Screening more people for cancer may mean less frequent screening for certain cancers

Screening more persons for breast, cervical, and colorectal cancers is a year 2000 goal of the United States. But meeting this goal means a tradeoff in terms of money and resources spent. Less frequent screening among the currently screened population would be necessary in order to increase the overall number of individuals screened. For colon and cervical cancer, this makes sense, but not for breast cancer, concludes this analysis by Graham A. Colditz, M.D., and his colleagues at Harvard University. In a study supported in part by the Agency for Health Care Policy and Research (HS07038), they examined criteria currently used to evaluate the effectiveness of screening for cancer.

Screening for cervical and colon cancer is meant to disclose precancerous lesions. The interval from a negative screening test to a subsequent new cancer is certainly longer than the interval between a negative mammogram for breast cancer and a later mammogram because this screening is for malignancies. Screening programs for cervical cancer are so effective that annual and biennial screens are unnecessary for women at average risk, in part because progression from the detectable, precancerous state to cancer is slow. Reallocating resources away from annual Pap smears should be an explicit public health policy, as should the emphasis on expanding coverage in the unscreened population, recommend the researchers.

Similarly, fecal occult blood testing or flexible sigmoidoscopy, recommended either annually by some to every 3 to 5 years by others, aims to detect cancer at an early stage or to detect and remove colon polyps, which are precursor lesions that over time will progress to colon cancer if left untreated. Screening less frequently but screening more of the population will result in fewer cases of cancer and fewer deaths, according to the researchers. On the other hand, mammography is moderately effective in decreasing breast cancer deaths among women aged 50 and older. When to start screening women and how often to screen are still being debated. But because the benefit from having a mammogram substantially dissipates beyond 2 years, the researchers do not propose any decrease in screening frequency for women already screened in order to increase the number of women screened. Yet expanded population coverage is needed to detect breast cancer, as only 48 percent of women have had a mammogram in the last 3 years.

Details are in "Cancer incidence and mortality: The priority of screening frequency and population coverage," by Dr. Colditz, David C. Hoaglin, Ph.D., and Catherine S. Berkey, D.Sc., in The Milbank Quarterly 75 (2), pp. 147-173, 1997.

Improvements are needed in the provision of preventive health services

Despite the establishment of rather modest national goals to deliver preventive care services and guidelines that recommend specific screening tests, current rates of many preventive services could be substantially improved. The following two studies, led by Leif I. Solberg, M.D., of the Group Health Foundation and supported by the Agency for Health Care Policy and Research (HS08091), examine factors affecting the provision of preventive services.

The first study suggests that the positive attitudes of physicians and nurses about providing preventive care do not affect actual provision of these services. The second study shows that physicians recommend preventive services at the same rate for all patients, regardless of income or insurance status. Yet, poor patients are 20 to 30 percent less likely to be up to date on vaccinations and most screening tests.

Solberg, L.I., Brekke, M.L., and Kottke, T.E. (1997, May). "How important are clinician and nurse attitudes to the delivery of clinical preventive services?" Journal of Family Practice 44(5), pp. 451-461.

There is little association between provider attitudes about preventive health services and the rate at which these services are provided, according to the findings of this study. Despite providers favorable attitudes toward providing preventive services, only 6 to 29 percent of patients needing particular preventive services, such as mammograms or vaccinations, received recommendations for them during an office visit. The exceptions were blood pressure measurements—nearly always part of every doctor's visit—and stop-smoking advice. Changing clinician attitudes may not be the key to improving delivery of preventive health services, conclude the researchers. They suggest instead that the focus should be on altering the environmental barriers which interfere with providing preventive services, such as building supportive office systems.

The researchers surveyed 647 clinic physicians, midlevel practitioners, and nurses in 44 primary care clinics contracting with health maintenance organizations about their attitudes toward preventive care. They also surveyed 6,830 patients visiting the clinics about their receipt of eight specific preventive care services: Pap smears, mammograms, breast exams, hypertension screening, cholesterol screening, stop-smoking advice, influenza shots, and pneumonia shots.

Most providers believed that preventive services were an important component of their practices and were willing to work on improving these services. They also supported the use of teams to work on problems or devise systematic ways to remind providers of the preventive services needed by individual patients. Despite these favorable attitudes, few patients were offered cholesterol measurement (7 percent), a breast exam (15 percent), mammogram (24 percent), Pap test (29 percent), influenza immunization (26 percent), or pneumonia immunization (6 percent), although 48 percent were advised to quit smoking.

Solberg, L.I., Brekke, M.L., and Kottke, T.E. (1997, May). "Are physicians less likely to recommend preventive services to low-SES patients?" Preventive Medicine 26, pp. 350-357.

According to this study, primary care physicians at 22 clinics that contract with health maintenance organizations recommend preventive screening tests and services at the same rate for all patients, regardless of their income or insurance status. Yet, with the exception of blood pressure readings, low-socioeconomic-status (SES) patients are substantially less likely to be up to date on immunizations and screening tests than other patients. This is despite the patients' similar interest in receiving these services, note the researchers. They based their findings on a survey of 4,245 patients (39.7 percent low-SES patients) within 2 weeks of a visit to one of the study clinics.

Results showed that all the preventive services needed by clinic patients were recommended at essentially the same rates, regardless of their SES. In fact, low-SES patients reported more queries about whether they smoked and advice about quitting smoking than other patients. Nevertheless, low-SES patients had 20 percent lower mean clinic rates for cholesterol testing within 5 years than other patients; 26 percent had lower rates for a breast exam within 2 years, a flu shot in the past year, or ever having had a pneumonia shot; and 9 percent had lower rates for a Pap smear within 2 years. If these differences are not due to differential efforts by physicians to encourage patients to obtain these services or to actually perform them, the researchers are puzzled about what might account for the differences.

Age, weight, and back injury influence the likelihood of chronic back disability

The number of persons with disabling low-back pain has increased dramatically in recent years, with about 5 percent of the U.S. population now affected. If you are older than 34 years, overweight, and have a history of back trauma, you are more likely to have chronic back disability, says a new study.

Using data from the 1989 National Health Interview Survey, the researchers correlated back problems in a representative sample of the adult population of the United States. Workers who were aged 35 to 54, had not graduated from high school, had disabling non-back medical problems, and had weight above the 50th percentile, were relatively more likely than others to have a disabling back condition. Those in technical, sales, clerical, private household, service, precision production and repair, and transportation occupations were more prone to this problem than professional workers. However, occupation was not correlated with nondisabling back conditions.

Both acutely disabled (at least one back-related restricted activity day in the preceding 2 weeks) and chronically disabled persons tended to be male, minority, less educated, never married, and unemployed. Age over 34 years, weight above the 50th percentile, and history of back trauma correlated only with chronic back disability. People working in truck driving, nursing, mining, customer service, construction, mechanics, and retail sales may have a high prevalence of disabling back problems because they are exposed to specific physical/work activities that increase their risk for back pain and subsequent disability. These include lifting, twisting, lifting while twisting, sitting, standing, driving, pulling, pushing, carrying, lowering, bending, stretching and reaching, and exposure to vibration.

While other studies have shown that various factors predict chronic back pain, this is the first study to compare predictors of long-term chronic and acute back disability, note Eric L. Hurwitz, D.C., Ph.D., and Hal Morgenstern, Ph.D., of the University of California, Los Angeles School of Public Health. Their work was supported by the Agency for Health Care Policy and Research (HS07968).

For more details, see "Correlates of back problems and back-related disability in the United States," by Drs. Hurwitz and Morgenstern, in the Journal of Clinical Epidemiology 50(6), pp. 669-681, 1997.

Return to Contents

Medical Informatics

Patient confidentiality must be safeguarded before life-saving information can be transmitted via the Internet

Emergency physicians treating an unconscious patient need to know whether the patient has allergies to certain medications, has any chronic health problems, or is currently taking medication. But these facts, which can be critical to saving a person's life, are difficult to obtain without quick access to a patient's full medical record. The Boston Electronic Medical Record Collaborative is working to solve this problem. It has developed the Word Wide Web Electronic Medical Record System (W3-EMRS) that will use the Internet and the World Wide Web to transfer hospitals' computer-based patient information to the emergency departments (EDs) of participating institutions.

With support from a cooperative agreement between the Agency for Health Care Policy and Research (HS08749) and the National Library of Medicine (LM05877), Charles Safran, M.D., F.A.C.P., of Harvard Medical School, and his colleagues recently proposed a system to maintain the confidentiality and security of patient information in an online system such as the W3-EMRS. Advances in Internet and Web security needed for financial transactions will be adequate to protect patient information during the transmission process, according to lead author David M. Rind, M.D., of Beth Israel Deaconess Medical Center in Boston. The most common threat to confidentiality is the inappropriate accessing of information by authorized providers.

The proposed protocol for the protection of confidentiality addresses a single important scenario—the treatment of a patient in the ED—in which patient care would be improved through the interinstitutional transfer of records. In the absence of exceptional circumstances, explicit patient consent must be obtained but would not be needed in serious medical situations where the doctrine of implied consent would apply if the patient were incapable of communicating consent. Both the patient's and the provider's identities must be authenticated, including a secure password by provider institutions. Patient consent must be verified, emergency need for access must be confirmed, integrity of data of the medical record must be maintained, and the patient has the right to review any data released. Clinicians who request information are required to have a hardware device that generates a time-limited password, so that breaches in confidentiality can be identified and rectified.

Details are in "Maintaining the confidentiality of medical records shared over the Internet and the World Wide Web," by Dr. Rind, Isaac S. Kohane, M.D., Ph.D., Peter Szolovits, Ph.D., and others, in the July 15, 1997, Annals of Internal Medicine 127, pp. 138-141.

Barriers continue to hinder integration of medical technology into home health care

In 1995, home health care expenditures soared over $2 billion but still accounted for only 3 percent of total U.S. spending on health care. Medical technology, including teleconsulting and teleradiology, can substantially enhance access to and delivery of home health care services. For instance, electronic home visits can improve the timeliness of medical interventions. But much has to be resolved before medical technology can be integrated into home health care, comments Diane L. Adams, M.D., M.P.H., of the Agency for Health Care Policy and Research, in a recent article.

Numerous industry and technical design issues have hindered the proliferation of and ability to deliver telemedicine applications into the mainstream of home health care. The lethargy with which telemedicine has been adopted is more indicative of historic business practices and restrictions within the health care industry than any technology limitations imposed on it, according to Dr. Adams. These barriers include variable, outdated, and often excessive long-distance telecommunication rates; concerns about liability; licensure of physicians and other providers within specific States when telemedicine may cross State lines; software and hardware incompatibility; absence of minimal standards for the practice of telemedicine; and the lack of reimbursement for telemedicine in most areas.

The author points out that several issues and their resolution will guide the progress of telecommunications in health care. These include the management of fiscal and political risks by Federal and State governments, public policy, and training of professionals to exploit the technology once it is deemed fiscally viable. Although telemedicine's original mandate was to improve access to health care in rural areas, it has now evolved to target the entire population. A current project involving the U.S. Department of Defense is focusing on developing an electronic house call system that integrates into home televisions.

For more information, see "Home healthcare: A new venue for telemedicine," by Dr. Adams, Kenneth L. Seymens, Gloria Rookard, R.N., P.N.P., and Brenda A. Leath, M.H.S.A., in World Medical Technology Update, 1997/1998, edited by Brian Kellock, London: Kensington Publications, Ltd., pp. 226-228, 1997. Reprints (AHCPR Publication No. 97-R088) are available from the AHCPR Publications Clearinghouse.

Return to Contents

Special Populations

Stressful life events linked with use of drugs, alcohol, and cigarettes by poor women during pregnancy

About 7 percent of poor black and Hispanic women who receive prenatal care at community clinics continue to use drugs during pregnancy, which is comparable to national estimates (5.5 percent to 11 percent). Black women are much more apt to use substances during pregnancy than Hispanic women, but stress and anxiety, not ethnicity, are highly associated with substance use during pregnancy, according to a recent study supported by the Agency for Health Care Policy and Research (HS05518). It found that women who continued to use substances during pregnancy, regardless of ethnicity, experienced twice as many stressful life events as women who did not use substances.

Ruth E. Zambrana, Ph.D., of George Mason University, and Susan C.M. Scrimshaw, Ph.D., of the University of Chicago, prospectively studied low-income, black (255), Mexican-American (525), and Mexican-immigrant (764) women visiting 22 prenatal care clinics in Los Angeles. They interviewed the women about prenatal life events, anxiety, sources of support, and their use of substances in the 3 months before they became pregnant and during pregnancy. Black women were more likely than Mexican-American or Mexican-immigrant women to be heavy users of alcohol (5.5 percent compared with 1.7 percent and .7 percent, respectively); to have used drugs (23.2 percent vs. 7.6 percent and 1.2 percent, respectively); and to have smoked cigarettes (34.6 percent vs. 27.8 percent and 18.8 percent, respectively).

Black women were the least likely to be employed full time or to live with the baby's father, and they were most likely to receive public health insurance. They also reported more stressful life events, such as death or injury of someone close, problems with alcohol and drugs, problems at work, an arrest, or problems with government agencies; more distress from these events; and higher anxiety than Hispanic women. These findings confirm other study results that link multiple prenatal risk factors to women who use substances during pregnancy.

See "Maternal psychosocial factors associated with substance use in Mexican-origin and African-American low-income pregnant women," by Drs. Zambrana and Scrimshaw, in the May/June 1997 Pediatric Nursing 23(3), pp. 253-259, 274.

Return to Contents
Proceed to Next Section