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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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

Marriage encourages healthy behaviors among the elderly, especially men

Marriage has a positive effect on the healthy behaviors of elderly people, and that effect is larger for elderly men than for elderly women. Previous studies have shown a positive relationship between marital status and health, but a new study by researchers at the Agency for Health Care Policy and Research demonstrates that marriage continues to have positive health benefits regardless of the age of the couple. The study focuses specifically on preventive behaviors practiced by the elderly.

The researchers, Barbara Schone, Ph.D., and Robin Weinick, Ph.D., looked at five health behaviors among the elderly and compared these behaviors for widowed and married men and women. They found that elderly people who are married are more likely to eat breakfast, wear seat belts, engage in physical activity, have their blood pressure checked, and not smoke than elderly men and women who are widowed. The researchers also found that the benefits of marriage tend to be more substantial for elderly men than for elderly women.

Drs. Schone and Weinick believe their research provides information that will be useful in identifying elderly people who may need the most encouragement to engage in these behaviors. In addition to marital status and sex, they found that the social and psychological circumstances of elderly people can help point out which of them are least likely to adopt healthy behaviors. The researchers analyzed data from the 1987 National Medical Expenditure Survey, now called the Medical Expenditure Panel Survey (MEPS), for this study.

For more information, see "Health-related behaviors and the benefits of marriage for elderly persons" in the October 1998 issue of The Gerontologist 38(5), pp. 618-627.

Reprints (AHCPR Publication No. 99-R007) are available from the AHCPR Publications Clearinghouse.

Return to Contents

Outcomes/Effectiveness Research

Even modest improvements in blood sugar control can substantially improve quality of life for people with diabetes

A recent study may alter how health care providers view treatment of people with adult onset (type 2) diabetes. Until now, physicians have typically sought to tightly control the blood sugar (glycemic) levels of patients with type 2 diabetes in order to prevent or reduce long-term complications of the illness, such as blindness and kidney disease. However, the new study shows for the first time that even relatively modest improvements in glycemic control can immediately improve quality of life for these patients by relieving fatigue, weakness, and difficulty with memory and mental tasks.

In this study, patients who received medication and dietary management had better glycemic control, greater physical and emotional well-being, fewer physician visits, and fewer missed days of work and restricted activities than those treated with diet alone. In fact, patients maintained on diet alone were nearly five times as likely to miss a day of work as those on diet and medication, costing employers an additional $91 monthly per employee in paid sick days, note Marcia Testa, M.P.H., Ph.D., of the Harvard School of Public Health, and Donald Simonson, M.D., of Harvard Medical School. Their study, which was supported in part by the Agency for Health Care Policy and Research (HS07767), focused on the 90 percent of diabetics who have type 2 diabetes. In this type of diabetes, the body is unable to make a sufficient amount of insulin or cannot properly use the insulin it does produce.

The researchers studied 594 type 2 diabetes patients at 62 clinical centers across the United States. Patients were randomized to receive treatments involving either placebo and diet management or daily doses of glipizide GITS (glipizide gastrointestinal therapeutic system—a drug that helps the body produce more insulin) and diet management. Physicians made periodic clinical assessments, and patients completed a questionnaire about their mental and physical functioning and reported days missed from work and days of reduced activity due to ill health.

For more information, see "Health economic benefits and quality of life during improved glycemic control in patients with type 2 diabetes mellitus" by Drs. Testa and Simonson, in the November 4, 1998, Journal of the American Medical Association 280(17), pp. 1490-1496.

Variations in provider characteristics, procedure use, and cost of care do not affect pneumonia outcomes

Elderly patients hospitalized with community-acquired pneumonia who are treated in urban teaching hospitals with specialists, high-technology procedures, and expensive resources fare no better than those treated at rural hospitals that do not have specialists or sophisticated procedures. In fact, the outcomes for these pneumonia patients are unrelated to hospital teaching status or location, physician specialty, number of procedures used, or cost of care, concludes a study supported in part by the Agency for Health Care Policy and Research (HS06468).

High-technology care provided by specialists has been shown to improve outcomes for certain illnesses. However, it is less clear that this approach can improve outcomes of patients who are treated for more "routine" illnesses such as pneumonia, explains Wishwa N. Kapoor, M.D., M.P.H., of the University of Pittsburgh Medical Center. Dr. Kapoor is principal investigator of the AHCPR-supported Patient Outcomes Research Team (PORT) on community-acquired pneumonia.

The researchers used diagnostic data from the Health Care Financing Administration to identify 21,194 Medicare patients hospitalized with pneumonia. They controlled for patient demographic characteristics and pneumonia etiology while examining the impact of provider characteristics on patient 30-day death rates and hospital readmission rates.

Patients treated at urban hospitals were more likely to obtain either a chest computed tomography scan (odds ratio, OR, of 1.64; 1 is equal odds) or a bronchoscopy (OR, 1.32) and pulmonary and infectious disease consults (OR, 3.49 and 3.11, respectively), and they had 15 percent higher costs than comparable patients treated in rural hospitals. Similarly, patients at teaching hospitals had 11 percent higher costs than those at nonteaching hospitals. Physician specialty was not independently associated with procedure use. However, when a patient's attending physician was a general internist or medical subspecialist, costs were 9 percent and 8 percent higher, respectively, than for a patient whose attending physician was a family practitioner.

See "Relationship of provider characteristics to outcomes, process, and costs of care for community-acquired pneumonia," by Jeff Whittle, M.D., M.P.H., Chyongchiou Jeng Lin, Ph.D., Judith R. Lave, Ph.D., and others, in Medical Care 36(7), pp. 977-987, 1998.

Researchers compare ways of treating low back pain

A comparison of different methods of treating low back pain supported by the Agency for Health Care Policy and Research (AHCPR grant HS07915) found no significant differences in outcomes among patients who were either treated by chiropractic spinal manipulation, by a popular form of physical therapy called the McKenzie method, or simply given an educational booklet and not treated.

According to the study, the chiropractic and McKenzie physical therapy patients were much more satisfied with their treatment than were the patients given the booklet. However, the short-term pain relief experienced by the chiropractic and McKenzie physical therapy patients was only marginally better than that of the no-treatment group. Furthermore, there were no significant differences among the three groups at any of the followup points (1, 4, 12, and 52 weeks after initial treatment) in their ability to function or in disability days spent in bed, home from work or school, or cutting down on usual activities. In addition, neither chiropractic manipulation of the spine nor McKenzie physical therapy reduced recurrence of back pain, subsequent visits, or costs of back care.

The study was based on 321 adult low back pain patients aged 20 to 64 seen in a primary care clinic who were randomly selected to undergo spinal manipulation by a chiropractor, see a physical therapist for McKenzie therapy, or simply be given a booklet. The study was led by Daniel C. Cherkin, Ph.D., of the Group Health Center for Health Studies—the research arm of the Group Health Cooperative of Puget Sound—a large health maintenance organization. Dr. Cherkin is also affiliated with the departments of health services and family medicine of the University of Washington, Seattle.

This study is one of several funded by AHCPR to examine the effectiveness and cost-effectiveness of the various ways used to treat low back pain, a problem that may affect up to half of all Americans in any given year and costs an estimated $50 billion to $100 billion annually, half of which goes to direct medical care.

For detailed findings and conclusions of the current study, see "A comparison of physical therapy, chiropractic manipulation, and provision of an educational booklet for the treatment of patients with low back pain," in the October 8, 1998, issue of the New England Journal of Medicine 339(15), pp. 1021-1029.

Low Birthweight PORT publishes latest findings

Nearly 70 percent of all infant deaths and about one-third of all handicapping conditions are associated with low birthweight (less than 2,500 grams or 5.5 pounds), which is caused by preterm birth and intrauterine growth retardation. Some minority groups, especially blacks, are at relatively high risk for low birthweight (LBW) infants. The Low Birthweight Patient Outcomes Research Team (PORT) focuses on the multiple causes of both full-term and preterm low birthweight. The PORT 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).

In two recently published articles, the researchers review approaches to preventing preterm birth and examine the adverse effects of short intervals between pregnancies on birth outcomes. A third article discusses how inclusion in State birth records of infants weighing less than a pound, most of whom die shortly after birth or are stillborn, masks strides made in improving neonatal survival. These articles are briefly summarized here.

Goldenberg, R.L., and Rouse, D.J. (1998, July). "Prevention of premature birth." New England Journal of Medicine 339(5), pp. 313-320.

The rate of preterm births in the United States is not declining. Most interventions designed to prevent preterm birth do not work, and the few that do—such as treatment of urinary tract infection, closure of a structurally weak cervix, and treatment of bacterial vaginosis in high-risk women—are not universally effective. Also, they are applicable to only a small percentage of women at risk for preterm birth. Without a better understanding of the mechanisms leading to preterm birth, substantial reductions in preterm delivery are unlikely to be achieved, conclude the researchers after reviewing the literature on this topic.

They found that making prenatal care available to more women or making more prenatal visits available generally has not reduced the number of preterm births. Programs ranging from those for cessation of tobacco, alcohol, and/or drug use to psychological counseling and nutritional supplementation among well-nourished women of normal body weight also have not reduced the rate of preterm births. Even home uterine monitoring for early labor contractions, bed rest, and hydration have not proven effective in this regard. Tocolytic drugs that interrupt or stop uterine contractions may help delay delivery up to 24 hours.

An incompetent, or structurally weak, cervix is diagnosed in 1 in 200 to 1 in 1,000 pregnant women with a history of second trimester preterm birth. The traditional treatment has been placement of one or several circumferential stitches (cerclage) in the cervix, which in one study showed a significant reduction in the rate of preterm birth before 33 weeks. Up to 80 percent of early preterm births are associated with an intrauterine infection that precedes the rupture of membranes. Also, symptomatic and asymptomatic urinary tract infections are associated with increased risk of preterm delivery. Since so many premature births are related to infection, antibiotic treatment has great promise for reducing early spontaneous preterm births, conclude the researchers.

Klerman, L.V., Cliver, S.P., and Goldenberg, R.L. (1998, August). "The impact of short interpregnancy intervals on pregnancy outcomes in a low-income population." American Journal of Public Health 88(8), pp. 1182-1185.

The shorter the time interval between pregnancies, the greater the likelihood of preterm birth in the second pregnancy, finds this study. It shows that women with only 13 weeks between delivery of the first child and conception of a second child had nearly double the rate of preterm delivery of the second child than women who did not get pregnant again until 104 weeks or longer after the first child was born. As the interval between pregnancies increased, there was a significant decrease in the rate of preterm delivery among women who had previously delivered a term baby. The risk of preterm birth decreased from 20 percent at the shortest interval (less than 13 weeks) between pregnancies to 16 percent at 13 to 25 weeks, 12 percent at 26 to 51 weeks, 10 percent at 52 to 103 weeks, and 11 percent at 104 weeks or longer.

Women, particularly women who are poor and young, should be advised of the potential harm to their infants of short interpregnancy intervals, recommend the Low Birthweight PORT researchers. They analyzed pregnancy intervals between first and second pregnancies in 4,400 women, most of whom were poor and on Medicaid, who had received prenatal care in Alabama county clinics between 1980 and 1990. The researchers used data drawn from the Obstetrical Automated Record System, a computerized patient information system of data on all women who delivered babies at the university or county hospital in Birmingham.

Phelan, S.T., Goldenberg, R., Alexander, G., and Cliver, S.P. (1998, August). "Perinatal mortality and its relationship to the reporting of low-birthweight infants." American Journal of Public Health 88(8), pp. 1236-1239.

In this study, the researchers examined changes in the reporting of infants born weighing less than 500g (1.1 pounds) in Alabama from 1974 to 1994 and the impact of the changes on perinatal mortality rates. During this time, the percentage of total newborns in the State weighing less than 500g increased by 155 percent, compared with a 25 percent increase in births of infants weighing 500g to 999g (2.2 pounds) and 7 percent among infants weighing 1,000g to 1,499g (3.3 pounds). As a result, 32 percent of neonatal deaths were attributable to newborns weighing below 500g compared with 3 percent previously.

This reporting of such extremely small babies masked the strides made in improving neonatal mortality in Alabama during the study period, say the authors of the study. For instance, mortality rates only dropped from 100 percent to 92 percent in the under-500-g group, for whom survival is tenuous at best. Yet the neonatal mortality rate declined from 86 percent to 31 percent for infants weighing 500g to 999g and from 39 percent to 4 percent for infants weighing from 1,000g to 1,499g.

The elimination of live-born infants weighing less than 500g from the calculations in 1994 would reduce the neonatal mortality rate by 32 percent and the stillbirth rate by 29 percent. Reporting of these births will continue to mask the full extent to which neonatal mortality and stillbirth rates are decreasing, conclude the authors. They linked Alabama live birth, infant death, and fetal death records for selected years from 1974 through 1994.

Prostate PORT studies review prostate cancer screening and treatment practices

The value of early detection of prostate cancer through prostate-specific antigen (PSA) testing is one of the great controversies in medicine. Two studies recently published by the Prostatic Diseases Patient Outcomes Research Team (PORT) show that both primary care physicians and urologists recommend PSA testing and believe that aggressive treatment of prostate cancer is more beneficial than existing evidence indicates. Primary care physicians, in particular (in contrast to urologists), tend to recommend PSA testing for older men, who are least likely to benefit from early cancer detection, and for men with symptoms of benign prostatic hyperplasia, who are most likely to have a false-positive PSA result.

An accompanying editorial notes that many patients and their physicians overestimate the likelihood of getting prostate cancer, dying of prostate cancer, and receiving benefits from early detection and treatment. At the same time, they underestimate the risks associated with early detection and intervention. The Prostate PORT, led by Michael J. Barry, M.D., of Massachusetts General Hospital, is supported by the Agency for Health Care Policy and Research (HS08397). The two articles and an accompanying editorial are summarized here.

Fowler, Jr., F.J., Bin, L., Collins, M.M., and others (1998). "Prostate cancer screening and beliefs about treatment efficacy: A national survey of primary care physicians and urologists." The American Journal of Medicine 104, pp. 526-532.

This national survey of 444 primary care physicians (PCPs) and 394 urologists found that most PCPs do PSA tests during routine examination of men older than 50 years of age. Most PCPs continue to do PSAs on patients over 80 years and to refer men with abnormal PSA values for biopsy. Only a minority of urologists, however, would recommend a PSA test or biopsy for abnormal values for men over 75 years of age.

More than 80 percent of PCPs and urologists doubt the value of radical prostatectomy for men with less than 10 years of life expectancy. However, more PCPs than urologists see probable survival benefit in radiation therapy for patients with life expectancy of less than 10 years (48 vs. 36 percent) and more than 10 years (67 vs. 53 percent). Thirteen percent of PCPs and only 3 percent of urologists consider watchful waiting to be as appropriate as aggressive therapy for men with more than 10 years of life expectancy.

The rationale for screening and early treatment is that untreated prostate cancer may grow locally, metastasize, or lead to death, and that early detection and appropriate treatment will delay or prevent these events. However, this has not yet been proven by controlled trials.

Meigs, J.B., Barry, M.J., Giovannucci, E., and others (1998). "High rates of prostate-specific antigen testing in men with evidence of benign prostatic hyperplasia." The American Journal of Medicine 104, pp. 517-525.

Men with urinary tract symptoms suggestive of benign prostatic hyperplasia (BPH) typically have higher PSA levels than similar-aged asymptomatic men, but they are not at higher risk for prostate cancer. Nevertheless, these men, who are more likely to have a false-positive PSA result, are often targeted by physicians for PSA testing, according to this study. Also, physicians often recommend PSA screening for older men, who are least likely to benefit from early cancer detection (due to shorter life expectancy and a higher risk of treatment complications).

Physician and patient education are needed to promote more rational and selective use of this screening test, the researchers conclude. They used self-reported histories of PSA testing in 1994 to examine the prevalence and correlates of this testing among 33,028 U.S. health professionals (aged 47 to 85 years) without prostate cancer. In 1995, a subset of 7,070 men provided additional information on diagnosis and treatment of BPH. PSA testing in the prior year was reported by 39 percent of men in their 50s to 53 percent of men in their 80s. Men were more apt to report PSA testing if they had lower urinary tract symptoms characteristic of BPH, a prior history of transurethral prostatectomy, or a physician diagnosis of BPH.

Wilt, T.J. (1998). "Prostate cancer screening: Practice what the evidence preaches." The American Journal of Medicine 104, pp. 602-604.

The current "epidemic" of routine PSA testing regardless of age, medical condition, or scientific evidence has created a generation of patients, families, physicians, and media focused on PSA values. However, routine screening for any condition is not warranted without evidence that the test accurately detects early disease, that early detection improves outcomes, and that benefits outweigh harms. Although PSA testing eventually may be proven to save lives and avert future cancer-related illness, such information is currently lacking, asserts the author of this editorial.

Certain U.S. groups and evidence-based groups from other countries have concluded that prostate cancer screening should not be routinely performed. Physicians may find it difficult to take the time required to inform patients in detail about prostate cancer screening. Routinely ordering a PSA test may seem like an easy option. But this author contends that a better alternative is to use fact sheets and videotapes to facilitate the delivery of unbiased information to patients. Men who are well informed about the known risks and uncertain benefits are less likely to be tested or desire surgery or radiation than men who are not well informed.

AHCPR along with other agencies has initiated randomized trials that aim to determine whether early detection and treatment of prostate cancer reduce morbidity and mortality. In the meantime, suggests this author, physicians should inform men about the risks and uncertain benefits of PSA screening and should not recommend it in men with a life expectancy less than 10 years. Also, men with lower urinary tract symptoms should be informed that they are not at greater risk for prostate cancer than men without symptoms, and that their PSA test is more apt to result in an elevated PSA level due to benign prostate disease. Finally, men should be informed that there is no consensus on the preferred therapy for clinically localized prostate cancer.

Return to Contents


Children's death rates fall sharply in intensive care units

A child given intensive care for a serious respiratory illness or other life-threatening disease has a significantly better chance of surviving now than he or she would have had as recently as a little over a decade ago, according to a recent study, which was jointly sponsored by the Agency for Health Care Policy and Research and the Health Resources and Services Administration (HRSA). The study indicates that the death rate of children admitted to hospital pediatric intensive care units (ICUs) for such killer diseases as asthma, bronchitis, and pneumonia declined by 45 percent between the early to mid-1980s and 1993.

According to the researchers, who were led by John M. Tilford, Ph.D., of Arkansas Children's Hospital in Little Rock, the study suggests that the falling death rates are the result of improvements over time in the quality of intensive care, such as better ventilation technology. These findings clearly show the payoff from investing in better quality of care, notes Dr. Tilford. Because of improvements in quality of care, children who would have died 15 years ago can now survive and mature into healthy adults.

The researchers also examined mortality risk by age, regardless of condition, and found that death rates declined the most for younger children. The death rate of infants less than 1 month old treated in hospital pediatric ICUs dropped by 39 percent, while that of babies 1 to 12 months of age fell by 28 percent. The death rate for children of all ages treated in hospital pediatric ICUs during the study period declined by 15 percent.

Details are in "Differences in pediatric ICU mortality risk over time," by Dr. Tilford, Paula K. Roberson, Ph.D., Shelly Lensing, M.S., and Debra H. Fiser, M.D., in the October 1998 issue of Critical Care Medicine 26(10), pp. 1737-1743.

Cost effectiveness of knee replacement surgery usually increases as surgical volume increases

Knee replacement (KR) is a relatively common and costly surgical procedure performed most often on elderly patients. Except for hospitals performing over 100 KRs per year, the more knee replacement surgeries a hospital performs, the less the associated treatment costs. In fact, the cost savings can amount to as much as 10 percent of the hospital's average KR cost, concludes a study supported by the Agency for Health Care Policy and Research (HS06432).

The research demonstrates that average treatment costs per KR surgery can be as much as 10 percent lower for large hospitals that perform 76 to 100 KRs per year compared with large hospitals that perform only 1 to 12 KRs per year. For example, hospitals that performed from 1 to 12 KRs in 1989 had an average KR treatment cost of $10,088 compared with an average cost of $9,527 for hospitals performing 51 to 75 KRs and $9,419 for those performing 76 to 100 KRs. However, the amount a hospital saved from increased KR volume also depended on its current volume of KRs and hospital size. Small hospitals that performed 25 to 50 KRs during 1989 had a KR treatment cost of $9,164 compared with $9,478 for medium-sized hospitals and $10,494 for large hospitals performing this number of procedures.

These findings are based on analysis of a 1989 national survey of hospitals and 1989 Medicare claims data by the AHCPR-funded Patient Outcomes Research Team (PORT) on total knee replacement, which was led by Deborah A. Freund, Ph.D., of Indiana University. Dr. Freund and her colleagues used data on Medicare beneficiaries who received KR surgery in U.S. acute care hospitals in 1989 to calculate average hospital treatment costs associated with all KR surgeries performed, adjusting for other factors affecting treatment cost, such as complications and case mix differences. They then correlated KR surgery volume and hospital size with KR treatment costs.

See "Does hospital procedure-specific volume affect treatment costs: A national study of knee replacement surgery," by Benjamin Gutierrez, Ph.D., Steven D. Culler, Ph.D., and Dr. Freund, in the August 1998 Health Services Research 33(3), pp. 489-511.

Return to Contents

Heart Disease

Cardiac arrests occur most often in the morning and early evening hours

Cardiac arrests do not occur randomly during the day but typically peak between the hours of 8 and 11 a.m. and again between 4 and 7 p.m., with a low incidence at night. This suggests that cardiac arrests are probably triggered by a person's activity rather than by biochemical processes. In most patients who experience more than one cardiac arrest, the time of the second arrest is unrelated to the time of the first one, explains Monika Peckova, Ph.D., of the University of Washington, Seattle.

In a study supported in part by the Agency for Health Care Policy and Research (HS08197), Dr. Peckova and her colleagues used three methods to assess temporal variations in 5,248 adult out-of-hospital cardiac arrests attended by the Seattle Fire Department that were related to a cardiac condition. They examined circadian (based on a 24-hour time period) variation in the patients according to age, sex, race, heart rhythm, and survivor status. The researchers found that the evening peak was primarily attributed to patients found in ventricular fibrillation (rapid, irregular heartbeat). Arrests that showed other rhythms typically exhibited a morning peak.

Cardiac arrests associated with survival had more pronounced variation in time of day than episodes in which the victims did not survive, even after adjusting for cardiac rhythm. For 597 patients who had at least two separate cardiac arrests, there was no overall association between the times of day of the first and second arrests. For women, on the other hand, the times of day for the first and second arrests were closer to each other than one would expect if the times were entirely unrelated. No difference in circadian variation of cardiac arrests was found according to sex, race, or day of the week. However, elderly patients had different circadian variations than middle-aged patients.

For details, see "Circadian variations in the occurrence of cardiac arrests: Initial and repeat episodes," by Dr. Peckova, Carol E. Fahrenbruch, M.S.P.H., Leonard A. Cobb, M.D., and Alfred P. Hallstrom, Ph.D., in the July 7, 1998, issue of Circulation 98, pp. 31-39.

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