prevent discrimination against chronically ill children

Some Medicaid and other health insurance programs use capitation payment rates to pay managed care plans and providers on a predetermined per capita basis. If a plan or provider is paid the same rate for a patient with chronic health problems as for a generally healthy patient, however, there is a financial incentive to discriminate against patients with more serious illness, so-called selection bias.

Capitation or risk-adjustment methods, which adjust capitated rates to compensate plans for increased costs associated with chronic medical problems, are one approach to eliminating the incentive to discriminate against children with chronic conditions. Patient sociodemographic characteristics, functional health status, clinical indicators, self-reported health status, and prior health system use are all factors that have been used to adjust capitation rates.

However, children with chronic health problems—such as asthma and diabetes—would probably remain at risk for discrimination in a competitive health care market even under risk-adjusted rates, according to a study conducted by Elizabeth J. Fowler, Ph.D., of HealthSystem Minnesota, and Gerard F. Anderson, Ph.D., of Johns Hopkins University. Regardless of the capitation adjustment method used, this study found significant underpayment to providers of care to high-risk children.

Drs. Fowler and Anderson tested current claims-based risk-adjustment methods on two pediatric populations based on data from 1990 and 1991. They tested five models at the individual and group level, using both randomly selected and nonrandom groups: a demographic model, ambulatory care groups, ambulatory diagnostic groups, diagnostic cost groups, and payment amounts for capitated systems. The researchers employed health care use and expenditure data for children enrolled in the Maryland Medicaid program and a private nonprofit health maintenance organization in Minnesota. One of the nonrandom groups represented children with chronic conditions.

The findings of this study indicate that the disincentive to enroll children with costly conditions and special health care needs would not be eliminated by any of the capitation methods tested. Although each of the alternative methods offers an improvement over a demographic model, significant underpayment remained for high-risk children. The authors call for further research in the area of pediatric capitation adjustment methods to address the limitations associated with current alternatives.

This study was supported by the Agency for Health Care Policy and Research (HS08441). For more information, see "Capitation adjustment for pediatric populations," by Drs. Fowler and Anderson, which appears in the July 1996 issue of Pediatrics 98(1), pp. 10-17.

Return to Contents

Hospital Use/Quality of Care

Researchers find quality of care for asthma affects pediatric hospitalization rates in three Northeastern communities

Pediatric hospitalizations account for almost 50 percent of total pediatric health care expenditures. Many of these hospitalizations—for example, for conditions such as asthma—can be prevented by good primary care, including the use of effective preventive therapies such as medications. The following two studies—conducted by researchers from Children's Hospital, Boston, Massachusetts General Hospital, Harvard University, Yale University, and the University of Rochester—examined the effects of quality of care on hospitalization rates for childhood asthma in Boston, MA, Rochester, NY, and New Haven, CT; and the effects of primary care involvement before and during a hospital episode for children in the same three communities.

The following studies were supported by the Agency for Health Care Policy and Research (HS06060).

Homer, C.J., Szilagyi, P., Rodewald, L., and others (1996, July). "Does quality of care affect rates of hospitalization for childhood asthma?" Pediatrics 98(1), pp. 18-23.

Use of effective medications may prevent many children with asthma, even children who have severe asthma, from being hospitalized for the condition, according to these researchers. Use of inhaled rather than oral bronchodilators (beta-agonists that open up constricted airways to improve oxygen flow), long-term anti-inflammatory therapy (inhaled cromolyn sodium or corticosteroids) for children with moderate to severe disease, and short-term anti-inflammatory therapy as soon as an acute episode becomes worse frequently prevents a child's asthma from progressing to the point at which hospitalization becomes necessary.

The researchers reviewed the medical charts of a random sample of 614 children (2 to 12 years of age) diagnosed with asthma, pneumonia, or bronchitis, who were discharged from hospitals in Boston, Rochester, and New Haven from 1988 through 1990. They found that Boston had the highest rate of hospitalizations for pediatric asthma and multiple deficiencies in asthma care compared with the other communities. Only 11 percent of Boston children had received maintenance anti-inflammatory therapy with either cromolyn or inhaled corticosteroids during the month prior to hospitalization, compared with 19 percent of New Haven and 33 percent of Rochester children. Only 30 percent of Boston children had received inhaled beta-agonist therapy during the month before admission compared with 51 percent of New Haven and 58 percent of Rochester children. About 5 percent of Boston children had received oral corticosteroids for acute asthma exacerbations, while 20 percent of New Haven and 14 percent of Rochester children had been treated with this therapy. Differences in patient characteristics, illness, and source of care did not fully explain these differences, according to the authors.

Boston pediatric patients, who were more often cared for at health centers, were less likely to be cared for by their primary care doctor while they were hospitalized. This finding suggests that a problem in integration of community-based primary care and tertiary hospital care may hinder communication between specialists and generalist physicians and interfere with the rapid introduction of new asthma treatments at the community level, conclude the researchers.

Perrin, J.M., Greenspan, P., Bloom, S.R., and others (1996). "Primary care involvement among hospitalized children." Archives of Pediatric and Adolescent Medicine 150, pp. 479-486.

This study found a lower rate of pediatric hospitalizations in communities in which primary care physicians are more involved in the care of children before and during hospitalization. In contrast to physicians who may not have seen the child prior to hospitalization, the primary care physicians often had examined the child prior to admission, referred the child to the emergency department, and served as the child's attending physician during the hospitalization.

The researchers reviewed the medical records of children admitted to all hospitals in Boston, New Haven, and Rochester during 1988 through 1990 for asthma, abdominal pain, meningitis, toxic ingestions, and head injury. The rate of hospitalization was nearly three times higher for children in Boston than children in Rochester (16 vs. 5.5 admissions per 1,000 children per year for the specified conditions) and twice as high as the rate for children in New Haven (16 vs. 8.4 admissions per 1,000). Children in Rochester had the highest likelihood of previous medical visits (59 percent vs. about 49 percent in Boston and New Haven), and once hospitalized, 81.5 percent of children in Rochester were attended by a familiar physician, compared with 50.2 percent for children in New Haven and 35.1 percent for children in Boston.

However, sources of care performed differently in each city. Private physicians in Boston were much less likely to be involved in the children's care than those in Rochester. Neighborhood health centers in New Haven had substantially closer ties to their patients during the admissions process than did Boston or Rochester health centers.

These findings indicate higher rates of hospitalization in communities with less primary care attachment and suggest that restructuring the components of primary care might diminish reliance on hospital care. However, the substantial variation by source of care and city suggests that different strategies may be needed in different settings to enhance primary care, conclude the authors.

Computer reminders fail to improve preventive care for hospitalized patients

Computer reminders do not improve preventive care—such as immunization against influenza and cholesterol screening—during hospitalization, according to a study supported by the Agency for Health Care Policy and Research (HS05626 and HS07719). These reminders have improved preventive care in outpatient settings. However, the fact that physicians providing care to hospitalized patients often are not their primary care physicians proved to be an important barrier to preventive care in the hospital, according to researchers at the Indiana University School of Medicine, the Regenstrief Institute for Health Care, and the VA Medical Center, Indianapolis.

The researchers analyzed physicians' responses to computer reminders on general medicine wards of a university-affiliated public hospital during 6 months in 1992. They randomized 12 rotating teams of physicians and medical students to receive the guideline-based, computer-generated reminders while 12 teams did not (a total of 78 house staff). Reminders about preventive care for which individual patients were eligible were printed on daily rounds reports, and suggested orders for preventive care were provided through the physicians' workstations.

Doctors who received the reminders failed to comply with preventive care guidelines at the same rate as doctors who did not receive the reminders (23 vs. 24 percent). This was despite the fact that most physicians were in favor of providing most kinds of preventive care for hospitalized patients.

About half of the staff felt that patients would prefer long-term preventive care treatments (for example, aspirin to prevent heart attack) to be started by their primary care physician in the clinic, and 60 percent thought that preventive care was best left to the primary care physician. On the other hand, 62 percent of physicians said they would like other physicians to provide their hospitalized patients with preventive care. Perhaps the most important barrier is embedded in incentives faced by physician teams to shorten stays and lower costs, suggest the researchers. Thus, providing preventive care for patients who are not their own carried only risk of more work and more responsibility.

Requiring physicians to respond to recommendations for preventive care (for example, saying why a recommendation is inappropriate for a particular patient) or requiring them to treat these care measures as "standing orders," unless the physician explicitly countermands them, might increase preventive care for hospitalized patients, conclude the researchers.

See "Computer reminders to implement preventive care guidelines for hospitalized patients," by J. Marc Overhage, M.D., Ph.D., William M. Tierney, M.D., and Clement J. McDonald, M.D., in the July 22, 1996, Archives of Internal Medicine 156, pp. 1551-1556.

Emergency department clinicians often experience conflicted relationships with patients who are "regulars"

Clinicians who practice in the emergency department (ED) setting often have ambivalent feelings about caring for heavy ED users, who may visit the ED as often as every week. These patients commonly have chronic, relatively intractable medical problems, including mental illness and substance abuse, which may be compounded by social problems such as homelessness or estrangement from families. Such characteristics—when encountered in an environment of urgency such as the ED—can contribute to feelings of futility and withdrawal on the part of ED clinicians from the patients they know best. A study supported in part by the Agency for Health Care Policy and Research (HS08412) suggests that clinicians need to distinguish between care and control in their expectations for and understanding of these patients.

The ED environment pushes clinicians toward a focus on control rather than a focus on care, explains Ruth E. Malone, R.N., Ph.D., CEN, of the University of California, San Francisco, author of the study. This is due to the legitimate need to treat patients quickly and clear space for new arrivals, but the kinds of chronic problems most of these patients experience are not amenable to the "quick fix" approach offered in the ED. Clinicians may feel they are doing nothing for such patients except treating symptoms.

However, observations and interviews at two inner-city trauma center EDs suggest that, for some patients who utilize these facilities most heavily, EDs are places to seek not only medical care but also reassurance of inclusion in the human community—often in the absence of other safe places in which to seek such recognition. Basic caring practices such as recognizing the patient on sight, inquiring about family, or referring to the patient by a familiar nickname were meaningful to patients, who frequently alluded to being "known" in the ED setting. Interviews with these "frequent flyer" patients suggest that such basic recognition practices might have more value in establishing trust and encouraging medically indicated behavior changes (such as reductions in substance use) than the typical practice of rapid medical stabilization and discharge with instructions to "stop."

See "Almost 'like family': emergency nurses and 'frequent flyers'," by Dr. Malone, which appears in the June 1996 Journal of Emergency Nursing 22(3), pp. 176-183.

Health maintenance organizations typically require patients to contact their primary care physician first for any new medical problem, with that physician deciding whether referral to a specialist is necessary. This first-contact approach to health care can save more than half of outpatient health care expenditures, conclude Christopher B. Forrest, M.D., Ph.D., and Barbara Starfield, M.D., M.P.H., of the Johns Hopkins School of Public Health. Their study, supported by the Agency for Health Care Policy and Research (National Research Service Award fellowship F32 HS00070 and training grant T32 HS00029), used data from the 1987 National Medical Expenditure Survey of a sample of noninstitutionalized persons in the United States to assess the impact of first-contact care on outpatient health expenditures.

Return to Contents

Health Care Costs and Financing

Costs of outpatient care are cut in half when a patient sees a primary care physician first

Analysis showed that nearly half (49 percent) of all episodes of care for 24 acute (not chronic) conditions—ranging from respiratory and ear infections to preventive care, sprains, and burns—began with a visit to the primary care clinician. This first-contact use of the primary care clinician was significantly associated with a 53 percent reduction in expenditures for all types of episodes except fatigue ($63 vs. $134), 62 percent for acute illness episodes ($62 vs. $164), and 20 percent for preventive care episodes ($64 vs. $80). Acute-care episodes that began with a visit to the emergency room were about four times more expensive than those that began with a visit to a primary care clinician ($255 vs. $63).

Increasing the current level of first-contact care just 10 percent would save at least $1.1 billion per year (1993 dollars) for the U.S. health care system, according to the researchers. They attribute the bulk of the cost savings of first-contact care to lower payments for primary care physician services, less resource-intensive practice styles of generalists compared with specialists, lower severity of illness, and lower reimbursement for established rather than new patients.

Details are in "The effect of first-contact care with primary care clinicians on ambulatory health care expenditures," by Drs. Forrest and Starfield, in the July 1996 issue of The Journal of Family Practice 43(1), pp. 40-48.

Nonclinical factors may determine whether depressed patients receive specialty care

Prepaid health insurance plans tend to rely less on psychiatrists to treat depressed patients and more on nonphysician mental health specialists than traditional fee-for-service (FFS) plans. The sicker a patient is, the more likely he or she will receive specialty care in both systems, but income, patient education, and ethnicity also influence the likelihood of seeing a psychiatrist, according to a study supported in part by the Agency for Health Care Policy and Research (HS06802).

Psychiatrists are more costly than generalists, but they typically provide more appropriate antidepressant medication and counseling than general physicians. This specialty care could be better targeted to patients most likely to benefit from it in both systems, notes Kenneth B. Wells, M.D., M.P.H., of RAND and the University of California, Los Angeles, the study's principal investigator. Dr. Wells and colleagues Roland Sturm, Ph.D., and Lisa S. Meredith, Ph.D., also of RAND, analyzed data on adult patients with depression and other health problems in alternative systems of care in three cities.

The researchers found that only 10 percent of depressed patients in prepaid plans considered a psychiatrist to be their main source of care, less than half the rate found among FFS patients (22 percent). And, prepaid patients were more likely to receive care from a nonphysician mental-health specialist or therapist than FFS patients (20 percent vs. 15 percent). In both payment systems, the majority of depressed patients received care from a general medical provider (63 percent in FFS, 70 percent in prepaid plans).

Worse psychological health increased the probability of obtaining care from a mental health specialist and worse physical health increased the probability of obtaining care from a general medical provider in both systems. Depressed individuals with higher incomes and better education were more likely to have a mental health specialist as their main provider. However, these factors were less predictive of specialty care than worse psychological health, regardless of payment system. Minorities were significantly less likely to see either type of mental health specialist in both payment systems.

Drs. Sturm, Meredith, and Wells point out that continuity of care may be particularly important for patients with depression because depression can be chronic and recurrent. They found that the duration of a patient-provider relationship is significantly shorter in prepaid than in fee-for-service plans for patients of psychiatrists and primary care physicians. In both types of plans, the duration of the patient-provider relationship was shorter for nonphysician therapists than for psychiatrists or primary care physicians.

The authors explored whether the end of a patient-provider relationship could have implications for quality of care because many patients in this sample were chronically depressed and could have benefited from long-term treatment. Focusing on patients receiving effective antidepressant medication at baseline, they found that the end of a patient-provider relationship is usually associated with discontinuing antidepressant medication. The authors conclude, therefore, that careful consideration of how new health care policies affect patient-provider relationships over time will be especially important.

More details are in "Provider choice and continuity for the treatment of depression," by Drs. Sturm, Meredith, and Wells, in Medical Care 34(7), pp. 723-734.

Recognizing depression in primary care can avoid unnecessary tests and referrals and save money

Missed depression diagnoses can lead to unnecessary and costly tests to uncover the cause of vague physical complaints such as headache and fatigue, which frequently mask depression. Physicians in training often do not recognize when a patient is depressed, according to a recent study. Improving physicians' recognition of depression could improve quality of care for these patients and save health care costs, conclude researchers at the University of California at Davis.

In a study supported by the Agency for Health Care Policy and Research (HS06167 and HS08029), Edward J. Callahan, Ph.D., and Klea D. Bertakis, M.D., M.P.H., and their colleagues found that when physicians did recognize depression, they treated patients differently. They spent less time chatting and taking a physical history and more time on counseling, as well as more overall time with the patient. Physicians who did not recognize depression spent significantly more time taking medical histories, perhaps to clarify confusing data in an effort to explain the patients' symptoms. This additional time may mark the beginning of expensive efforts to diagnose patients using more laboratory tests, more return visits, and more referrals to subspecialists for diagnostic testing, all of which increase depression-related cost of care, according to the researchers.

They videotaped the physician visits of 508 new adult patients randomly assigned to 105 third-year resident primary care providers. The patients had previously taken the Beck Depression Inventory (BDI), a 13-item questionnaire to detect depression, but results were not revealed to their physicians. The physicians identified only 15 percent of patients as depressed, while 26 percent of patients had a BDI score of 9 or greater indicating moderate to severe depression by this test. Whether the patients actually had depression was not confirmed.

More details are in "The influence of depression on physician-patient interaction in primary care," by Drs. Callahan and Bertakis, Rahman Azari, Ph.D., and others, in the May 1996 issue of Family Medicine 28(5), pp. 346-351.

High social costs found for home-based care of ventilator-assisted individuals

Pressure to contain costs is prompting hospitals to discharge a growing number of ventilator-assisted individuals (VAIs) to be cared for at home. However, studies that demonstrate the cost-effectiveness of home care do not take into consideration many of the economic burdens assumed by the families of VAIs who return home. A recent study, supported in part by the Agency for Health Care Policy and Research, examined the extent to which home care resulted in a shift of costs from the hospital to the family, rather than a true reduction in the total cost of care. This study, one of the first to measure the social costs associated with home-based care for VAIs, was conducted by Mary Ann Sevick, Sc.D., R.N., of Bowman-Gray School of Medicine at Wake Forest University (National Research Service Award fellowship F32 HS00054).

Dr. Sevick and her colleagues surveyed 277 primary family caregivers of VAIs residing in 37 States. They found that family members cared for VAIs an average of 12 hours per day, had done so for an average of 8 years, and reported a loss of 4 to 5 hours of daily leisure since becoming a caregiver. About half of respondents adjusted their work schedules to accommodate their caregiving activities. Of those adjusting their employment, nearly 50 percent stopped working, 33.6 percent decreased their working hours, 14.5 percent changed jobs, and 2.7 percent increased their working hours. These employment changes resulted in a median loss of $400 in monthly earnings.

The total direct and indirect costs of home care for health personnel, equipment rental, oxygen, medications, supplies, ambulance transport, cost of one-time purchases and/or home remodeling, physician and hospital costs, and lost wages averaged $7,642 to $8,596 per month. This is significantly lower than the estimated monthly cost of long-term-care facility placement found in other studies, which ranges from $13,578 to $27,133 (in 1995 dollars). However, as Dr. Sevick points out, the figures derived in this study do not take into consideration intangible costs such as the value of the caregivers' lost leisure time and the impact of caregiving on quality of life.

Previous studies have found that caregivers do not perceive their experience to be a negative one, and many are unwilling to consider alternative living arrangements for their family member. The development of policies and programs to support the growing number of VAIs being discharged home requires additional investigation regarding the cost-effectiveness of home placement, as well as the ability and/or willingness of families to accept the cost and responsibility of home care, conclude the researchers.

For more information, see "Economic cost of home-based care for ventilator-assisted individuals," by Dr. Sevick, Mark S. Kamlet, Ph.D., Leslie A. Hoffman, Ph.D., R.N., and Ian Rawson, Ph.D., in the June 1996 issue of Chest 109(6), pp. 1597-1606.

Return to Contents

Patient Outcomes/Effectiveness Research

Increased risk of preterm birth among pregnant black women may be due in part to a higher rate of vaginal infections

Black women have two to three times more preterm births than white women, and the gap appears to be increasing. A recent study suggests that pregnant black women are far more likely to have vaginal infections that have been associated with preterm births than other women.

This multivariate analysis, conducted by Robert L. Goldenberg, M.D., of the University of Alabama at Birmingham, principal investigator of the Low Birthweight Patient Outcomes Research Team (PORT), and colleagues, was supported in part by the Agency for Health Care Policy and Research (PORT contract 282-92-0055). Data for the analysis were derived from a computerized database containing the results of vaginal cultures for various types of bacteria in 13,747 pregnant women (23 to 26 weeks' gestation) of low socioeconomic status from various ethnic groups at seven urban medical centers from 1984 to 1989.

Results showed that black women were two to six times more likely than white women to have vaginal infections: Chlamydia trachomatis (16 percent vs. 5 percent), Neisseria gonorrhoeae (2.5 percent vs. 0.4 percent), Bacteroides sp. (25 percent vs. 14 percent), and bacterial vaginosis (23 percent vs. 9 percent), as well as four other organisms. Infection rates in Hispanic women were higher than in white and Asian-Pacific Islander women but lower than in black women. Asian-Pacific Islander women had the lowest rates of vaginal infections.

Differences in socioeconomic and health status, medical conditions, health behaviors (for example, number of sexual partners or smoking or alcohol use during pregnancy), and psychological characteristics do not explain the great disparity in pregnancy outcomes between black and other women, according to the authors.

For the various organisms evaluated in this study, when black and white women of similar incomes are compared, black women still have a two-fold greater preterm delivery rate, according to Dr. Goldenberg. The authors also note that Hispanic women—especially those of Mexican origin—have one of the lowest rates of low birthweight, despite average income levels that are lower than many other ethnic groups. The presence of bacterial vaginosis in more than 20 percent of the black women studied suggests a significant association with the risk of preterm birth. Vaginal infections during pregnancy may well provide a new explanation for the disparity in preterm delivery rates between black and other women, concludes Dr. Goldenberg.

Details are in "Bacterial colonization of the vagina during pregnancy in four ethnic groups," by Dr. Goldenberg, Mark A. Klebenoff, M.D., Robert Nugent, Ph.D., and others, in the May 1996 American Journal of Obstetrics and Gynecology 174(5), pp. 1618-1621.

Return to Contents
Proceed to Next Section

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


AHRQ Advancing Excellence in Health Care