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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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

Half the cases of early childhood asthma are due to secondhand cigarette smoke

About half of the cases of early childhood asthma, chronic bronchitis, and wheezing are due to exposure to secondhand cigarette smoke, according to a study by the Agency for Health Care Policy and Research. It shows that 38 percent of children aged 2 months to 5 years in a representative sample of the U.S. population (7,680) are exposed to environmental tobacco smoke (ETS) in the home, and 24 percent of these children were exposed by maternal smoking during pregnancy.

The impact of secondhand smoke was seen at 20 or more cigarettes per day. Children 2 months to 2 years of age who were exposed to this level of household smoke were 2.5 times as likely to develop chronic bronchitis and nearly 3 times as likely to have had three or more episodes of wheezing. Children 2 months to 5 years of age were twice as apt to develop asthma.

Children 2 months to 2 years of age exposed to maternal smoke in the womb were twice as apt to develop chronic bronchitis and twice as likely to have three or more episodes of wheezing. Those 2 months to 5 years of age with prenatal exposure were nearly twice (1.8) as apt to develop asthma. Peter J. Gergen, M.D., M.P.H., of AHCPR's Center for Primary Care Research, and his colleagues estimate that an extra 160,000 cases of asthma per year, 79,000 cases of bronchitis, and 172,000 cases of wheezing are due to ETS. However, in this study exposure to ETS was not significantly associated with the prevalence of upper respiratory infection, pneumonia, or cough.

ETS appears to increase the prevalence of asthma rather than the severity of the illness, which the researchers measured by medication use. Children exposed to tobacco smoke had similar use of prescription asthma medications to unexposed children. These findings are based on data from the Third National Health and Nutrition Examination Survey, 1988 to 1994, which included reports on household smoking and maternal smoking during pregnancy.

See "The burden of environmental tobacco smoke exposure on the respiratory health of children 2 months through 5 years of age in the United States: Third National Health and Nutrition Examination Survey, 1988-1994," by Dr. Gergen, Jean A. Fowler, M.S., Kurt R. Maurer, Ph.D., and others in the February 1998 Pediatrics Electronic Pages 101(2). Reprints (AHCPR Publication No. 98-R033) are available from the AHCPR Publications Clearinghouse.

Editor's note: Select for a related article by Dr. Gergen and other researchers on hospital admissions for asthma in New York City.

Return to Contents

Health Care Delivery

Nearly one-third of U.S. toddlers have not received complete immunizations

It is estimated that 30 percent of 2-year-olds in the United States have not received the recommended four doses of the diphtheria-tetanus-pertussis (DTP) vaccine, three doses of the polio vaccine, and single dose of the measles-mumps-rubella (MMR) vaccine. This vaccination series should be completed by age 2.

Two recent studies, supported by the Agency for Health Care Policy and Research and summarized here, highlight barriers and aids to childhood immunization. The first study (AHCPR grant HS08068) shows that in some cases physician concern about vaccine side effects and related lawsuits hinders timely vaccination of children. The second study (AHCPR contract 290-93-0038) demonstrates that public health clinics are playing an important role in the immunization of rural children.

Zimmerman, R.K., Schlesselman, J.J., Mieczkowski, T.A., and others (1998, January). "Physician concerns about vaccine adverse effects and potential litigation." Archives of Pediatric and Adolescent Medicine 152, pp. 12-19.

Physician concerns about the adverse effects of vaccines and—to a lesser degree—potential litigation hinder timely vaccination of children, concludes this study. This is a problem, especially given that the major measles epidemic in the United States in 1989 to 1991 was attributed to inadequate vaccination levels.

The authors of this study recommend development of new educational materials for health care providers that explain the risks of vaccine-preventable diseases, vaccine safety, and protection against litigation afforded by the Vaccine Injury Compensation Program (VICP). They surveyed by telephone a national random sample of family physicians, pediatricians, and general practitioners who frequently saw children younger than 6 years of age and had office-based practices in 1995.

Of the 1,236 physicians surveyed, 32 percent overestimated the risk for serious adverse effects related to DTP vaccines, and 13 percent overestimated the risks of measles vaccines. Among physicians who were not concerned about DTP-related risks, 15 percent were highly concerned about litigation. However, 38 percent of physicians who were concerned about the adverse effects of DTP vaccines also were highly concerned about vaccine litigation.

Among doctors most concerned about vaccine litigation, 22 percent—compared with 12 percent of doctors who expressed little concern—said they were unlikely to recommend the third dose of DTP for a child with a fever of 39.4o C and no other symptoms after the second DTP dose. The VICP provides no-fault compensation to individuals with a permanent injury that is temporally related to vaccination. Nevertheless, only 41 percent of doctors who were aware of the VICP believed that it afforded them substantial liability protection.

Slifkin, R.T., Clark, S.J., Strandhoy, S.E., and Konrad, T.R. (1997, Fall). "Public-sector immunization coverage in 11 States: The status of rural areas." Journal of Rural Health 13(4), pp. 334-341.

Are rural children less likely to be immunized due to less Medicaid insurance, poverty, fewer physicians, and greater distance to health care facilities? Not rural children who are dependent on the public health system, finds this study. It shows that in 11 Eastern States, rural counties had average immunization completion rates that were 2.5 percent higher than metropolitan counties, even after controlling for county socioeconomic characteristics.

States varied widely. High mean public-sector immunization completion rates in Georgia (91 percent), South Carolina (84 percent), and Alabama (79 percent) probably reflect the South's strong tradition of providing direct clinical services to the public sector. In contrast, many Northern industrial States have not traditionally viewed immunization delivery and promotion as the responsibility of the public sector. For example, Michigan and Ohio, which had the lowest public-sector immunization rates among the 11 States studied (46 percent and 53 percent, respectively), also had overall State immunization rates that were below the national average.

Rural counties had higher rates than urban counties in every State except for West Virginia. The researchers conclude that rural children fare comparatively well using the present system of immunization delivery, and they recommend maintaining it. Their findings are based on analysis of 1995 county-level immunization data generated by State public health agencies in 882 counties in 11 States east of the Mississippi River.

Laboratory testing and antibiotic therapy are often at odds with expert advice about Lyme disease care

Lyme disease (LD) is caused by the tickborne bacteria Borrelia burgdorferi, but tick bites rarely result in LD. In 1995, 11,603 cases of LD were reported to the Centers for Disease Control and Prevention by 43 States and the District of Columbia (overall incidence of 4.4 per 100,000 population). Nevertheless, tick bites are a common reason for people in endemic areas to seek medical care.

It takes 2 weeks or longer for antibodies to B. burgdorferi to show up in the blood, but half of patients who seek care for tick bites in Lyme-infested areas have a blood test performed right away. Most receive prophylactic antibiotics regardless of whether they have a blood test or the results of the test. Thus, most patients with tick bites undergo costly blood testing before they would have detectable antibodies, and most receive antibiotic therapy of unproven benefit, concludes a study supported in part by the Agency for Health Care Policy and Research (HS07813).

Researchers led by G. Thomas Strickland, M.D., Ph.D., of the University of Maryland School of Medicine, studied the use patterns of serologic testing and antibiotic therapy for 232 patients with tick bites, LD, or suspected LD by physicians in primary care practices on Maryland's Eastern Shore in 1995. This is an area endemic for LD (more than 100 cases per 100,000 population).

Blood testing for LD did not appear to influence treatment of patients diagnosed as having LD. About half of the patients who sought care for a tick bite had a blood test performed at the initial visit, but did not have the followup "convalescent" assay needed to verify LD (when antibodies have had time to build up). Fifty-five percent of patients with a tick bite received antibiotic therapy, but in most cases, antibiotics were prescribed before serologic test results became available. Convalescent testing was not performed for 86 percent of the 43 patients with suspected LD who had initial negative or equivocal results. Of these patients, 68 percent did not receive antibiotic therapy. No patient with a tick bite developed clinical LD.

Details are in "Tick bites and Lyme disease in an endemic setting: Problematic use of serologic testing and prophylactic antibiotic therapy," by Alan D. Fix, M.D., M.S., Dr. Strickland, and John Grant, M.D., M.P.H., in the January 21, 1998, Journal of the American Medical Association 279(3), pp. 206-210.

Editor's note: Select for a summary of a related article by Dr. Strickland and colleagues on the use and costs of serologic testing for LD in Maryland (Journal of Infectious Diseases 176, pp. 819-821, 1997).

Return to Contents

Clinical Decisionmaking

Treadmill testing of certain ED patients with chest pain may substantially reduce hospital admissions and costs

About 2 million people with chest pain are admitted to the hospital each year for tests to rule out a heart attack. However, careful selection of low-risk patients to undergo early exercise stress testing—after 6 hours of uneventful observation in the emergency department (ED)—could alter this scenario. This approach could reduce the number of hospital admissions by 17 percent and days of hospitalization by 11 percent, with a potential savings of more than $500 million a year, according to a recent study supported in part by the Agency for Health Care Policy and Research (HS06452).

The researchers used published evidence, the experiences of local medical opinion leaders, and physician-investigators to develop a critical pathway for management of patients with acute chest pain who are at low risk for complications of ischemic heart disease. Criteria were developed to determine which patients seen in one hospital's ED for chest pain should be admitted to the hospital, directly discharged from the ED, or observed for 6 hours in the ED and, if appropriate, subjected to an exercise treadmill test. Of 4,585 patients, 63 percent were admitted to the hospital. Among those admitted, 40 percent were classified as potentially eligible for observation in the ED, and 93 percent had a benign clinical course during the initial 6-hour observation period.

The 1,068 patients observed in the ED had a mean length of stay of 3 days (plus or minus 5 days). If 47 percent of these patients had been discharged after observation and exercise testing, implementation of the critical pathway would have reduced the number of admissions by 505 (17 percent) and days of hospitalization by 1,407 (11 percent). The pathway is a step-by-step procedure for selection and management of patients. Based on the pathway, if an ED patient who was initially thought to be at low risk for heart attack (probability of 5 percent) had a benign clinical course, including a negative result cardiac enzyme testing during the 6-hour observation period, the post-test probability of heart attack would be low enough to consider noninvasive exercise stress testing. Among the patients who were stable after at least 6 hours of observation, 1.2 percent had a heart attack compared with 17.5 percent of ineligible patients.

Details are in "A critical pathway for management of patients with acute chest pain who are at low risk for myocardial ischemia: Recommendations and potential impact," by Graham Nichol, M.D., Ron Walls, M.D., Lee Goldman, M.D., and others, in the December 1, 1997, Annals of Internal Medicine 127(11), pp. 996-1005.

Heparin reduces catheter-related clot formation and may reduce catheter-related infections

Many hospitalized patients require a central venous catheter to provide access for medication, fluids, and parenteral nutrition. Also, many critically ill patients require central venous or pulmonary arterial catheters for hemodynamic monitoring and blood sampling. These catheters put them at risk for vascular thrombosis (blood clots that partially or completely block the blood vessel) and systemic infections. But administration of the anticoagulant heparin effectively reduces thrombus formation and may reduce catheter-related infections, finds a study supported by the Agency for Health Care Policy and Research (National Research Service Award fellowship F32 HS00106).

Researchers led by Adrienne G. Randolph, M.D., M.Sc., of Harvard Medical School performed a meta-analysis of 14 randomized controlled trials evaluating prophylactic doses of heparin (as a heparin flush of the catheter between uses) or heparin-bonded catheters. They found that prophylactic heparin decreased central venous catheter-related thrombosis by 57 percent and bacterial colonization of these catheters by 82 percent. This reduction in bacterial colonization may decrease central venous catheter-related bacterial infections by 74 percent.

Analysis of two trials comparing heparin-bonded versus unbonded pulmonary artery catheters showed that heparin bonding reduced by 92 percent the risk of catheter clot formation within the first 24 hours after catheter insertion.This is significant, especially when one considers that blood clots form on these catheters in the first few hours, and thrombosis of large vessels occurs after long-term catheterization in 35 to 67 percent of patients.

For more details, see "Benefit of heparin in central venous and pulmonary artery catheters: A meta-analysis of randomized controlled trials," by Dr. Randolph, Deborah J. Cook, M.D., M.Sc., F.C.C.P., Calle A. Gonzales, M.D., M.P.H., and Maureen Andrew, M.D., in the January 1998 Chest 113 (1), pp. 165-171.

Moderately anemic elderly patients who undergo surgery don't necessarily require blood transfusions

The risks of blood transfusions have been evaluated extensively in recent years; surprisingly, the benefits of transfusion are far less well understood. Older patients with anemia (low blood count) who lose blood during surgery may be at increased risk of adverse events such as heart attacks, but it is not clear which patients are likely to benefit from transfusions.

A recent study, supported by the Agency for Health Care Policy and Research (HS07322), examined whether transfusions given to elderly patients undergoing surgery for a fractured hip had a positive impact on postoperative outcomes. Other goals of the study were to identify which patients were most likely to benefit and the level of hemoglobin (blood count) that signaled the need for a transfusion to decrease the risk of adverse events due to anemia.

According to the study's findings, some elderly patients may not need red blood cell transfusions following surgery, even at hemoglobin levels as low as 80 g/L (moderate anemia). Ten years ago the National Institutes of Health Consensus Development Conference recommended that otherwise healthy patients with hemoglobin values of 100 g/L or greater rarely require postoperative transfusion, whereas those with acute anemia (hemoglobin values of less than 70 g/L) frequently do.

This retrospective study of 8,787 high-risk elderly patients who had extensive coexisting illnesses and underwent surgery for hip fracture suggests a lower transfusion trigger. The researchers studied the medical charts of these patients, who underwent surgical repair between 1983 and 1993 at one of 24 study hospitals. They found that postoperative transfusion did not influence 30-day mortality (adjusted odds ratio 0.96 [1 is equal risk], 95 percent confidence interval, 0.74-1.26). Results were similar for the 7.2 percent who underwent a preoperative transfusion.

Since there is good evidence about the risks of transfusion, such as infections and allergic reactions, physicians should reconsider whether transfusion is warranted for these patients, says Jeffrey L. Carson, M.D., of the University of Medicine and Dentistry of New Jersey. He points out, however, that it is possible that transfusion may affect other patient outcomes, such as morbidity, readmission to the hospital, speed of recovery, and functional status.

See "Perioperative blood transfusion and postoperative mortality," by Dr. Carson, Amy Duff, M.H.S., Jesse A. Berlin, Sc.D., and others in the January 21, 1998/ Journal of the American Medical Association 279(3), pp. 199-205.

Nonclinical factors affect use of maternal corticosteroids to reduce respiratory complications in premature infants

Clinical factors primarily determine physicians' use of intravenous tocolytic medication to halt premature labor contractions, a commonly accepted practice. But patient race and site of care affect physicians' use of maternal corticosteroid therapy, an effective but less widely accepted practice that hastens maturation of fetal lung tissue to prevent respiratory complications in premature infants.

In a recent study of the use of tocolysis and corticosteroid therapy in the treatment of premature labor, black women were one-third less likely and Latino women were nearly 100 percent less likely to receive corticosteroid therapy than white women in similar clinical situations. There were no differences by race in the use of tocolysis. The study was conducted by the Patient Outcomes Research Team (PORT) on Prevention of Low Birthweight in Minority and High-Risk Women, which is led by Robert L. Goldenberg, M.D., of the University of Alabama at Birmingham and supported by the Agency for Health Care Policy and Research (PORT contract 290-92-0055). The researchers analyzed the preterm labor or delivery records of 33,792 women in the March of Dimes Prematurity Prevention Trial database from 1983 to 1986.

According to the researchers, the variation in use of corticosteroids during the 1980s appears to have stemmed from nonclinical factors, since most clinical indications, technological context, and site variations were controlled for in the analysis. This study confirms the premise that practice variation on the basis of nonclinical factors occurs more commonly for interventions where there is more uncertainty about clinical indications and effectiveness. It also identifies another area of clinical care in which aggressive and relatively uncertain interventions are used less often for minority patients, concludes lead author Janet M. Bronstein, Ph.D.

Details are in "Practice variation in the use of interventions in high-risk obstetrics," by Dr. Bronstein, Suzanne P. Cliver, B.A., and Dr. Goldenberg, in the February 1998 Health Services Research 32(6), pp. 825-839.

Return to Contents

Outcomes/Effectiveness Research

Using relatively inexpensive antibiotics to treat pneumonia does not lead to poorer medical outcomes

Pneumonia caused by bacteria is typically treated with antibiotics, which range in price from less than $1 to over $100 a dose. But less expensive antibiotics seem to work just as well as more costly ones, according to a study conducted by the Pneumonia Patient Outcomes Research Team (PORT) led by Wishwa N. Kapoor, M.D., M.P.H., of the University of Pittsburgh, and supported by the Agency for Health Care Policy and Research (HS06468). The researchers compared patterns of antibiotic use at five medical institutions to assess the relationship between antibiotic use, antimicrobial costs, and medical outcomes in 927 outpatients and 1,328 hospitalized patients with community-acquired pneumonia.

Each institution varied significantly in its use of 17 of 23 oral antibiotics in outpatients and 18 of 20 parenteral (injected) agents used in hospital patients. The overall median cost of antimicrobial therapy was $12.90 for outpatients (range of $10.80 to $58.90) and $228.70 for inpatients (range of $183.70 to $315.60) among treatment sites. On an institutional level, the use of less costly antibiotics did not lead to poorer medical outcomes. Death and hospital readmission rates for inpatients were not significantly different across sites after adjustments were made for differences in patient demographic characteristics, other coexisting medical problems, and illness severity.

In fact, subsequent hospitalization rates for outpatients and rehospitalization rates for inpatients were lower among the treatment sites with the lowest antibiotic costs. Hospitals that restricted use of expensive antibiotics to extreme medical circumstances or required prior approval for their use by an infectious disease specialist used fewer expensive antibiotics than less restrictive hospitals.

For more information, see "Variations in antimicrobial use and cost in more than 2,000 patients with community- acquired pneumonia," by Kenneth Gilbert, M.D., M.P.H., Patrick P. Gleason, Pharm.D., Daniel E. Singer, M.D., and others, in the January 1998 American Journal of Medicine 104, pp. 17-27.

You can help us show how health services research makes a difference

We in the health services research community already know that our work makes a difference; now, we need to enhance awareness and understanding outside of our community to inform policymakers, providers, purchasers, the business community, and consumers. The easiest way to do this is to present them with examples of health services research findings "in action." Simply put, we are looking for user stories to illustrate the far-reaching and beneficial effects of our research.

Examples include both products—such as decision aids for use in emergency rooms, software programs that can detect medication errors, treatment algorithms for use in physicians' offices, and data products that policymakers and others can use—and other, less concrete effects—such as new ways of communicating with patients and information on the effectiveness and outcomes of early treatment for specific conditions.

Tell us how findings from your AHCPR-funded research are being used. Please send your user stories to Mary L. Grady, Managing Editor, Research Activities.

Stroke PORT examines post-stroke risks and outcomes of carotid endarterectomy

Stroke affects 500,000 Americans each year. Currently, there are few reliable guidelines for identifying and managing patients at risk for stroke. As a result stroke-related clinical practices vary significantly. The Stroke Prevention Patient Outcomes Research Team (PORT), led by David Matchar, M.D., of Duke University, and supported by the Agency for Health Care Policy and Research (PORT contract 290-91-0028), is investigating clinical strategies to prevent stroke. The PORT recently published the two studies summarized here.

The first study identifies factors that increase the risk of death and recurrent stroke following a first stroke. The second study, also supported by an AHCPR National Research Service Award training grant (T32 HS00044), shows that despite resource-conserving changes in management of carotid endarterectomy, outcomes remain about the same.

Petty, G.W., Brown, R.D., Whisnant, J.P., and others (1998, January). "Survival and recurrence after first cerebral infarction." Neurology 50, pp. 208-216.

Within 1 year of a first stroke, 27 percent of those affected are at risk for death, and 12 percent are at risk for another stroke. Five years after the first stroke, 53 percent are at risk for death, and 29 percent are at risk for another stroke. The strongest predictors of death after the first stroke are congestive heart failure and ischemic heart disease. Persistent atrial fibrillation is a somewhat less powerful determinant of death, but more important than hypertension, diabetes, the patient's sex, and preceding transient ischemic attack (often called mini-stroke). Aside from older age, diabetes is the single most important predictor of stroke recurrence, according to this study.

The Stroke Prevention PORT researchers used a model to determine independent predictors of and temporal trends in survival and recurrent stroke among 1,111 residents of Rochester, MN, who had a first cerebral infarction (stroke) between 1975 and 1989. Stroke as a cause of death subsequent to first stroke declined over the study period. However, deaths due to other causes, such as pulmonary embolism or pneumonia, increased. This suggests that stroke may have become less lethal, but that the disabled survivors are more apt to succumb to pulmonary embolism, pneumonia, or other non-neurologic conditions.

Holloway, R.G., Witter, D.M., Mushlin, A.I., and others (1998, January). "Carotid endarterectomy trends in the patterns and outcomes of care at academic medical centers, 1990 through 1995." Archives of Neurology 55, pp. 25-32.

In carotid endarterectomy (CEA), plaque that is blocking one or more of the carotid arteries is surgically removed to prevent stroke. Even though CEA is being offered to increasingly older patients, resource-conserving changes in the hospital management of CEA patients have not affected clinical outcomes, concludes this study.

The researchers retrospectively analyzed hospital discharge data on 7,019 patients undergoing CEA from 1990 through 1995 at 10 U.S. academic medical centers. Over the 6-year study period, the number of CEA procedures performed more than doubled, and the percentage of hospital admissions for patients 65 years or older increased from 65 percent to 75 percent. The mean and median length of stay was cut in half, and the percentage of admissions with transfers to the intensive care unit decreased from 56 percent to 26 percent of cases.

In addition, the percentage of cases with a cerebral arteriogram performed during the same admission but prior to the day of the CEA decreased from 52 percent to 27 percent.

Despite these CEA management changes, there were no trends in inpatient deaths, discharge to an institution, or 30-day hospital readmission rate. There were no significant trends indicative of poorer quality of care as measured by the frequency of secondary diagnoses or postprocedure diagnostic test use. This suggests no detrimental outcomes related to different inpatient management policies that limited hospital resource use, conclude the researchers.

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