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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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: https://info.ahrq.gov. Let us know the nature of the problem, the Web address of what you want, and your contact information.
Please go to www.ahrq.gov for current information.
One-fifth of California's nonprofit community hospitals do not
provide community benefits sufficient to balance their tax
Nonprofit hospitals in the United States are exempt from Federal corporate income taxes and
other taxes, amounting to annual tax subsidies of $8.5 billion nationwide. About 80 percent of
nonprofit community hospitals in California provide uncompensated care for community patients
equal to the tax subsidies they receive, but another 20 percent do not. The average shortfall per
hospital was nearly $815,000 in 1991 for a total shortfall of more than $40 million that year,
according to a study supported by the Agency for Health Care Policy and Research (HS07266).
And, most hospitals that do not provide community benefits (i.e., charity care and forgiving bad
debts) sufficient to balance their tax subsidies tend not to do so consistently year after year.
It may be time to revisit the tax codes and define standards for tax exemption more narrowly and
explicitly, conclude Michael A. Morrisey, Ph.D., and Mahmud Hassan, Ph.D., of the Lister Hill
Center for Health Policy at the University of Alabama at Birmingham, and Gerald J. Wedig,
Ph.D., of Indiana University.
The researchers used 1988 to 1991 financial data from a Statewide hospital disclosure report on
189 nonprofit voluntary community hospitals in California. Analysis showed that an average
nonprofit hospital in California received $1.58 million in tax subsidies in 1991. These subsidies
increased with the size of the hospital, with large hospitals receiving nearly $4 million.
California hospitals provided on average more than $3 million in uncompensated care in 1991.
Larger hospitals were more likely than smaller hospitals to provide uncompensated care in excess
of their tax subsidies. Major teaching hospitals provided nearly $15 million in uncompensated
care. On average, the amount of uncompensated care, measured on a cost basis, exceeded the tax
subsidies by almost two to one. However, one in five nonprofit California hospitals failed to
provide charity care at least equal to the amount of tax subsidies it received, with an average
shortfall of about $815,000.
For more information, see "Do nonprofit hospitals pay their way?" by Drs. Morrisey, Wedig, and
Hassan, in the Winter 1996 issue of Health Affairs 15(4), pp. 132-144.
Return to Contents
Research center focuses on health of Asian and Pacific
Asian and Pacific Islander Americans (APIs) are the fastest growing minority group in the United
States, doubling from 3.5 million to 7.3 million between 1980 and 1990. Yet few studies on the
health of APIs have been reported in the medical research literature. The University of California,
Los Angeles Medical Treatment Effectiveness Program (MEDTEP) Center for Asians and Pacific
Islanders was established in 1992, as one of a group of research centers funded by the Agency for
Health Care Policy and Research (HS07370) to focus on health care for minority populations.
The goal of the UCLA MEDTEP Center is to find out whether prevailing health care strategies
and practices are appropriate and effective for the API community. The Center has found that as a
group, APIs have better or equal health compared with non-Latino American whites but are less
satisfied and feel less a partner in the doctor-patient relationship compared with other ethnic
A few MEDTEP studies provide insight into the health of APIs. One pilot study on falls and hip
fractures in APIs showed that 24 percent of the women studied (mostly Chinese Americans with
an average age of 86) died within 12 months of hip fracture. Women who had the poorest ability
to function and those who were taking sedative medications prior to hip fracture were at the
greatest risk of fracture. Another study showed that 200 API patients changed medication twice
as often and reported side effects two times more frequently than 196 age- and sex-matched
non-Latino American whites treated at the UCLA outpatient clinic.
Finally, MEDTEP researchers found that the increase in tuberculosis (TB) among Asian
Americans between 1985 and 1993 occurred primarily in immigrants from countries with high
endemic rates of TB, such as the Philippines, Vietnam, South Korea, and the People's Republic of
China. The peak incidence of TB occurred in the 2 months after immigrants arrived in the United
States. The risk of TB among newly arrived Vietnamese, Filipino, Chinese (mainland), and
Korean Americans was 112, 60, 37, and 28 times, respectively, relative to the risk of the U.S.
More details on the Center's research development strategy and core research projects are in
"Medical health outcome studies in the Asian and Pacific Islander American Population," by
Takashi Makinodan, Ph.D., and Nancy Harada, Ph.D., in the Asian American and Pacific
Islander Journal of Health 4(1-3), pp. 97-104, 1996.
Return to Contents
AHCPR News and Notes
President's budget includes increased funding for AHCPR in
President Bill Clinton's budget for fiscal year 1998, presented to Congress and the Nation on
February 6, includes $149 million for the Agency for Health Care Policy and Research, an
increase of $5.5 million over the FY 1997 level. This request reaffirms AHCPRs unique and
important niche to provide the public and private sectors with the information and data they need
to improve the quality and value of health care services.
With this amount, AHCPR will be able to fund existing commitments, support the Medical
Expenditure Panel Survey (MEPS), and support the new Quality and Cost-Effectiveness of
Clinical Care Initiative. To the extent that any additional efficiencies can be found, AHCPR is also
committed to maximizing the number of grants funded in 1998. The FY 1997 appropriation will
support a modest number of new small grants and dissertations and full funding of grant
commitments, based on expected cost savings to be identified for MEPS.
AHCPR's new $5 million quality initiative consists of two components. The first will increase
scientific knowledge by supporting outcomes, effectiveness, and cost-effectiveness research on
common, costly conditions for which the evidence base is incomplete or controversial. The second
component will fund research studies that will assess the effectiveness of strategies to implement
quality improvement techniques and clinical interventions in actual medical practice.
Data from MEPS, one of AHCPR's highest priorities, will be released for use by researchers
beginning in late March 1997. These data, which are not available from other surveys supported
by the Federal Government or the private sector, are essential for analyzing issues such as
insurance portability, access to health care, spending by different population groups, and changes
in overall health care financing and delivery.
AHCPR names new director for outcomes research program
Carolyn M. Clancy, M.D., has been appointed as the new Director of the Center for Outcomes
and Effectiveness Research (COER) at the Agency for Health Care Policy and Research. Dr.
Clancy, a clinical researcher and practicing internist, has been with AHCPR since 1990 and has
served as Acting Director of COER since the departure of Richard Green, M.D., in March
During her time with AHCPR, Dr. Clancy has held a number of senior positions. Most recently,
she served as Director of the Center for Primary Care Research. She will remain as Acting
Director of that Center while a national search is conducted for her replacement. Dr. Clancy is a
member of the Public Health Service's Commissioned Corps and is board certified in internal
Under her direction, COER will place a special emphasis on filling vital gaps in knowledge about
the effectiveness of treatments for prevalent, costly conditions, and continue to focus on the
application of patient-reported outcomes measures, pursue research to inform the effective use of
pharmaceutical interventions, and work toward the development of tools to help patients make
good health care decisions.
New publications and data products are available from
AHCPR and NTIS
The following new publications and data products are now available from the Agency for Health
Care Policy and Research and the National Technical Information Service.
Child Health Services: Building a Research Agenda. Over the past 25 years, health services
researchers have made impressive strides in developing methods to understand the organization
and financing of health care systems and assess the effectiveness of health care services and their
impact on patient outcomes. However, most of this work focuses on health care for adults,
leaving important gaps in our understanding of children's health care services. A new report
developed by AHCPR in response to a request from the U.S. House of Representatives
Committee on Appropriations is intended to facilitate further development of an agenda for health
services research focused on children. The report explains the need for such an agenda; describes
AHCPRs activities and progress in this area, including its collaborative approach to developing a
children's research agenda; and outlines principal priorities identified to date. The report also
highlights examples of new findings with potentially significant implications for children's health.
The report (AHCPR Publication No. 97-R022) is available from the AHCPR Publication
Most Frequent Diagnoses and Procedures for DRGs, by Insurance Status. Healthcare Cost
and Utilization Project (HCUP-3) Research Note 4. Elixhauser, A., Duffy, S.Q., and
Sommers, J.P. December 1996. AHCPR Publication No. 97-0006. This Research Note
presents information on the most frequent diagnoses and procedures for the top 50
diagnosis-related groups (DRGs) in U.S. community hospitals. The analysis is based on data from
the 1992 Nationwide Inpatient Sample (NIS) of the Healthcare Cost and Utilization Project
(HCUP-3). The five most common principal diagnoses and five most commonly performed
principal procedures are listed for each of the 50 most frequent DRGs. Mean and median charges
and length of stay for each DRG-diagnosis combination and each DRG-procedure combination
are provided along with estimates of standard errors. Results are provided for all patients
combined and for three patient groups defined by their insurance status: privately insured,
Medicaid, and self-pay patients. This information will be useful to several audiences: medical
professionals can compare their own practices to a nationwide sample; third-party payers and
managed care organizations can use it as a starting point for examining the impact of payment
policies and insurance status on practice patterns; health services researchers can generate
hypotheses for future research on the treatment of specific conditions and the variation in resource
use within DRGs, as well as differences in morbidity, use of services, and patterns of care by
Research Note 4 is available from the AHCPR
Nationwide Inpatient Sample (NIS) Release 3. NIS Release 3, which contains 1994 data
hospital inpatient care, conditions, services, and costs, is now available from the National
Technical Information Service. Developed as part of AHCPR's Healthcare Cost and Utilization
Project (HCUP-3), the NIS consists of 6.4 million patient records from a geographically dispersed
sample of 904 community hospitals in 17 States. NIS Release 3, along with NIS Releases 1 and 2,
form the largest, most powerful source of discharge-level information for hospital inpatients,
regardless of payer, in the United States. NIS Release 3, which consists of six CD-ROMs
designed to run on desktop computers, contains information on more than 100 variables,
including: patient demographic characteristics, total patient charges, median income for patient's
area of residence, diagnoses and procedures, admission and discharge status, links to hospital and
community characteristics, length of stay, weights for national estimates, and expected source of
payment. As with past releases, the data will prove useful to public- and private-sector
organizations interested in hospital care. Data from previous releases have been used to compare
services of various types of payers, assess treatment variations, prioritize strategies for product
development by studying conditions treated in hospitals, and determine the costs of treating
specific conditions. More than 200 organizations are now using NIS data. In addition, the
hospital identifiers in NIS Release 3 allow users to obtain detailed information on individual
hospitals by linking the data with the American Hospital Association's Survey of Hospitals. Users
also can obtain county-level information, such as the availability of medical personnel and
services, by linking the NIS data to the Bureau of Health Professions' Area Resource File. The
weights included in the NIS data allow users to produce national estimates.
NIS Release 3 is
available from National Technical Institute Service (NTIS
accession no. PB97-500433,
$160.00) as a set of six CD-ROMs in ASCII format. The set includes documentation in Adobe®
Acrobat® (a Windows® Reader™ is provided) and NIS tools (to assist SAS and SPSS users). NIS
Release 1, which presents data for 1988 to 1992, is available from National Technical Institute
Service (NTIS accession no. PB95-503710, $300.00) on a set of 26 CD-ROMs. NIS
covering 1993, is available from National Technical Institute
Service (NTIS accession no.
PB96-501325, $160.00) on six CD-ROMs. Users of NIS data must sign and return a Data Use
Agreement supplied by NTIS before their CD-ROM set can be shipped. For more information on
HCUP-3 or the NIS, phone toll-free 866-290-HCUP (4287) or E-mail email@example.com.
Calling all AHCPR Web site users
We'd like to hear your evaluation of the AHCPR Web site. Customer feedback will help us make the site responsive to your needs and the needs of other
We are conducting a 1-month, online survey during March 1997. We want you to tell us how we are doing and what we could do better. AHCPR's Web site offers access to information and data related to AHCPR's activities in the areas of health services research, quality measurement and improvement, clinical care, health technology assessment, and medical expenditures. The site also features consumer information for health care decisionmaking. Your ideas on how we can make this electronic
dissemination system more responsive to your information needs are welcome.
AHCPR funds new conference grants
The following five conference grants were awarded recently by the Agency for Health Care Policy
Advances in research for rural community and migrant health center operations
Project dir: Keith J. Mueller, Ph.D.
Organization: University of Nebraska
Medical Center, Omaha, NE
Project no: AHCPR grant HS09364
Period: 2/1/97 to 10/31/97
Chicago health services research symposium
Project dir: Edward F. Lawlor, Ph.D.
Organization: University of Chicago, Chicago, IL
Project no: AHCPR grant HS09361
Period: 12/1/96 to 3/31/97
Efficacy of prenatal care for women and children's health
Project dir: Marie C. McCormick, M..D. Sc.D.
Organization: Harvard School of Public Health
Project no: AHCPR grant HS09528
Period: 1/1/97 to 12/31/97
Healthy People 2000: Taking action with children and families
Project dir: Ardene Brown, Ph.D.
Organization: Marquette University,
Project no: AHCPR grant HS09362
Period: 2/1/97 to 1/31/98
Value-based health care purchasing by employers
Project dir: James D. Mortimer, B.A.
Organization: Midwest Business Group on Health,
Project no: AHCPR grant HS09363
Period: 2/1/97 to 4/30/97
Return to Contents
American Medical Informatics Association (1996). "Proceedings of the 1996 AMIA annual fall
symposium: Beyond the superhighway: Exploiting the internet with medical informatics."
(AHCPR grants HS08362, HS08751, HS07632, HS07719, HS06594, HS07763, HS06575,
HS08749, and HS00028). Journal of the American Medical Informatics Association
The theme of this medical informatics conference was use of the Internet and World Wide Web
for medical computing. There were seven sessions. The first session, Training, Education and
Cognition, focused on bridging health informatics and education, social-cognitive approaches to
understanding human-computer interaction, using the Web for accessing and designing medical
curricula, Web-based approaches to professional training, and clinical lexicons. The second
session, Policies and Standards, examined health information standards and terminologies. The
third session, Expert Systems and Algorithms, offered papers ranging from the use of neural
networks to predict survival in AIDS patients and use of expert systems via the Internet for
diabetic patient management, to image and signal processing such as brain mapping and
defibrillator pattern recognition. The fourth session, Information Retrieval and Digital Libraries,
covered data confidentiality, publishing biomedical journals on the Web, concept-based
information retrieval, and Web information resources. The fifth session, Clinical Information:
Storage and Use, focused on advancing medical records, computer-based guidelines, and
computer-based records; natural language processing and guideline implementation; and user
interfaces and hospital systems. The sixth session, Health Information Networks, explored such
issues as implementing Web-based clinical information and "clinics without walls." The final
session, User Interfaces and Hospital Systems, examined the uses and evaluation of physician
order entry, Web-based interfaces to clinical information systems, and bedside workstations.
Cappelleri, J.C., Ioannidis, J.P., Schmid, C.H., and others. (1996).
"Large trials vs.
of smaller trials." (AHCPR grant HS08532). Journal of the American Medical Association 276(16), pp. 1332-1338.
Large-scale clinical trials and meta-analyses of smaller trials are both used to obtain evidence for
medical treatment decisions. Although some previous studies have shown differences in the results
of these two approaches, this study found that results of meta-analyses of smaller studies are
usually compatible with the results of large studies, based on an examination of 79 meta-analyses
with at least one large study of 1,000 or more patients. Random effects calculations showed
agreement between large and smaller trials in 90 percent of meta-analyses selected by the
sample-size approach and in 82 percent of the meta-analyses selected by the statistical power
approach. Of the 15 disagreements between results of large and smaller trials using the random
effects model, plausible explanations were identified in 10 meta-analyses. Two other
disagreements were not clinically important, and tentative reasons could be identified for two of
the remaining three disagreements.
Iezzoni, L.I., Shwartz, M., Ash, A.S., and Makiernan, Y.D. (1996,
in-hospital mortality for stroke patients." (AHCPR grant HS06742). Medical Decision
Making 16, pp. 348-356.
Patients with stroke vary considerably in their use of hospital resources and risk of in-hospital
death, depending on how ill they are. Sicker patients are more likely to do poorly than others
managed by the same treatment or provider. However, this study shows that different measures of
patient severity of illness yield a different probability of hospital death for stroke patients. The
researchers studied 9,407 stroke patients from 94 hospitals, with 10 percent in-hospital deaths.
They calculated probability of death, adjusting for age and sex, using five severity measures.
Results showed that two severity measures, MedisGroups and the physiology scores, predicted
similar odds of death for 82 percent of patients. MedisGroups and Patient Management
Categories (PMCs) disagreed the most, with very different odds predicted for 62 percent of
patients. Patients viewed as more severely ill by MedisGroups and the physiology scores were
more likely to have clinical stroke findings than patients seen as sicker by the other severity
measures. This suggests that MedisGroups and the physiology score are more clinically credible
than the other measures used.
Katz, D.A., Griffith, J.L., Beshansky, J.R., and Selker, H.P. (1996). "The
use of empiric
clinical data in the evaluation of practice guidelines for unstable angina." (AHCPR grant
HS07360). Journal of the American Medical Association 276(19), pp.
A clinical practice guideline released by Agency for Health Care Policy and Research in 1994
recommends outpatient management of low-risk patients and intensive care unit (ICU) admission
of high-risk patients who arrive at hospital emergency departments (EDs) with unstable angina.
However, this study by researchers at Tufts University School of Medicine found that low-risk
patients with unstable angina account for only 6 percent of patients who arrive at the ED with
chest pain, and that often there are no ICU beds available for high-risk patients (about 40 percent
of patients). The researchers studied placement of patients with symptoms suggestive of acute
cardiac ischemia and unstable angina prior to and after dissemination of the guideline. These
10,785 patients arrived at the EDs of 10 hospitals during 1993 and participated in the Acute
Cardiac Ischemia-Time-Insensitive Predictive Instrument (ACI-TIPI) Trial. Only 4 percent of
low-risk patients were discharged directly home with outpatient followup; the others were placed
in telemetry beds with constant heart monitoring. Only 40 percent of high-risk patients were
admitted to ICUs, with the rest placed in telemetry beds. For many patients, chest pain was
controlled without the use of medications that would necessitate ICU admission. Often, there
were no ICU beds available for high-risk patients. The small number of low-risk ED patients with
unstable angina suggests that following the guidelines recommendations will result in only
modest reductions in hospital admissions but may increase demand for the limited number of ICU
beds to accommodate high-risk patients. According to the researchers, these findings underscore
the importance of obtaining data on the likely actual impact of guidelines on clinical care prior to
their national dissemination.
Musen, M.A., Tu, S.W., Das, A.K., and Shahar, Y. (1996, December).
component-based approach to automation of protocol-directed therapy." (AHCPR grant
HS06330). Journal of the American Medical Informatics Association 3(6), pp.
Providing automated support for planning protocol-directed therapy requires a computer program
to take as input clinical data stored in an electronic patient-record system and to generate as
output recommendations for therapeutic interventions and laboratory testing that are defined by
applicable protocols. The authors present a synthesis of research carried out at Stanford
University to model the therapy-planning task and to demonstrate a component-based architecture
for building protocol-based decision-support systems. The researchers developed general purpose
software components that interpret abstract protocol specifications to construct appropriate
patient-specific treatment plans; infer from time-stamped patient data higher-level, interval-based,
abstract concepts; perform time-oriented queries on a time-oriented patient database; and allow
acquisition and maintenance of protocol knowledge. They implemented these components in a
computer system known as EON. They tested the EON architecture within the context of
T-HELPER, a computer-based patient record system that uses EON to offer advice regarding the
management of patients who are following clinical trial protocols for AIDS or HIV infection.
Their experience demonstrates that EONs ability to reason about protocol-based care is
sufficient to provide therapy recommendations for a wide range of AIDS-related clinical trials. In
addition, a test of the reuse of the software components in a different clinical domain
demonstrated rapid development of a prototype application to support protocol-based care of
patients who have breast cancer.
Ohsfeldt, R.L., Morrisey, M.A., Johnson, V., and Treat, R. (1996).
assessment of rural emergency medical service trauma transport." (AHCPR grant
HS06814). Medical Care 34(11), pp. 1085-1092.
Managed care organizations and State licensure and emergency medical service (EMS) programs
would like to know whether the choice of an ambulance company has any impact on time to care
or patient outcomes. Assessing expected versus actual outcome is ideally based on the patients
initial Trauma Related Injury Severity Score (TRISS). However, this measure is usually only
available for patients ultimately admitted to a trauma center or hospital that maintains a trauma
registry. This study found that in rural settings, where a patient's TRISS generally is not available,
studies of rural EMS characteristics and trauma outcomes may use Revised Trauma Score, patient
age, and type of trauma to control for expected survival. Data were taken from ambulance trip
reports and merged with the trauma registry data for the Georgia EMS region VI trauma center
hospital. The researchers used these factors and TRISS to predict the probability of patient death
and compared it with predicted probability of death when the model omitted TRISS.
Osmond, D.H., Vranizan, K., Schillinger, D., and others. (1996,
need for medical care in an ethnically diverse population."(AHCPR grant HS07373). HSR:
Health Services Research 31(5), pp. 551-571.
Need for medical care can be measured in several ways, for example by number of days spent in
bed, existence of a major chronic condition, use of prescription medication, and self-rated health.
How various ethnic groups judge these factors as indicating the need for care varies, and it is not
clear how medical need relates to use of outpatient health services, according to this study. None
of the measures of medical need used by the researchers (self-rated health, activity limitation,
major chronic conditions, need for ongoing treatment, bed days, and prescription medication)
could be applied to adjust health care use uniformly across five ethnic groups. Compared with
whites, one or more of the other ethnic groups varied significantly for each need-for-care
measure, after adjustment for health insurance, age, sex, and income. Overall, blacks reported the
most need for care and Latinos the least. Latinos interviewed in Spanish reported the worst
health, and whites reported the best mean self-rated health. Asians and Latinos interviewed in
English reported less need for care than whites on four measures, but reported more bed days in
the previous 3 months, as well as worse self-rated health.
Return to Contents
AHCPR Publication No. 97-0021
Current as of February 1997