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


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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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

Both hospital and market factors affect a rural hospital's likelihood of survival or organizational change

Because of the disproportionate numbers of the elderly and poor in rural areas, rural community hospitals often find themselves in financial trouble and forced to choose between closure and a shift in core strategies away from acute inpatient care. Conversion to an organization that provides nonacute health care service (e.g., a primary care clinic, long-term care facility, or specialty hospital) is more apt to occur than closure when the population's demand for health care and ability to pay for it are high, competition from other hospitals is substantial, and hospitals have established strategies to provide alternative forms of health care, according to a study supported by the Agency for Health Care Policy and Research (HS07047).

Conversion can represent a viable alternative for hospitals threatened with closure, one that may allow rural communities to retain access to essential health services, notes Jeffrey A. Alexander, Ph.D., of the University of Michigan, lead author of the study. Dr. Alexander and his colleagues studied 2,780 rural community hospitals at risk for conversion from 1984 to 1991 as well as hospitals that emerged during the study period as newly formed organizations that may have resulted from mergers and consolidations of rural hospitals. They used several databases to identify conversions and closures, hospital organizational characteristics, and market variables.

Analysis showed that the closeness of a rural hospital to another hospital (competition), poor hospital performance (for example, less cash flow or fewer admissions), and smaller hospital size put the hospital at risk for both closure and conversion. Membership in a multihospital system increased a hospital's likelihood of conversion, while investor-owned control increased a hospitals likelihood of closure.

For every 10,000 residents or $1,000 increase in per capita income in a county's population, conversion was 22 percent more likely than closure. Conversion was 28 percent more likely than closure with each 10 percent increase in the proportion of a hospital's nonacute care specialization. For each additional mile between the focal hospital and its closest neighboring provider (indicating decreased competition), conversion was 6 percent more likely than closure.

For more information, see "Determinants of profound organizational change: Choice of conversion or closure among rural hospitals," by Dr. Alexander, Thomas A. D'Aunno, Ph.D., and Melissa J. Succi, in the Journal of Health and Social Behavior 37, pp. 238-251, 1996.

ICUs vary substantially in their use of do-not-resuscitate orders

Whether intensive care unit (ICU) patients receive do-not-resuscitate (DNR) orders usually depends on their physicians assessment of their likelihood of survival and family preferences. It also depends on the ICU where the patients are treated, according to a study supported in part by the Agency for Health Care Policy and Research (HS05787).

The study, which involved 17,440 patients admitted to 42 ICUs from 1988 to 1990, showed that 9 percent of patients admitted to ICUs received DNR orders during their first week in the ICU. However, the range of DNR orders written at various ICUs extended from 1.5 to 22 percent of ICU patients. DNR orders were written significantly less frequently than was predicted by the patients severity of illness in five ICUs and more frequently than predicted in three of the ICUs studied.

Patient characteristics most associated with DNR orders were severity of physiologic abnormalities (explaining 46 percent of DNR orders), age (18 percent), and ICU admission diagnosis (18 percent). Nearly half of ICU variation in DNR orders was due to clinically relevant patient characteristics. But the remaining variation may have resulted from unmeasured patient differences such as treatment preferences, religious affiliation, and educational level, or to physician or ICU practice style, according to researchers from the George Washington University and the University of Virginia who authored the study.

For example, site visits to nine ICUs showed substantial variations in ICU adherence to standards recommending DNR protocols and provisions for family support. At some ICUs, physicians lacked knowledge about providing comfort care or terminal weaning (versus "futile" aggressive care). There was no DNR protocol or evidence of medical leadership in this area, and physicians often were unable to give clear DNR instructions to the nursing staff. This often resulted in confused and upset families and staff. In contrast, other ICUs promoted family meetings where prognosis and plans were clearly outlined to help families prepare for what was to come next, provided family support via social workers or ministers, and had well-established protocols for writing DNR orders. Frequency of DNR orders was not correlated with ICU adherence to published DNR standards.

See "Variations in the use of do-not-resuscitate orders in ICUs," by Robert L. Jayes, M.D., Jack Zimmerman, M.D., Douglas P. Wagner, Ph.D., and William A. Knaus, M.D., in Chest 110(5), pp. 1332-1339, 1996.

ICU studies focus on triage of low-risk patients and outcomes for patients with organ system failure

The following two studies, led by researchers from the George Washington University and supported in part by the Agency for Health Care Policy and Research (HS05787 and HS07137) examine care for intensive care unit (ICU) patients. The first study describes the type and amount of patient services needed to care for low-risk patients transferred from an ICU to an intermediate or step-down unit. The second study compares outcomes for patients with one or more organ system failures treated in an ICU between 1979 and 1982 with similar patients treated between 1988 and 1990.

Zimmerman, J.E., Wagner, D.P., Sun, X., and others (1996). "Planning patient services for intermediate care units: Insights based on care for intensive care unit low-risk monitor admissions." Critical Care Medicine 24(10), pp. 1626-1632.

ICU patients who have less than a 10 percent chance of needing life-support therapies can be safely admitted to an intermediate care unit as long as they receive concentrated nursing care (ratio of one nurse to three or four patients) and some technologic monitoring, according to this study. Using intermediate care units in this manner can promote earlier ICU discharge, facilitate patient triage, decrease costs, improve ICU use, avoid ICU readmission, and decrease the mortality rate on hospital wards, according to the researchers. They obtained demographic, physiologic, and treatment information on 8,040 ICU patients, who were admitted to ICUs at 40 U.S. hospitals for monitoring but received no active life-support treatment on their first day in the ICU. The investigators identified 6,180 ICU patients at low risk for receiving active treatment in the ICU. They used the daily Therapeutic Intervention Scoring System (TISS) to record patient care services during ICU days 1 to 7 to project the types of equipment and services and level of nursing care that would be required for patient care in an intermediate care unit. The most frequent technologic monitoring service for all low-risk ICU patients was electrocardiogram monitoring (nearly 100 percent), arterial catheter or blood pressure monitoring (51 percent), and pulse oximetry (34 percent). Concentrated nursing care accounted for 89 percent and technologic monitoring for 11 percent of the interventions required for low-risk patients during their first day in the ICU. Subsequent active life-support therapy was received by about 4 percent of these ICU low-risk patients. Based on these findings, the researchers estimated that low-risk surgical and medical patients would require slightly more than a 1:3 and 1:4 nurse/patient ratio, respectively, if transferred to an intermediate care unit, as well as limited technologic monitoring.

Zimmerman, J.E., Knaus, W.A., Wagner, D.P., and others (1996). "A comparison of risks and outcomes for patients with organ system failure: 1982-1990." Critical Care Medicine 24(10), pp. 1633-1641.

Nearly half of ICU patients suffer from organ system failure, and about 14 percent develop failure of several organ systems, which is the leading cause of death in noncoronary ICUs. Nevertheless, the severity of physiologic disturbance on the first day of organ failure is a better predictor of hospital death than the number of organ system failures, according to this study. The investigators examined the outcomes of patients treated in 60 ICUs at 53 U.S. hospitals to identify risk factors for organ system failure (by comparing ICU patients with and without organ system failure) and to compare ICU treatment outcomes over an 8-year period from 1982 to 1990. They recorded the type and number of organ system failures as well as physiologic responses, using the Acute Physiology and Chronic Health Evaluation (APACHE) score, for about 1 week of ICU treatment and followed the patients until hospital discharge. About 48 percent of ICU patients treated between 1988 and 1990 and 44 percent of those treated between 1979 and 1982 developed organ system failure; 14 percent of both groups of patients developed multiple organ system failure. ICU patients most likely to develop organ system failure were older, had preexisting severe chronic disease, and nonsurgical diagnoses, particularly sepsis and cardiac arrest. The APACHE score was higher (indicating greater severity of physiologic disturbance) on the first day of organ system failure for patients who had either single or multiple organ system failure compared with patients who did not have organ system failure. The risk of hospital death varied widely for patients with single or multiple organ system failure, depending on the severity of physiologic derangement on the first day of system failure, ICU admission diagnosis, and the combination of different types of organ system failure (reflected in the exact profile of physiologic abnormalities). For example, hospital death ranged from 20 percent for patients with hematologic and cardiovascular failure to 76 percent for cardiovascular and neurologic failures.

Return to Contents

Managed Care/Primary Care

Women receive more health care services if they have an internist as their primary care doctor

Women who have an internist as their primary care physician have more tests and receive more ancillary services and referrals to specialists than women whose primary care doctor is an obstetrician/gynecologist (OB/GYN) or family/general practitioner. Its not clear whether internists do too much (overtreat) or whether OB/GYNs and general practitioners do too little (undertreat). It also is unclear whether women who want more health care services choose internists as their primary care doctors. More research is needed to answer these questions, which are at the heart of the ongoing debate about which type of doctors should provide primary care to women, concludes Carolyn M. Clancy, M.D., Director of the Center for Outcomes and Effectiveness Research, and Acting Director of the Center for Primary Care Research, at the Agency for Health Care Policy and Research.

Dr. Clancy and her colleagues used data from the 1987 National Medical Expenditure Survey (NMES), a population-based survey of civilian noninstitutionalized individuals in the United States on their health care use, expenditures, access, and health status. The researchers estimated the independent effects of the specialty of a usual physician (primary source of care) on women's use of and expenditures for outpatient care, controlling for patient characteristics, insurance status, geographic region, and location of care. Compared with internists and OB/GYNs, family/general practitioners were most likely to serve as the usual source of care to large segments of women across all demographic and socioeconomic categories.

Women under age 65 with an internist as their usual source of care had more total visits, more ancillary service visits, and received more care outside their usual source of care, and they spent significantly more on outpatient services than women with a family/general practitioner or an OB/GYN as their usual source of care. There were no differences in outpatient use or spending between women having an OB/GYN and those having a family/general practitioner as their usual source of care.

Dr. Clancy and her colleagues point to the need for additional studies to examine interspecialty differences in various managed care models that will enhance understanding of the relationship among patient characteristics, physician training, and health care use to provide cost-effective primary care to all women.

Details are in "Cost differences among women's primary care physicians," by Barbara A. Bartman, M.D., M.P.H., Dr. Clancy, Ernest Moy, M.D., and Patricia Langenberg, Ph.D., in the Winter 1996 issue of Health Affairs 15(4), pp. 177-182. Reprints (AHCPR Publication No. 97-R048) are available from the AHCPR Publications Clearinghouse.

Geographic region and type of health plan influence length of hospital stay for childbirth

The length of time a woman stays in the hospital after delivering a baby has continually declined since 1970. The recent decline to stays as short as 24 hours has been attributed primarily to the cost-cutting initiatives of managed care plans. However, the type of managed care plan, as well as several other factors including geographic region, influence how long a mother will stay in the hospital after a normal vaginal delivery, according to Julie A. Gazmarian, Ph.D., and Jeffrey P. Koplan, M.D., of the Prudential Center for Health Care Research in Atlanta.

The researchers examined Prudential Health claims data for factors that were most strongly associated with length-of-stay for delivery: type of plan, region, and maternal age. Although the majority of newborns (62 percent) and mothers (64 percent) were discharged within 1 day after delivery, there was considerable variation by plan type. For example, 82 percent of health maintenance organization (HMO) enrollees were discharged within 1 day compared with 61 percent of point-of-service enrollees (who can choose to receive services out of network and still obtain some coverage for those services) and 48 percent of traditional fee-for-service enrollees.

Length of hospital stay also varied by region, which may reflect the level of managed care penetration in each region. Interestingly, some of the States that recently have enacted legislation mandating hospital stays of 48 hours after normal delivery are in the Northeast (New Jersey, Massachusetts, and New York), where length-of-stay is the longest and does not vary by plan type and where managed care penetration is lower than it is in other parts of the country.

Overall, there was a higher frequency of 1-day hospital stays in the West (88 percent) compared with the South (77 percent) and the North Central (64 percent) and Northeast (34 percent) regions. For the South and the North Central and Western regions, length-of-stay varied by type of plan: HMOs consistently had the highest frequency of 1-day stays, followed by point-of-service and fee-for-service plans. As noted above, there was no significant difference in length-of-stay by type of plan in the Northeast. Data from this study were presented at the Building Bridges conference, sponsored by the American Association of Health Plans and the Agency for Health Care Policy and Research, and held in March 1996 in San Diego, CA.

Details are in "Length-of-stay after delivery: Managed care versus fee-for-service," by Dr. Gazmarian and Dr. Koplan, in the Winter 1996 issue of Health Affairs 15(4), pp. 74-80.

Less than one-fourth of community hospitals participate in systems that facilitate managed care contracting

There is more talk of physician/hospital organizations (PHOs) and other integrated systems to facilitate contracting with managed care firms than actual involvement in such arrangements by hospitals, concludes a study supported by the Agency for Health Care Policy and Research (HS09183). The study found that only 23 percent of U.S. community hospitals were participating in PHOs, management services organizations (MSOs), foundations, and integrated health care organizations (IHOs). The vast majority (64 percent) of hospitals that did participate were in the relatively simple PHOs, in which neither physicians nor hospitals give up much autonomy or freedom of action. The more tightly linked integrated organizations, such as MSOs and IHOs, were much less prevalent, with only 3 percent of community hospitals belonging to an IHO in 1993.

The study was led by Michael A. Morrisey, Ph.D., of the University of Alabama at Birmingham. Dr. Morrisey and his colleagues analyzed data from the 1993 Prospective Payment Assessment Commission (ProPAC) survey of Hospital-Physician Relationships and the American Hospital Associations (AHA) Annual Survey of Hospitals. Their analysis showed that hospital participation, while quite low across the board, nevertheless varied by region, hospital size, rural/urban location, ownership, teaching role, and system affiliation. Larger, urban, teaching, and nonprofit hospitals or those owned, leased, or sponsored by a multihospital system were more apt to form these new arrangements.

Finally, hospitals with more than 15 percent of their revenues from managed care were twice as likely to participate in organizations that provide tighter linkages with physicians. Of hospitals with 15 percent or less of their revenue coming from managed care, no more than 26 percent participated in such organizations. This suggests that the dependence of providers on managed care firms as a source of patients and revenue, not the existence of a contract or the number of contracts, leads to organizational change, according to Dr. Morrisey.

As with most preliminary studies, this one raises at least as many questions as it answers. The authors point out two sets of questions in particular. First, to what extent has the local market-wide growth of managed care led to the adoption of the various forms of provider integration? More generally, what factors have led physicians and hospitals to engage in these new organizations? Second, to what extent has the adoption of these organizational structures enhanced the economic viability of their participants? Do the organizational structures per se make any difference, or is it the skills, location, and fees charged by the participants that matter?

In conclusion, the authors note that the linkages between elements of physician/hospital integration and hospital performance appear to be positive in nature but weak in magnitude. Of course, as these integrated arrangements mature, they may grow in strength. On the other hand, they may only be the most recent fad in the hospital industry, destined to be abandoned as the industry moves on.

For more details, see "Managed care and physician/hospital integration," by Dr. Morrisey, Jeffrey Alexander, Ph.D., Lawton R. Burns, Ph.D., and Victoria Johnson, Ph.D., in the Winter 1996 issue of Health Affairs 15(4), pp. 62-73.

Return to Contents

Elderly/Long-term Care


Primary care providers should ask seemingly depressed older patients about suicidal thoughts

Suicide, the ninth leading cause of death in the United States, is most prevalent among older Americans. Less than 1 percent of older primary care patients are potentially suicidal (think that life is not worth living, have a suicide plan, or have tried to commit suicide before). But 5 percent of older primary care patients who are depressed are suicidal, according to a study supported in part by the Agency for Health Care Policy and Research (HS07632 and HS07763). Often these patients are moderately to severely impaired in their ability to carry out everyday tasks. The medical problems that plague them are conditions typically seen by primary care physicians, who are in a unique position to help prevent suicide in their routine contacts with the patients, suggests Christopher M. Callahan, M.D., of the Indiana University School of Medicine and Regenstrief Institute for Health Care.

For instance, studies show that older patients suffering from medical illness are much more likely to be suicidal and to be successful in their suicide attempts than other patients. But thoughts of suicide are typically transient; that is, an older patient who is helped through the crisis may never again have thoughts of suicide. Since depressed older patients typically do not volunteer that they have suicidal thoughts and may even deny depression, primary care physicians should simply ask seemingly depressed older patients very directly if they have any suicidal thoughts, suggests Dr. Callahan.

The researchers compared the severity of depressive symptoms and functional impairment among older patients with suicidal ideation (thoughts of committing suicide) with those of older patients without suicidal ideation at an outpatient clinic of an academic primary care group practice. They compared 301 patients who screened positive for depression and a random sample of 101 patients who screened negative from among 3,767 patients aged 60 and older who completed screening for depression during routine office visits.

More details are in "Suicidal ideation among older primary care patients," by Dr. Callahan, Hugh C. Hendrie, M.B.,Ch.B., Nancy A. Nienaber, M.A., and William M. Tierney, M.D., in the Journal of the American Geriatrics Society 44, pp. 1205-1209, 1996.

Help from family members increases elderly diabetics adherence to treatment and dietary regimens

Elderly persons with adult-onset, non-insulin-dependent diabetes must follow a fairly strict regimen to maintain nearly normal blood-glucose levels. They often must take daily medication, adhere to a low-sugar, carbohydrate-restricted diet, exercise, regularly examine their feet (for sores or wounds that can become gangrenous and limb-threatening due to poor circulation typical of diabetes), get regular eye care (to detect any diabetes-related damage), and monitor blood and/or urine glucose levels each day. Elderly diabetics are much more likely to follow this regimen and to have lower blood-sugar levels if family members are involved in their daily routine and medical visits, according to a study supported in part by the Agency for Health Care Policy and Research (HS06665).

The study was conducted by the Variations in the Management and Outcomes of Diabetes Patient Outcomes Research Team (PORT), which is led by Sheldon M. Greenfield, M.D., of the New England Medical Center. It shows that from 22 percent to 55 percent of family members assisted older diabetics with diabetes-specific care, 53 percent accompanied them to doctors visits, and 36 percent actively participated in these visits by asking questions about diabetes care and helping the patient communicate with the doctor. Family members were more apt to be involved if they had a good understanding of diabetes care issues or if the patient was functionally impaired or was their spouse.

More elderly diabetics whose family members were involved in their care reported taking prescribed medications and following diabetic diets than patients with uninvolved family members, according to lead author Rebecca A. Silliman, M.D., Ph.D., of the Boston University School of Medicine. Dr. Silliman and her colleagues recommend that health care systems and physicians place more emphasis on educating older patients and family members about diabetes care issues and be supportive of them in their management of diabetes and other chronic diseases. The researchers followed the involvement of family members of diabetic patients 70 years of age or older who participated in a larger study of diabetics at three primary care practice settings.

More details are in "The care of older persons with diabetes mellitus: Families and primary care physicians," by Dr. Silliman, Shelah Bhatti, M.S., Amina Khan, M.A., and others, in the November 1996 Journal of the American Geriatrics Society 44(11), pp. 1314-1321.

Return to Contents

Special Populations

Exposure to TB and vision problems affect more than one-third of homeless persons in Los Angeles County

More than 3 million Americans are homeless. Health problems that typically affect this group are vision problems; tuberculosis (TB); hypertension; skin disorders ranging from impetigo and eczema to scabies and head lice; peripheral vascular disease of the foot and ankle evidenced by conditions such as edema and leg ulcers; and foot problems ranging from painful calluses or corns to severe athlete's foot. But little is known about the prevalence of these conditions among the homeless, who because they have no residence, are difficult to survey. And their poor access to health care facilities prevents researchers from studying their medical records.

However, an innovative study of homeless persons in Los Angeles County found that all of these conditions are prevalent among the homeless there and that more than one-third of this group had vision problems or had been exposed to TB. The study was conducted by researchers at the University of California, Los Angeles, and RAND and was supported in part by the Agency for Health Care Policy and Research (HS06696). The researchers used laypersons to conduct a structured physical exam of 366 persons in Los Angeles County, who were enrolled in the RAND Course of Homelessness Study. The RAND study was conducted in 1990 to detect the presence of typical conditions that afflict the homeless.

The lay examiners were trained to assess blood pressure, visual acuity, and skin problems and to test for TB before they conducted exams in the field office, outdoor areas, or homeless shelters. Their clinical findings and measurements were operationalized as a Physical Exam Grid (PEG) for field assessment of these conditions. The researchers conclude that laypersons can collect reliable and valid physical exam data on disadvantaged populations using the PEG, and that the PEG is a reliable instrument for assessing and monitoring the health of homeless persons.

Details are in "Homing in on the homeless: Assessing the physical health of homeless adults in Los Angeles County using an original method to obtain physical examination data in a survey," by Lawrence C. Kleinman, M.D., M.P.H., Howard Freeman, Ph.D., (deceased), Judy Perlman, M.A., and Lillian Gelberg, M.D., M.S.P.H., in the December 1996 Health Services Research 31(5), pp. 533-549.

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