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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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.
Despite greater risk factors, black men and women are
only half as likely as whites to have heart attacks
Among patients presenting to the emergency department (ED), black men and women are half as
likely as their white counterparts to develop acute myocardial infarction (AMI or heart attack)
and 60 percent as likely to develop acute cardiac ischemia (ACI, insufficient blood supply to the
heart muscle), even though they are more at risk for heart disease because of higher rates of
smoking, hypertension, and diabetes. Whites, on the other hand, are more likely to have existing
coronary artery disease, as indicated by their more frequent histories of angina, AMI, and use of
cardiac medications, according to a study supported by the Agency for Health Care Policy and
As part of the ACI-TIPI Clinical Trial, investigators studied 10,673 patients (including 6,600
whites and 3,401 blacks) at 10 hospitals around the country, including all those patients 30 years
of age or older who arrived at the ED with chest pain or other symptoms suggestive of ACI.
Upon ED admission, more black women had chest pain as a primary symptom, and a higher
proportion of black men and women had additional symptoms, including shortness of breath,
abdominal pain, nausea, vomiting, and dizziness, as well as high blood pressure.
These multiple symptoms more often resulted in noncardiac diagnoses for blacks than whites,
according to Harry P. Selker, M.D., M.S.P.H., of New England Medical Center, the study's
principal investigator. Six percent of black men versus 12 percent of white men were diagnosed
with AMI, while 8 percent of black men versus 20 percent of white men were diagnosed with
angina pectoris (searing chest pain). Only 14 percent of black men had chest pain or other
symptoms due to ACI, compared with 32 percent of white men. Noncardiac diagnoses were
established for half of black men and women, one-third of white men, and 39 percent of white
women. Only 4 percent of black women were diagnosed with AMI compared with 8 percent of
white women, and 12 percent of black women versus 17 percent of white women were diagnosed
with angina pectoris. Chest pain and other symptoms were attributed to ACI in 16 percent of
black women and 25 percent of white women.
Details are in "Causes of chest pain and symptoms suggestive of acute cardiac ischemia in
African-American patients presenting to the emergency department: A multicenter study," by
Charles Maynard, Ph.D., Joni R. Beshansky, R.N., M.P.H., John L. Griffith, Ph.D., and Dr.
Selker, in the Journal of the National Medical Association 89(10), pp. 665-671, 1997.
Return to Contents
Effective treatment of geriatric urinary incontinence
depends on proper assessment
Urinary incontinence (UI) affects 15 to 30 percent of the population and half of those living in
nursing homes. But this often embarrassing condition is underreported by patients and
underdiagnosed by doctors. Proper diagnosis of geriatric UI includes distinguishing temporary
and persistent UI and the factors causing them from permanent UI, according researchers at the
University of North Carolina at Chapel Hill, who were supported by the Agency for Health Care
Policy and Research (National Research Service Award training grant T32 HS00032).
For instance, delirium promotes incontinence by increasing disorientation and decreasing patient
awareness of the need to void. Mobility impairment affects the ability to reach the toilet in time.
These factors should be managed by environmental manipulations, scheduled toileting, toilet
substitutes and pads, and attention to skin care. Also, UI due to anticholinergic, narcotic, and
beta-adrenergic drugs can be improved by switching or discontinuing medication or modifying the
dosage schedule when appropriate, explain Theodore M. Johnson II, M.D., and Jan
There are three broad types of treatment for management of UI: behavioral (kegel exercises,
bladder training, dietary changes, and prompted voiding), medication (anticholinergic agents,
alpha-adrenergic agents, and estrogen), and surgery (collagen injection, sling procedure, and
bladder neck suspension). According to Drs. Johnson and Busby-Whitehead, distinguishing the
type of UI the person has is critical in order to prescribe the appropriate treatment.
It is very difficult to correctly diagnose UI based solely on the patient's medical history, note
the researchers. They recommend that physicians evaluate all patients for urinary tract infection,
post-void residual volume, and simple cystometry measures of bladder capacity and stability.
Physicians should not use urodynamic tests such as uroflowmetry and cystometry as screening
tools because they do not differ significantly for continent and incontinent patients. Instead,
urodynamic tests should be used selectively as confirmatory tests to help determine the
See "Diagnostic assessment of geriatric urinary incontinence," by Drs. Johnson and
Busby-Whitehead, in the October 1997 American Journal of Medical Science 314(4), pp.
Characteristics of a nursing home's management
team affect the likelihood it will adopt innovations
Nursing homes whose top managers (director of nursing [DON] or administrator) are
well-educated, involved in professional societies, and who have been in the job a long time, are
more likely to adopt innovations that cut costs and meet Government regulations than homes
whose top managers lack these attributes, according to a study supported in part by the Agency
for Health Care Policy and Research (National Research Service Award training grant T32
The Minimum Data Set (MDS) is a tool used for the preliminary assessment of physical and
psychosocial functioning of nursing home residents that is still being introduced in nursing homes.
It identifies areas for further evaluation and helps facilities track resident outcomes for quality
assurance and improvement. To determine which nursing home management teams were early
adopters of a computerized version of the MDS, Nicholas G. Castle, Ph.D., of AtlaniCare Health
Systems, and Jane Banaszak-Holl, Ph.D., of the University of Michigan, examined tenure,
education, and involvement in a professional society, as well as relative homogeneity within the
top management team of 236 nursing homes in 10 States. Of these, 52 percent were chain-owned.
The researchers analyzed data from a telephone survey of nursing home administrators and
directors of nursing after implementation in 1993 of a Federal law requiring nursing homes to
identify and track residents' care preferences and their physical and psychosocial functioning.
They also used data from the Medicare/Medicaid Automated Certification Survey and the Area
Resource File. The researchers found that 50 percent of chain facilities and 44 percent of nonchain
facilities had computerized the MDS. As combined educational levels of the administrator and
DON increased, the likelihood of the nursing home adopting a computerized version of the MDS
increased by about one-third. As the degree of professional involvement of the top management
team increased (society membership, attendance at meetings), the likelihood of the home
computerizing the MDS almost doubled. Nursing homes belonging to chains that had asked their
members to use computers in the past were nine times more likely to have computerized the
MDS. In general, chain facilities were not as affected by external factors that played a role in
whether nonchain homes adopted the MDS, such as local income, size of the community's
elderly population, and Medicaid reimbursement level.
More details are in "Top management team characteristics and innovation in nursing homes," by
Drs. Castle and Banaszak-Holl, in The Gerontologist 37(5), pp. 572-580, 1997.
Return to Contents
Clinical trial results may be markedly affected when
small numbers of high-risk patients are included
The results of clinical trials generally apply to average patients. Some treatments may be effective
for all patients and under all conditions. But in other cases, it would be a mistake to extrapolate
results driven by the experiences of high-risk patients to patients at low risk, or vice versa. For
some diseases, both the disease and the patient populations change over time.
Instead of aiming at average estimates, mega-trials should study very heterogeneous groups but
stratify them into risk subgroups. This would give particularly low-risk or high-risk patients a
better idea how they might benefit from the treatment, concludes a study supported in part by the
Agency for Health Care Policy and Research (HS07782).
This conclusion is based on analysis of human immunodeficiency virus (HIV)-related trials and
trials of magnesium in acute myocardial infarction conducted by John P. Ioannidis, M.D., and
Joseph Lau, M.D., of Tufts University School of Medicine in Boston. They found that subtle
differences in patient inclusion or exclusion criteria can result in important differences in
representation of patients with distinct responses to treatment. For instance, if substantial
treatment benefit is seen mostly in very sick patients but patients at lower risk experience excess
harm due to toxicity, only high-risk patients would be expected to experience a net benefit from
treatment. Patients who are not at high risk may be the majority of those treated, however, and
they may experience an excess of harm from the therapy. Therefore, a clinical trial enrolling
patients from a homogeneous low-risk population will probably find the therapy to be useless, but
at the same time, the trial may miss detecting its efficacy for high-risk patients.
On the other hand, when clinical trials do not strictly enroll a very homogeneous population,
high-risk patients may inadvertently predominate in the control or the treatment group and distort
results. In other words, this small group of high-risk patients may be responsible for most of the
treatment complications or the greatest improvement due to treatment. In small trials enrolling
fewer persons, this effect often results in "surprising" or "unexpected" results, according to Drs.
Ioannidis and Lau.
For more information, see "The impact of high-risk patients on the results of clinical trials," by
Drs. Ioannidis and Lau, in the Journal of Clinical Epidemiology 50(10), pp. 1089-1098,
Southeast stroke belt has two to three times the rate of
elevated blood pressure of other U.S. regions
The proportion of persons with elevated blood pressure (EBP), a precursor of hypertension,
which is a well-established risk factor for stroke, varies regionally across the United States, with
the so-called "stroke belt" in the Southeast having the highest prevalence of EBP.
25 percent of black men in Birmingham, AL, have EBP, as compared with 9 percent in Chicago,
where rates were lowest. High salt diets or family history of hypertension, although linked to
hypertension, do not fully explain the Southeast's higher rates of EBP. Accounting for known
dietary, behavioral, and familial correlates of hypertension did not change regional differences in
EBP in a study supported in part by the Agency for Health Care Policy and Research (HS09446).
Catarina I. Kiefe, Ph.D., M.D., of the University of Alabama at Birmingham, and her colleagues
followed a group of 5,115 black and white men and women, aged 18 to 30 in 1985 or 1986, for 7
years at four centers: Birmingham, AL; Chicago, IL; Minneapolis, MN; and Oakland, CA.
Regional disparities did not show up initially but became more apparent as the population aged.
Compared with Birmingham, Chicago and Minneapolis had one-third the prevalence of EBP and
Oakland one-half. EBP for black and white men ranged from a low in Chicago (9 percent for
black men and 5 percent for white men) to a high in Birmingham (25 percent for black men and
14 percent for white men). Although not significantly different, the incidence of EBP for black
and white women was highest in Birmingham. Birmingham also had the highest 7-year incidence
(11 percent) and overall prevalence at year 7 (14 percent). These differences were consistent
regardless of race or sex. Birmingham showed the highest sodium intake and lowest potassium
and magnesium intakes, as well as the highest proportion of persons with a family history of
hypertension (67 percent). After further analysis, known dietary, behavioral, and familial
correlates of hypertension did not fully explain disparities between Birmingham and other regions.
In fact, adjusting for many factors that correlate with hypertension—such as body mass
weight gain, physical activity, dietary intake, alcohol and tobacco use, education, oral
contraceptive use, and family history of hypertension—had only limited impact on
differences in EBP rates.
The authors call for further studies to investigate whether these differences become more marked
as the study population ages. They believe continued followup will lead to a better understanding
of the known geographic variability in stroke morbidity and mortality.
More details are in "Regional disparities in the incidence of elevated blood pressure among young
adults: The CARDIA study," by Dr. Kiefe, O. Dale Williams, Ph.D., Diane E. Bild, M.D.,
M.P.H., and others, in Circulation 96(4), pp. 1082-1088, 1997.
Health status questionnaire accurately assesses
changes over time in quality of life after elective
One multidimensional measure of health status can capture changes in health-related quality of life
(HRQL) after elective surgery and help assess a person's need for in-home services after hospital
discharge, according to a study supported by the Agency for Health Care Policy and Research
(HS06573). The Short Form Health Survey (SF-36) questionnaire, for example, found that
patients undergoing hip arthroplasty, thoracic surgery, or abdominal aortic aneurysm repair had
worse physical function and more role limitations due to health problems 1 month after surgery
than they had prior to surgery. However, by 6 months after surgery, the majority of patients
experienced significant gains in most health dimensions, and these were sustained 1 year after surgery.
The SF-36 assesses health status in eight areas: health perception, bodily pain, physical function,
physical role, social function, emotional role, mental health, and vitality. How much each of these
areas changes depends on the type of surgery and timing of followup after surgery. For instance,
for total hip arthroplasty patients, responsiveness was greatest for the SF-36 scales that measure
physical constructs. However, for patients undergoing the other two procedures, all eight areas
changed significantly at various points of recovery.
The SF-36 moderately correlated with other indicators of physical function and health perception
measured by other instruments used in the study. The patterns of change and consistency of
results with other instruments suggest that differences detected with the SF-36 are valid, conclude
Thomas Lee, M.D., of Harvard Medical School, Lee Goldman, M.D., M.P.H., of the University
California, San Francisco School of Medicine, and their colleagues. They analyzed responses
given by 528 adult patients admitted for the three elective surgeries to the SF-36, the Specific
Activity Scale, five validated health transition questions, and a 0-to-100 scale measure of global
health at 1, 6, and 12 months after surgery.
See "Health-related quality of life after elective surgery," by Carol M. Mangione, M.D.,
M.S.P.H., Dr. Goldman, E. John Orav, Ph.D., and others, in the November 1997 Journal of
General Internal Medicine 12, pp. 686-697.
Return to Contents
Health Care Delivery
Patients' treatment preferences reflect the value they
place on different health outcomes
Patients' choices for life-sustaining treatments are generally consistent with their preferences for
different health conditions, and this relationship remains consistent over time. When these
preferences are inconsistent, it is for understandable reasons, according to a study supported in
part by the Agency for Health Care Policy and Research (HS06343).
Researchers from the Veterans Affairs Puget Sound Health Care System and the University of
Washington, led by Robert A. Pearlman, M.D., M.P.H., interviewed seven groups of people at
baseline and 6, 18, and 30 months later. Participants were asked whether they would want six
treatments in five different health states, which they were asked to rate in varying degrees as
better or worse than death. The researchers measured the increase in odds of refusing
life-sustaining treatment for each one-point change in health state ratings. They considered
preferences to be concordant (consistent) if treatments were refused in health states rated as
worse than death and accepted in health states rated as better than death.
Treatment refusal differed by health state. Coma had the highest rate of treatment refusal.
Fifty-two percent of study participants rated permanent coma as worse than death, 31 percent
rated it as better than death, and the majority of treatments were refused. The current health state
had the lowest rate of treatment refusal; dementia, severe stroke, and severe pain had similar rates
of refusal in the middle range for every treatment.
As would be expected, the percentage of treatment refusals for all treatments increased as
the health state rating went from much better than death to much worse than death. For example,
a change in health state rating from a little worse than death to somewhat worse than death was
associated with a 70 to 94 percent increase in the odds of treatment refusal. Preferences were
generally consistent over time, particularly when the health state ratings were stronger, that is,
much worse than death or much better than death.
The 115 participants who rated their current health state as better than death but did not want
treatment had good reasons. Either they believed that receiving treatment would lead to a worse
state of health—such as machine dependency or increased dysfunction (71
percent)—or that the
treatments themselves were unacceptable (46 percent). Being a burden to others was another
stated reason for not wanting treatment in both current and hypothetical health states (22 and 23
See "Validation of preferences for life-sustaining treatment: Implications for advance care
planning," by Donald L. Patrick, Ph.D., M.S.P.H., Dr. Pearlman, Helene E. Starks, M.P.H., and
others in the October 1, 1997 Annals of Internal Medicine 127, pp. 509-517.
Doctors' counseling about reducing risky behaviors
falls short of national goals
About 40 percent of U.S. deaths due to nongenetic causes in 1990 were attributable to unhealthy
behaviors such as excessive eating, alcoholism, smoking, unsafe sex, and not wearing seat belts.
Unfortunately, doctors still are not doing enough to meet national goals for counseling patients
about reducing these risky behaviors, concludes a study supported by the Agency for Health Care
Policy and Research (HS08841).
Physicians who counsel patients about unhealthful behaviors can make a difference. In this study,
49 percent of patients who were counseled by their doctors about smoking tried to cut down or
quit smoking based on the doctors' advice. Nearly half of those who tried to stop smoking were
successful. A 50 percent increase in smoking discussion by doctors (from 53 percent to 80
percent) would result in a 6 percent decrease in smoking, a potential 24,000 annual deaths
delayed, and $3 billion in annual cost savings to society, note Deborah A. Taira, Sc.D., and Dana
Gelb Safran, Sc.D., of New England Medical Center.
Drs. Taira and Safran and their colleagues mailed a survey to a random sample of Massachusetts
State employees in 12 health plans. With the resulting responses (n=6,549), they examined the
relationship between patient income, health risk behaviors, prevalence of physician discussion of
these behaviors, and receptiveness of patients to their doctors' advice.
Low-income patients were more likely to be obese and to smoke and less likely to wear seat belts
and exercise than those with higher household incomes. Stress and alcohol consumption increased
with income, but the proportion of heavy drinkers (more than 20 drinks per week) did not vary
significantly. According to reports from patients at risk in each area, 73 percent of physicians
discussed exercise, 70 percent discussed diet, 61 percent discussed stress, 53 percent discussed
smoking, 39 percent discussed alcohol consumption, 19 percent discussed safe sex, and 16
percent discussed seat belt use. Overall, 19 percent of patients surveyed believed that their doctor
gave them too little advice on health risk behaviors.
Doctors were more apt to discuss diet and exercise with high-income patients in need of
counseling in these areas than with low-income patients. On the other hand, they were more apt
to discuss smoking with low-income than high-income patients who smoked. Low-income
patients in general were much more likely than high-income patients to try to change their
behavior based on their doctors' advice.
Details are in "The relationship between patient income and physician discussion of health risk
behaviors," by Drs. Taira and Safran, Todd B. Seto, M.D., and others, in the November 5, 1997
Journal of the American Medical Association 278(17), pp. 1412-1417.
Return to Contents
Use of Data/Medical Informatics
Medical informatics conference explores cutting-edge
Medical data mining to uncover factors contributing to preterm birth, Internet-based patient
interviews, and Web-based clinical information systems that enable referring physicians at remote
sites to access a hospital's clinical database are just a few of the many topics addressed at the
1997 Annual Fall Symposium of the American Medical Informatics Association held in Nashville,
TN. The conference theme, "Emergence of Internetable Health Care: Systems That Really Work,"
was addressed in 273 scientific manuscripts, 29 panels, and 131 posters that the Society published
in the 1997 Annual Fall Supplement to the Journal of the American Medical Informatics
Association and in the CD-ROM Extended Proceedings.
Seventeen of the studies presented at the conference were supported in whole or in part by the
Agency for Health Care Policy and Research. They are:
- Bichindaritz, I., Bradshaw, J.M., Sullivan, K.M., and others. "Distributed reuse of
in a computerized decision support system for bone-marrow post-transplant care over the
World-Wide Web," p. 844, AHCPR grant HS09407.
- Bourie, P.Q., Dresch, J. and Chapman, R.H. "Usability evaluation of an on-line nursing
assessment," p. 914, AHCPR grant HS08749.
- Das, A.K., and Musen, M.A., "A foundational model of time for heterogeneous clinical
databases," p. 106, AHCPR grant HS06330.
- de Estrada, W.D., Murphy, S., and Barnett, G.O. "Puya: A method of attracting attention
relevant physical findings," p. 509, AHCPR grant HS06575.
- Einbinder, J.S., Klein, D.A., and Safran, C.S. "Making effective referrals: A
knowledge-management approach," p. 330, AHCPR grant HS08749.
- Flowers, C.R., Garber, A.M., Bergen, M.R., and Lenert, L.A. "Willingness-to-pay utility
assessment: Feasibility of use in normative patient decision support systems," p. 223, AHCPR
- Hulse, M., Zielstorff, R.D., and Estey, G. "User-interface design of a Web-based clinical
decision support system," p. 951, AHCPR grant HS06575.
- Jones, P.C., van Wingerde, F.J., and Safran, C., "An Internet-based patient interview," p.
AHCPR grant HS08749.
- Kittredge, R., Usman, R., Melanson, F., and others. "Experiences in deployment of a
Web-based CIS for referring physicians," p. 320, AHCPR grant HS06575.
- Lobach, D.F., Gadd, C.S., and Hales, J.W., "Structuring clinical practice guidelines in a
relational database model for decision support on the Internet," p. 158, AHCPR grant
- Lush, M.T., and Henry, S.B. "Nurses' use of health status data to plan for patient care:
Implications for the development of a computer-based outcomes infrastructure," p. 136,
AHCPR NRSA training grant T32 HS00026.
- McDonald, C.J., Overhage, J.M., Tierney, W.M., and others. "The Regenstrief medical
system 1997: A system for clinical pervasive and city-wide computing," p. 1027, AHCPR
- Murphy, S.N., Rabbani, U.H., and Barnett, G.O. "Using software agents to maintain
autonomous patient registries for clinical research," p. 71, AHCPR grant HS06575.
- Prather, J.C., Lobach, D.F., Goodwin, L.K., and others. "Medical data mining:
discovery in a clinical data warehouse," p. 101, AHCPR grant HS09436.
- Prather, J.C., Lobach, D.F., Hales, J.W., and others, "Exploratory data analysis to detect
preterm risk factors," p. 984, AHCPR grant HS09331.
- Sands, D.Z., Chapman, R.H., Goodman, S., and others. "Drug allergies and reactions
A new paradigm to record complete information," p. 998, AHCPR grant HS08749.
- Zielstorff, R.D., Estey, G., Vickery, A., and others. "Evaluation of a decision support
for pressure ulcer prevention and management: Preliminary findings," p. 248, AHCPR grant
Return to Contents
AHCPR News and Notes
AHCPR and Kaiser join forces to study mental health
in managed care
The Agency for Health Care Policy and Research and the Kaiser Permanente Medical Care
Program in Northern California have begun a research project that could help all managed care
organizations target and provide appropriate health services for enrollees and family members
suffering from depression or other mental health problems. As a result, health plan enrollees could
benefit from improved access to high quality preventive services, more productive mental health
treatment, and added resources invested in continued service enhancements.
The research effort, led by Kaiser Permanente's Enid M. Hunkeler, M.A., and AHCPR's William
D. Spector, Ph.D., will focus on mild depression—one of the most prevalent mental health
problems among managed care patients—and on its interactions with chronic diseases,
heart disease, and risky behaviors, such as alcohol, drug, and cigarette use, on nonpsychiatric
health services utilization and costs. The researchers also will explore whether family members of
enrollees who have mental health symptoms differ from family members of enrollees who do not
have these symptoms in their use of other kinds of health services.
Findings from the project are expected to contribute to a more complete understanding of the
actual health care costs of depression and other mental health symptoms, such as panic disorder,
anxiety, and violent impulses in people who have other illnesses and those who do not. The
research will also look at persons who may exhibit these symptoms but do not use health
Current research suggests that better targeting and appropriate use of mental health services are
likely to have the most significant impact on persons with less severe mental problems and the
elderly. Studies also suggest that many visits for physical problems actually stem from mental
health problems. Around 10 percent of the population suffers from some form of depressive
illness. About 4 percent of the population meets the criteria for major depression. These rates are
twice as high among people seeking health care.
The California Division of Kaiser Permanente serves more than 5.3 million members throughout
the State and has more than 7,000 physicians and over 55,000 employees. The research will be
based on Kaiser Permanente data from a recent, large-scale telephone survey of a random sample
of enrollees regarding their mental health symptoms, chronic disease, and alcohol, cigarette, and
drug use, as well as data regarding their use and expenditures for care in Kaiser Permanente
hospitals, emergency rooms, outpatient facilities, and other medical services in Northern
California. The first findings from this research are expected in 1999.
For more information, select the press release.
AHCPR recognizes contributions of peer reviewers
Each year, hundreds of reviewers contribute their expertise to assist in the peer review of research
grant applications. The Agency for Health Care Policy and Research would like to acknowledge
and thank those who served as reviewers in 1997.
If you are interested in serving as a peer reviewer, please forward a current curriculum vitae to the Agency for Health Care Policy and Research, Office of Scientific Affairs, Attention Bonnie Campbell, 540 Gaither Road, Suite 2000, Rockville, MD 20850; or, you may fax a copy of your CV to Ms. Campbell at (301) 427-1561.
Return to Contents
Proceed to Next Section