Depression and coexisting illnesses should be treated together to reduce costs and improve outcomes

Older primary care patients who are depressed incur greater average charges for diagnostic tests than patients who are not depressed. These higher charges are not due to depression, per se, but instead can be attributed to the greater burden of coexisting medical illnesses borne by depressed patients compared with patients who are not depressed, according to Christopher M. Callahan, M.D., of the Regenstrief Institute for Health Care, lead author of a recent study.

Older persons, in particular, are more apt to suffer from coexisting illness, to be functionally impaired, or to receive medication that can lead to symptoms of depression or mask existing depression, according to the study, which was supported in part by the Agency for Health Care Policy and Research (HS07632 and HS07763). Thus, improved treatment of mental illness is not the answer to reducing excess use of medical services by patients with mental disorders. Instead, primary care physicians and mental health specialists need to address organic and psychiatric illnesses simultaneously, notes Dr. Callahan.

Dr. Callahan and his colleagues compared the amount and charges for inpatient and outpatient diagnostic testing for 3,767 patients 60 years of age and older who completed testing on the Centers for Epidemiologic Studies Depression Scale (CES-D) during routine office visits at a primary care group practice from 1991 to 1993.

The researchers found that depressed patients were more apt than nondepressed patients to have nonpsychiatric medical problems and more visits to the outpatient care center (nine visits vs. eight visits, respectively), to use the emergency department (52 percent vs. 40 percent), and to be hospitalized (22 percent vs. 17 percent). These patients also incurred $200 more in diagnostic test charges for a 1-year period ($583 vs. $387) than nondepressed patients—a large sum if calculated for an entire population of older adults over several years. The 54 percent difference in charges, mostly for clinical pathology (for example, routine blood and urine tests), persisted into the second year.

See "Association of symptoms of depression with diagnostic test charges among older adults," by Dr. Callahan, Joseph G. Kesterson, M.A., and William M. Tierney, M.D., in the March 15, 1997, issue of Annals of Internal Medicine 126, pp. 426-432.

Primary care physicians are more likely to recognize depression in women and more educated patients

Primary care physicians often fail to recognize depression when it is present in their patients, yet they are less likely to miss it in women and more educated patients than in men and less educated patients, according to a study supported by the Agency for Health Care Policy and Research (HS08029 and HS06167). The study shows that depressed women are twice as apt to be diagnosed as such than depressed men and that for each additional year of education, the odds increase by 24 percent that a patient will be correctly diagnosed as depressed.

Health status measures, such as level of energy and physical and social functioning, declined among patients in this study as their depression symptoms increased. Therefore, Edward J. Callahan, Ph.D., and Klea K. Bertakis, M.D., M.P.H., of the University of California, Davis, and their colleagues suggest that clinicians incorporate health status measures when assessing primary care patients. When a patient demonstrates an unexpectedly sudden decline or progressive deterioration in health status scores that is not accounted for by a physical illness or other medical problem, the clinician should suspect a mental disorder. The clinician can then use other diagnostic techniques and instruments such as the Beck Depression Inventory (BDI) to aid in the correct diagnosis.

The researchers gathered data from 1990 to 1993 on 508 consecutive new patients who were randomly assigned to 105 family practice and general internal medicine residents. Each physician saw an average of five patients. The patients completed the BDI, which measures the presence and severity of 13 symptoms of depression, and the SF-36 health status questionnaire. The researchers compared the test results, which were not revealed to the physicians, with the presence or absence of depression diagnosis on the patients medical charts.

Physicians made a chart notation of depression for only 28 percent of the 130 moderately or severely depressed patients (BDI score of 9 or more). The physicians were more apt to correctly diagnose depressed women and more educated patients, even after adjusting for differences in health status, BDI scores, and other factors. Age and race did not influence recognition of depression by the primary care physicians.

Details are in "Depression in primary care: Patient factors that influence recognition," by Drs. Callahan and Bertakis, Rahman Azari, Ph.D., and others, in the March 1997 issue of Family Medicine 29(3), pp. 172-176.

Return to Contents

HIV/AIDS Research

Primary care physicians fail to identify half of patients at risk for HIV infection

Primary care physicians do not routinely assess patients for HIV risk, even when the patients have obvious risk behaviors for HIV, such as multiple sex partners or intravenous drug use. In fact, they fail to identify risk behaviors for HIV infection or include HIV in the diagnosis of more than half of the patients in whom active HIV infection should be considered, according to a study supported by the Agency for Health Care Policy and Research (HS06454).

Paul G. Ramsey, M.D., principal investigator, and other researchers randomly selected 134 primary care physicians to examine standardized patients (SPs, individuals trained to enact specific case presentations) to assess overall clinical skills, as well as the physicians skill at identifying and managing HIV infection. In a day of testing, physicians saw 13 to 16 SPs with diverse case presentations. Nine SPs were at high risk for HIV infections with unidentified risk behaviors for HIV, which they offered if asked. However, only 12 percent of at-risk SPs were asked about their history of sexually transmitted disease, which increases the risk of HIV infection; 49 percent were asked if they were sexually active; 15 percent were asked about the number of sexual partners they had; 39 percent were asked about injection drug use; and 50 percent of at-risk male SPs were asked about sex with other men.

Individual physicians initiated screening in at least one of the five risk assessment categories with an average of 60 percent of at-risk SPs they saw. General internists initiated screening for a higher percentage of SPs than family practitioners (67 percent vs. 55 percent of SPs seen), and physicians with the most HIV experience initiated screening more often than physicians with the least experience (69 percent vs. 55 percent of SPs seen).

Primary care physicians are inconsistent and sporadic in screening for and identifying HIV risk factors, concludes Marjorie D. Wenrich, M.P.H., of the University of Washington, Seattle, the lead author of a recent paper describing these findings. She points out that standardized screening techniques are needed to identify persons at risk for HIV. One approach would be the inclusion of specific questions about HIV risk behaviors in standard medical intake forms.

Details are in "HIV risk screening in the primary care setting," by Ms. Wenrich, J. Randall Curtis, M.D., M.P.H., Jan D. Carline, Ph.D., and others, in the February 1997 Journal of General Internal Medicine 12, pp. 107-113.

Many health care providers do not test TB patients for HIV infection as recommended

Nearly half of urban tuberculosis (TB) patients in the United States are also infected with the human immunodeficiency virus, or HIV. But despite recommendations to test all TB patients for HIV, many health care providers choose to test only those they perceive to be at risk of HIV infection, according to a study supported in part by the Agency for Health Care Policy and Research (National Research Service Award training grant T32 HS00046). It shows that one-third of TB patients in AIDS-dense areas of Los Angeles were not tested for HIV, and that provider choice not to test, rather than patient refusal to be tested for HIV, was the critical variable.

Physicians tested mostly persons with HIV risk factors, but at least 2 to 7 percent of low-risk patients were infected with HIV. Thus, the opportunity to diagnose and treat HIV coinfection may be missed in a substantial number of TB patients, conclude the California researchers. They reviewed the medical charts of 500 patients at 20 health centers, who were listed in the Los Angeles County tuberculosis registry in 1993, to document whether or not the patients had been tested for HIV infection.

The researchers found that females were 42 percent less likely to have been tested compared with males, and persons aged 20 to 49 years were more apt to be tested than younger and older persons. Patients with HIV risk factors were nearly nine times more likely to be tested for HIV than patients without risk factors. Overall, physicians tested only 63 percent of TB patients for HIV infection. Patients of private practitioners were 68 percent less likely to have been tested than were patients treated by public health authorities. This may be because public health providers are better educated about the relationship between TB and HIV or that they have a stronger commitment to infectious disease control, suggest the researchers. They conclude that further education of health care providers is essential to achieve universal HIV testing among TB patients.

For more information, see "Testing for human immunodeficiency virus infection among tuberculosis patients in Los Angeles," by Steven M. Asch, M.D., Andrew S. London, Ph.D., Peter F. Barnes, M.D., and Lillian Gelberg, M.D., M.S.P.H., in the American Journal of Respiratory and Critical Care Medicine 155, pp. 378-381, 1997.

Return to Contents

Special Populations

Homeless individuals living in congested urban areas are at high risk for TB infection

Homeless persons living in crowded shelters are vulnerable to infection with tuberculosis (TB). They come into contact with former prison inmates, substance abusers, and others who are at high risk of already being infected, and their often crowded quarters make it easier for the TB bacteria to pass into their lungs via air droplets when infected persons cough or sneeze. As a result, about one-third of the homeless persons in Los Angeles and San Francisco are infected with TB; this translates to a prevalence rate that is three to six times greater than the 5 to 10 percent prevalence of TB skin test positivity in the general adult population.

A recent study, supported in part by the Agency for Health Care Policy and Research (HS06696), examined factors that increase the risk of homeless persons becoming infected with TB. It found that 40 percent of homeless persons in the urban Skid Row area of Los Angeles had positive TB skin tests compared with 14 percent of homeless persons in the suburban Westside area (an overall 32 percent prevalence). This was probably due to the crowded conditions in the urban area, explains lead investigator Lillian Gelberg, M.D., M.S.P.H., of the University of California, Los Angeles, School of Medicine.

The researchers performed TB skin tests on 260 homeless adults in shelters, soup lines, and outdoor locations in two areas of L.A. Compared with persons living in traditional housing (hotel, apartment, or house), persons living in a shelter or institution were four times more likely to have a positive TB test. Individuals living in limited shelters—such as an abandoned building, car, or all-night theater—were 3.5 times more apt to test positive, and those living on the streets were 2.5 times as likely to test positive.

The likelihood of skin test positivity increased dramatically after an individual had been homeless for 10 years (quadrupling the risk) but was relatively constant up to that point. Older homeless persons were more apt to test positive, with persons aged 18 to 30 years showing a prevalence of 19 percent; those 31 to 40, 27 percent; and those over 40, 44 percent. Although alcohol and drug abuse and prison history have been shown to be predictors of TB in the general population, they were not important predictors of TB test positivity in this study.

Details are in "Tuberculosis skin testing among homeless adults," by Dr. Gelberg, Christopher J. Panarites, Ph.D., Hal Morgenstern, Ph.D., and others, in the January 1997 Journal of General Internal Medicine 12, pp. 2-33.

Return to Contents

AHCPR News and Notes

From the Administrator

I would like to take this opportunity to introduce myself, and tell you how pleased I am to be working with you as Administrator of the Agency for Health Care Policy and Research. During my years as a health services researcher, I was impressed with the high caliber of the research supported by AHCPR and conducted by researchers on its staff. This continues to be true as I learn more about the Agency from the inside.

AHCPR is the nation's lead Agency in ensuring that there is a scholarly and scientific foundation for a rapidly changing health care system. However, we cannot succeed without your expertise and dedication to excellence in research. Our shared goal is to enhance the way health care services are organized, financed, and provided. We want to improve the care patients receive and help them take responsibility for their own health and health care.

My agenda is still evolving, but one of my priorities is to ensure that AHCPR continues to enhance the quality of health care and its value by meeting the needs of the health services research community. One of the most critical needs is to ensure that we continue to build capacity for health services research. We already are expanding our efforts to nurture the careers of new investigators beyond our dissertation grants program and our National Research Service Awards.

I also plan to work with you to create centers of excellence where teams of established researchers can do state-of-the-art work, and where young researchers can be trained. We already have excellent examples in our Patient Outcomes Research Teams (PORTs), upcoming Evidence-based Practice Centers, and MEDTEP Minority Research Centers.

We also will continue to play a critical role in fostering innovation among researchers in our field. To that end, AHCPR is carefully weighing the balance between the funding of investigator-initiated research and targeted research. We are reviewing the criteria we use in targeting funds to particular areas and will be seeking funding partners for targeted research projects. In addition, AHCPR will be working with you to translate health services research findings into a language that can be used by decisionmakers at all levels, from patients, clinicians, and health systems to the Federal and State governments. I firmly believe that we must disseminate our findings and do everything we can to ensure that they are applied or at least considered so that the public may benefit from the work we do.

I encourage you to let us know how we can do a better job of communicating and working with you and of providing the basic tools and leadership needed to ensure that health services research can meet the challenges that lie ahead. Please send your comments and suggestions to the Managing Editor, Research Activities. We will let you know in an upcoming issue of Research Activities about the comments and recommendations we receive and how we plan to address them.

AHCPR to collaborate with AMA and AAHP to develop a National Guideline Clearinghouse™

Department of Health and Human Services (HHS) Secretary Donna E. Shalala recently announced that the Agency for Health Care Policy and Research—a HHS agency—will collaborate with two private-sector organizations to develop a comprehensive Internet-based source for clinical practice guidelines. The new National Guideline Clearinghouse™ (NGC) will make available a full range of current guidance on treatments for specific medical conditions.

Under the plan, AHCPR, the American Association of Health Plans, and the American Medical Association will work jointly to develop the new guideline clearinghouse. It is anticipated that AHCPR will award a contract later this year for the technical work to establish the NGC. The target date for launching the new Internet clearinghouse site is fall 1998.

Internet technology makes it possible to provide rapid access to the latest information on medical treatments. According to AHCPR Administrator John M. Eisenberg, M.D., the NGC will provide access to the widest selection of guidelines available from public and private organizations by establishing an independent, interactive Web site, accessible by using any standard Web browser or through the Web sites of the three collaborating organizations.

The development and use of clinical practice guidelines have grown markedly in the past 5 years. However, many existing and potential guideline users have difficulty gaining access to and keeping abreast of the many clinical practice guidelines currently in use. In addition, existing guidelines often differ in their development and content, further complicating their use. To help address these issues, the NGC Web site will:

• Contain standardized information for thousands of guidelines such as title, sponsoring organization, author(s), and methodology used.
• Provide guideline abstracts and where possible the full text of guidelines.
• Compare and contrast the recommendations of guidelines on similar topics, and provide summaries covering major areas of agreement and disagreement.
• Have topic-specific electronic mailing lists to enable registered users to communicate with one another on guideline development, dissemination, implementation, and use.

AHCPR updates CONQUEST database

The Agency for Health Care Policy and Research has released an update to the clinical performance measures database first released last year. CONQUEST 1.1, the Computerized Needs-Oriented Quality Measurement Evaluation System, has been changed to make it more user-friendly and compatible with a variety of operating systems.

CONQUEST is a quality improvement software tool that helps users identify, understand, evaluate and select measures to assess and improve clinical performance. It summarizes information on approximately 1,200 clinical performance measures developed by public- and private-sector organizations to examine the quality of clinical care. CONQUEST describes measures using a common language and provides a research short-cut by identifying specific types of measures that may be modifiable or adaptable for a users own unique needs. If modified, the measures should be retested for validity and reliability.

Organized in an easy-to-use format, CONQUEST consists of two interlocking databases—a Condition Database and a Measure Database—that allow users to target specific conditions or measures. The Condition Database includes information on 52 clinical conditions; the Measure Database contains information on the 1,200 measures contained in 53 measure sets. Users can produce preprogrammed or custom reports and add measures and conditions to the database.

The 1.1 version functions as a run-time program in Windows® 3.1 and higher, including Windows® 95. Microsoft®'s Access® 2.0 is no longer required but still may be used to complete keyword searches or add to the database. The User's Guide has been updated with clearer instructions and an index.

CONQUEST may be downloaded from AHCPR's Web site. The diskettes (AHCPR Publication No. 97-DP01) and the User's Guide (AHCPR Publication No. 97-R001) also are available free from the AHCPR Publications Clearinghouse.

CONQUEST users can obtain free technical assistance through AHCPR's Quality Measurement Network (QMNet) contractor by calling (800) 865-5380.

Return to Contents
Proceed to Next Section