Is Domestic Violence Screening Helpful?

From 1993 through 1998, a period when professional medical organizations such as the American Medical Association and the American College of Obstetricians and Gynecologists recommended that physicians screen female patients for intimate partner abuse, violence against women by their intimate partners declined by 21%, according to the US Department of Justice.

Experts in the prevention of intimate partner violence would like to know whether medical screening accounted for some of this reported decline. Unfortunately, said Linda Saltzman, PhD, a senior scientist in the Division of Violence Prevention at the National Center for Injury Prevention and Control, Centers for Disease Control and Prevention (CDC), in a recent interview, "we have no data that would help us answer that."

Department of Justice data do not include data on screening, said Saltzman, and it is not possible to make inferences about the effect of screening without data. However, it is unlikely that screening had a major impact on rates of violence against women because few physicians routinely screen their patients for intimate partner abuse unless they have injuries (JAMA. 1999;282:468-474). Practicing physicians say one of the reasons they are reluctant to screen all their patients is the lack of scientific evidence that screening makes a difference for women's health.

Obviously, it makes sense to ask about intimate partner violence if a patient has symptoms or signs of an illness or injury that could have been caused by violence, said Nancy Sugg, MD, associate professor of medicine at the University of Washington School of Medicine. Suppose, she said, that a patient complained of abdominal pain. "I would ask anyone with acute abdominal pain about a history of trauma, and I would also ask if anyone had tried to hit her or hurt her," said Sugg. Like that of most physicians, her diagnostic evaluation would consider trauma as a direct cause of abdominal injury or an indirect cause of abdominal pain resulting from the stress of an abusive relationship, she said.

To address questions about screening and begin to develop a research agenda on the detection and clinical management of intimate partner abuse, the CDC hosted a workshop in April for Saltzman, Sugg, and other experts.

Diagnostic evaluation and screening are two different things, said David Atkins, MD, MPH, coordinator of the US Preventive Services Task Force (USPSTF) Guide to Clinical Preventive Services. According to Atkins, screening for intimate partner abuse implies a standardized assessment of patients, regardless of their reasons for seeking medical attention.

The US Preventive Services Task Force recommendation states "There is insufficient evidence to recommend for or against the use of specific screening instruments to detect family violence, but recommendations to include questions about physical abuse when taking a history from adult patients may be made on other grounds." (US Preventive Services Task Force. Guide to Clinical Preventive Services. 2nd ed. Baltimore, Md: Williams & Wilkins; 1996).

The USPSTF recommends screening and other clinical preventive services on the basis of scientific evidence of their effectiveness. For screening tests, it has two criteria: there must be an accurate test for the condition, and there must be scientific evidence that screening can prevent adverse health outcomes. According to the most recent (1996) guide, there is insufficient evidence to recommend for or against the use of specific screening instruments to detect family violence, but it says that recommendations to include questions about physical abuse when taking a history from adult patients may be made on other grounds.

Atkins said that asking patients about abuse as part of routine history-taking may be indicated on the basis of the substantial prevalence of undetected abuse among women, the potential value of this information in the care of the patient, and the low risk of harm in asking.

For these reasons, Sugg said she asks all her female patients about abuse as part of a complete history-taking and physical examination. "It's part of my health care maintenance screening," she said, adding that an important reason to ask about abuse is that it may be related to a variety of common, stress-related medical conditions, such as asthma and hypertension. Unfortunately, said Sugg, there is no proof that she is making a difference in her patient's lives by asking about abuse. Her recommendation: "We need long-term outcomes. We need controlled trials."

In addition to her clinical practice, Sugg trains physicians to assess their patients for intimate partner abuse. "They are always asking me, ‘What proof do you have that asking about abuse will make a difference for this woman?' I have to say to them, "Wait a minute—I have no data.'"

Herb Garrison, MD, MPH, professor of emergency medicine at East Carolina University School of Medicine in Greenville, NC, said that the lack of knowledge about interventions and outcomes of intimate partner violence screening compares unfavorably with the impact of other screening tests. "If your patient has a positive Pap smear," said Garrison, "you know exactly what to do. You send her to a gynecologist" for further evaluation and treatment. In contrast, the lack of evidence for effective interventions is a disincentive for physicians to ask all their patients about intimate partner abuse, he said.

This is why physicians find the USPSTF guidelines so useful, said Robert F. Thompson, MD, director of the Department of Preventive Care at Group Health Cooperative of Puget Sound. "We care what the guidelines say because they help us prioritize what we do," he said. Thompson, who has undertaken one study of intimate partner violence screening and is planning another, said that screening is "the right thing to do. But what is the best way to do it?"

Another participant in the CDC workshop was Heidi Bauer, MD, MPH, a clinical fellow in sexually transmitted diseases prevention in the California Department of Health Services. Recalling that the effectiveness of mammography in early detection of breast cancer was demonstrated in a randomized controlled trial, Bauer advocated similar research on screening for intimate partner abuse. "But if I were to design it," she said, "I would give myself the best chance of success, with multiple-level interventions and lots of intermediate outcomes."

Breast cancer death rates were not expected to fall immediately after instituting mammography, said Bauer, nor should unreasonable outcomes be expected right away with screening for intimate partner abuse. "We should have a gold standard for abuse, an accurate screening test, a measure of patient acceptability, follow-up on referrals, and then down the road be looking at health benefits and cost savings," she advised.

Laura Sadowski, MD, MPH, codirector of the Collaborative Research Unit at Rush Medical College of Rush University in Chicago, another workshop participant, agreed that screening for intimate partner abuse, like all screening tests, needs formal scientific evaluation. "How do you define a false-negative test? Some women aren't ready to disclose that they have been abused. Do you ask all your patients at every visit? How many times do you ask?" Sadowski pointed out that biopsy-confirmed studies of colorectal cancer screening, for example, have shown that patients can be considered cancer-free after six consecutive negative fecal occult blood tests, but similar evidence for a "true-negative" test for intimate partner abuse is lacking.

Equally important is evidence that early detection will improve health outcomes, said Thompson. By analogy, he said, the USPSTF considered the published evidence from randomized controlled trials and well-done nonexperimental studies and concluded that prostate cancer can be detected early, "but do the patients do any better? Nobody knows." Moreover, he said, there may be potential for harm when screening leads to a cascade of interventions. "We figured out that if we tested all 35,000 men over age 50 in the Group Health Cooperative population, we would have over 550 serious adverse events, including 13 deaths as a result of surgery. We still offer the test to patients who ask for it, but now we tell them the pros and cons."

Because of their evidence-based approach to the evaluation of screening tests, the USPSTF guidelines carry great weight with institutional purchasers of managed care services, said Thompson. Moreover, the guidelines help define research agendas by identifying gaps in knowledge about screening tests. Participants in the CDC-sponsored workshop agreed that screening for intimate partner abuse should be evaluated scientifically. However, there was also consensus that physicians should not wait for the results of randomized clinical trials to begin screening in their own practices.

Sugg noted that surprisingly few clinical interventions are supported by evidence from clinical trials. "I feel that people who want evidence for everything they do just can't practice primary care medicine," she said, adding that her concern is not whether she can prove to her colleagues that it is important to screen for intimate partner abuse. "In my mind, that's a no-brainer," she said. "But I do want to make sure that what we are doing about intimate partner abuse is the right thing. Are we really doing what we should do? That's my question, not should we do it at all."

The test must be able to detect the target condition earlier than without screening and with sufficient accuracy to avoid producing large numbers of false-positive and false-negative results (accuracy of screening test). Screening for and treating persons with early disease should improve the likelihood of favorable health outcomes (eg, disease-specific morbidity or mortality) compared with treating patients when they present with signs or symptoms of the disease.

Source: US Preventive Services Task Force. Guide to Clinical Preventive Services. 2nd ed. Baltimore, Md: Williams & Wilkins; 1996.

Chicago—Every year in the United States, at least one physician is killed by a patient.

Psychiatrists and other mental health professionals rank fourth—just below taxicab drivers, convenience store clerks, and police—in likelihood of being killed in the workplace.

Eighty percent of nurses report being assaulted on the job at least once in their careers, the highest rate for any occupational group.

Residents of the United States own more than 200 million guns. Firearm injuries caused more than 32,000 deaths in this country in 1997 (the most recent figures available). An additional 64,000 persons were injured by guns that year and survived (Morb Mortal Wkly Rep. 1999;48:1029-1033).

These sobering statistics helped fuel a forum on strategies to ensure practitioner safety and other psychiatric aspects of violence at the annual meeting here of the American Psychiatric Association (APA).

The American public increasingly views violence as a serious public health problem, noted Carl Bell, MD, who chaired the forum and is vice chair of the APA's Task Force on Psychiatric Aspects of Violence. Yet many physicians deny that they are potential targets of violence, said Bell, who directs the Community Mental Health Council in Chicago. Moreover, he asserted, "physicians' commitment to the principle ‘do no harm' often gets in the way when they are threatened."

Some murders of physicians probably could have been prevented by appropriate interventions by the physicians or the institutions in which they worked, said Arthur Berg, MD, of Harvard Medical School. Learning to assess and manage violent patients, he said, should be part of all medical school, residency, and allied health professional training programs.

Physicians should be alert to warning signs of imminent violence. In an escalating situation, he said, patients often pace or show other agitated behavior, and open and close their fists. Their eyes dart, and they intrude into the physician's personal space, making him or her feel uncomfortable or fearful.

These signals, Berg said, call for action. "Administrators should tell practitioners that they have a right to defend themselves when threatened with bodily harm," he said. This declaration, he maintained, will remove some potentially immobilizing ambivalence. Clinicians who ignore their discomfort because of uncertainty, inexperience, or a false sense of bravado, he asserted, are more likely to suffer harm.

Drug or alcohol intoxication and loud outbursts often indicate impending loss of control. The clinician needs to get away quickly from patients in such states, said Berg, who is a member of the American Society of Law Enforcement Trainers. Nonthreatening postures by the clinician may help defuse the situation. One is the "thinker" pose: one hand on the cheek, with the elbow of that arm resting in the hand of the opposite arm, which is crossed in front of the body. Another involves keeping both arms at the sides or slightly forward, palms up. "Visualize the situation," Berg suggests. "Imagine someone grabbing you and plan what you might do."

Violent situations can be categorized as emergent, urgent, and potential, said Joe Tupin, MD, of the University of California, Davis, Medical Center. An emergent situation involves an assaultive patient in the same space: an inpatient unit, emergency department, or private office. The practice environment influences safety planning, he said. In emergency departments and inpatient units, for example, furniture should be heavy or soft, so that it cannot be used as a weapon. Interview rooms ideally will have two doors to offer a potential escape route.

Those who work in a busy emergency room need to know how to contact other professionals and nearby security personnel— for example, by using a buzzer system—and to mount a coordinated response. A single practitioner in an isolated office should avoid seeing potentially dangerous patients there. Clinicians should not keep potential lethal weapons such as letter openers (sharp objects) or sculptures (blunt ones) on their desks.

Clinicians who are threatened should seek any way to escape or to restrain the patient, he said, "not definitive, not elegant, just a way to keep themselves and their staff out of harm's way for minutes to hours, long enough to regroup and move forward." In emergent situations, he said, physical control tactics, such as take-down procedures, become important. Several people may be needed to subdue a violent person. The staff needs to practice this procedure.

Sedating medications, such as benzodiazepines or high-potency neuroleptics, Tupin said, typically are needed in an emergent situation and usually are given intramuscularly or intravenously. Haloperidol is an example of a commonly used and effective medication, he said, although it has not been approved by the Food and Drug Administration for intravenous use.

An urgent or imminent situation, with a patient on the verge of losing control, Tupin said, presents an opportunity for verbal and interpersonal interventions, and for oral medications. Judicious use of seclusion and removal from the environment often prove helpful. Safety still is the main concern, he said, but treatment may be conducted with a more therapeutic eye.

A potentially violent situation, Tupin said, is one in which the clinician learns that someone who comes for consultation or who already is in treatment has a history of violence or also abuses alcohol or drugs, all risk factors for violent behavior. Here, he said, there are opportunities for a thoughtful diagnostic assessment that provides a better handle on long-term intervention.

Advance preparation is the best preventive strategy, these clinicians agreed. "I don't think doctors should carry guns," said Bell, "but self-defense is important." He cited the adage, "It is better to be judged by 12 than carried by 6." Though trained in martial arts himself, he aims to prevent violence by building rapport and being careful in his choice of words. "I advise clinicians not to put hands on patients," Bell said, "unless they are being assaulted."

A new book, Psychiatric Aspects of Violence: Issues in Prevention and Treatment, edited by Carl Bell (San Francisco, Calif: Jossey-Bass; 2000; $25), includes a chapter by Berg, Bell, and Tupin with detailed guidelines for clinician safety and management of violent patients. Other chapters review the biology of violence, assessment of the risk for violence, prevention of violence, family violence, victims and perpetrators of sexual violence, and treatment of traumatized patients and victims of violence.

The APA's Task Force on Psychiatric Aspects of Violence, chaired by Paul Fink, MD, of Temple University School of Medicine, Philadelphia, expects to issue a comprehensive report later this year on these and additional topics, including working with the media to ensure accurate reporting in case of a violent incident. The report will aim to make information readily accessible, with fact sheets, summaries of best practices, resources, references, and other guidelines. The goal, Fink said, is to produce a practical manual for medical students, residents, and organizations.

Cool Horror, a life-sized white marble figure on a hospital gurney, with chart and murder guns, was one of 75 works in a show entitled "Guns in the Hands of Artists," held in New Orleans in 1996.

The chart lists the medical record numbers, nature of injuries, and fate of more than 3000 gunshot victims treated by the sculptor, Steven Lesser, MD, a specialist in emergency medicine, in the Accident Room of Charity Hospital in New Orleans in the 18 months preceding the show. "On a typical 12-hour shift, we admit five men, women, or children with gunshot wounds," Lesser said in a recent interview. "We usually can save four out of five."

The figure is faceless, Lesser said, to show that "all of us are victims of the shootings." It is armless to suggest everyone's vulnerability. Shackled at neck and feet with chains made from handguns, it rests on a bed of rifles.

Some of the guns in this work likely were used to shoot Lesser's patients; all were used in crimes committed in New Orleans in the same time period. While rendered inoperative, they retain their police identification tags, a grim reminder that each gun wounded or killed a specific person. The figure, according to Lesser, "reposes on a troubled sea representing the epidemic of senseless violence that grips our country."

Other sculpture by Lesser, who is an associate professor of medicine at Louisiana State University School of Medicine, will be on display at the Contemporary Art Center in Virginia Beach, Va, from April 28 to September 16, 2001.—L. L.

Chicago—Family violence has a ripple effect. Those who experience it and those who witness it suffer physical and emotional injuries. Those who perpetrate it also wound themselves. Violence damages family and other relationships, often from childhood onward.

Clinicians often struggle when trying to address domestic violence. They may not know what to say or do. They may find it traumatic to listen to a patient's report. Some have trouble empathizing with the victim's helplessness. Some are involved in abusive relationships themselves, according to speakers at a symposium on domestic violence at the annual meeting of the American Psychiatric Association here.

In emergency departments, physicians may miss or dismiss abuse, according to Carole Warshaw, MD, who directs the domestic violence and mental health policy initiative at Cook County Hospital in Chicago. She cited a surgeon who noted blunt trauma to a woman's face in the medical record without saying how that occurred.

People in abusive situations may visit physicians for isolated injuries, multiple somatic complaints, chemical dependency, depression, and other problems. Physicians may give them prescriptions for psychoactive and analgesic agents without exploring the reasons for their complaints. As a result, Warshaw said, patients' feelings of despair and isolation increase.

About one in four women will be subjected to domestic abuse during her lifetime, the US Department of Justice estimates. Women are up to eight times more likely to be victimized by an intimate partner then men are (Violence by Intimates: Analysis of Data on Crimes by Current or Former Spouses, Boyfriends, and Girlfriends. Washington, DC: US Department of Justice; 1998). When women do assault male partners, it is more likely to be in self-defense, Warshaw noted, and rarely as part of an ongoing pattern of coercion and control. Domestic violence also is a significant problem, she said, in same-sex relationships.

Clinicians need to recognize, she said, that abusers often use psychological as well as physical coercion. Abusers may withhold medications, change a prescription, cancel appointments, and say their spouse is mentally ill and hallucinates the abuse. They may keep partners awake, or threaten to carry out attacks while the partner sleeps. One man, Warshaw said, forced his diabetic wife to eat the wrong foods. The husband or partner often controls family health insurance and income, restricting the partner's independence.

Contrary to widespread belief, Warshaw said, there is no personality profile that causes women to stay in abusive relationships. Abused women generally make many attempts to leave their abusers. They may appear passive and compliant, she said, because they have learned that such behavior blunts the pain of living with abuse. This is an appropriate response to ongoing danger. Their symptoms often lessen once they feel safe. For gay men and lesbians, she said, there may be the added trauma of having to explain the relationship to the physician, as well as acknowledge the abuse.

Physicians should never ask about abuse in the presence of the abuser, she said. When alone with the patient, it's appropriate to say, "I'm concerned that some of your medical problems may be the result of someone hurting you," and to ask direct questions: Do you feel safe at home? Does your partner humiliate you? Has your partner ever made you engage in sex when you didn't want to? When you are with your partner, do you feel like you are walking on eggshells?

Developing a safety plan is critical, she said. That might include tactics such as locating a safe place to go in an emergency and teaching children to dial 911. The physician needs to treat patients with compassion, she said, and to facilitate interactions within the family, with social service resources, and with the local legal system. While all states mandate reporting child abuse, few require it for domestic violence.

Early exposure to abuse vastly increases children's risk of violent behavior toward others or themselves later on, said Bessel van der Kolk, MD, of Boston University School of Medicine. Boys who see their mothers beaten are far more likely to be abusive adults, he said (Can J Psychiatry. 1990;35:466-470), and girls who witness beatings are 300 times more likely to be in abusive relationships (J Marriage and Family. 1984;46:11-19). People abused as children are 18 times more likely to commit suicide than those who were not abused (Am J Prev Med. 1998;14:245-258).

Well-publicized campaigns tell children not to take candy or accept rides from strangers, van der Kolk said, yet 80% of domestic trauma in children is inflicted by their own caretakers. Other relatives account for an additional 10% of instances of trauma in children. More than 1 million confirmed cases of child abuse and neglect occur in the United States annually.

Mothers who were abused when young have trouble relating to their own children, van der Kolk said. Serial videotapes show their babies starting to turn away from them after the early weeks of life. The mother then displays frustration and neglects the child. After repeated abandonment, he said, children experience an emotional shutdown. They often don't recognize their own feelings or those of others.

Abuse also puts children at higher risk of later medical problems, including HIV infection, heart disease, and diabetes. Abused children may try to reregulate their emotional state with self-mutilation, anorexia or binge eating, or abuse of drugs (Am J Prev Med. 1998;14:245-258).

In treating abused adults, an injury model works better than a sickness model, asserted Sandra Bloom, MD, who directs The Sanctuary, a short-term therapeutic milieu program located at two sites, the Horsham Clinic in Ambler, Pa, and the Hampton Behavioral Health Center in Rancocas, NJ. The sickness model implies the etiology is within the sufferer, she said. By contrast, an injury model connects suffering with the environment in which it occurred and with the person who caused it. It implies recovery. It also empowers sufferers to participate actively in their own recovery.

"We look at our patients as normal people reacting to abnormal stress," Bloom said. "We try to stabilize their environment and not to retraumatize them by placing them in situations that reinforce helplessness, scapegoating, and isolation." She and her colleagues seek to teach people to put their feelings into words, to stay focused on the present, and to grieve for their losses. The aim, she said, "is to free them from being stuck in the past."

The myth that children are not violent was shattered by recent school shootings, noted Elissa Benedek, MD, of the University of Michigan. Such events, while rare, represent the culmination of many slights and episodes of victimization, she said. Clinicians must be more aware of these widespread forms of violence. Children encounter violence on television, via the Internet, in their music, and in the computer games they play. They may participate in vandalism or witness it and steal or have things stolen. Some are assaulted at school. Sexual harassment and bullying are common. One of 7 juveniles arrested is charged with a violent crime.

Perpetrators are getting younger, she noted. In Michigan, a 5-year-old boy brought a gun to school, and shot and killed a classmate. Violence by girls, she said, also is increasing, although males still account for 94% of homicides.

Youngsters who harm others often lack empathy and have explosive anger and chronic hatred, Benedek said, because they were treated poorly themselves. Some have rigid parents who treat them harshly when they cannot achieve the parents' unrealistic goals. Some overreact to failure.

Most clinicians are not trained to assess violence, Benedek asserted, particularly in children. Perpetrators of recent school shootings provided multiple hints, she said, but nobody took them seriously. One boy drew numerous pictures of knives and guns, yet no one asked him what was going on in his life.

A clinician evaluating a child for violent behavior, she said, should query the child's family and friends about abusive acts toward other children or pets. It's useful to review poems, diaries, and e-mail, if available. One should ask children to describe the most violent act they have ever committed and how they felt about it. Also ask about what makes them angry and how they deal with it. It's important to learn if they have access to guns. One also must consider biological factors, such as brain injury.

Clinicians often are too sympathetic, Benedek said, excusing violence in children whom they know were abused. Appropriate treatment may involve psychotherapy to help these children verbalize their thoughts and feelings, cognitive-behavioral therapy to help them learn better ways to deal with aggression, group therapy to improve social skills, and possibly medication. Foster care also may need consideration.

Victims of violence, she stressed, also need treatment. Some students who see classmates shot and killed by another student, for example, will be able to talk about their feelings at the time, but others may not be able to do so for years. Critical incident debriefing, she said, can help keep victims of violence from feeling alone and promote their healing.

A Web site, Domestic Violence Education, at http://www.dvcme.org/, offers physicians two free continuing medical education (CME) credits for taking a comprehensive course online. Developed for the American Medical Women's Association by Marjorie Braude, MD, of Los Angeles, the course is based on manuals of the Family Violence Prevention Fund and is designed for physicians and medical students, as well as other health care professionals.

Course topics include the nature and dynamics of domestic violence, screening, assessment and documentation, presentations and assessment in the clinic, safety planning, referrals, legal aspects of domestic violence, children of domestic violence, and the perpetrator. It also puts a human face on these issues via video and audio first-person accounts by male and female survivors of domestic violence.

Useful phone numbers include the National Domestic Violence Hotline, (800) 799-7233; and the National Resource Center on Domestic Violence, (800) 537-2238.—L.L.

Durban, South Africa—Not since the Black Death devastated medieval Europe has humankind observed infectious disease deaths on such a massive scale that a country's population has shrunk rather than grown. But that scenario is playing out again in the 21st century, with HIV/AIDS replacing bubonic plague as the killer, according to new data presented here at the XIII International AIDS Conference.

For the first time, this conference is taking place in Africa, the epicenter of the epidemic. And while it has been clear for some time that HIV/AIDS has stricken sub-Saharan Africa with unparalleled savagery, experts who thought themselves incapable of being shocked by high HIV/AIDS prevalence rates and mortality are stunned by the latest figures documenting the impact the infection is having on this continent.

According to projections from a new study commissioned by the US Agency for International Development (USAID), by 2003, Botswana, South Africa, and Zimbabwe will be experiencing negative population growth, and five other countries will be experiencing a growth rate of nearly zero, said Karen A. Stanecki, MPH, chief of the health studies branch of the US Census Bureau's population division.

"This is the first time the Census Bureau is estimating negative population growth due to AIDS for any country," said Stanecki, who presented the new findings at a press briefing. As a result of a combination of the HIV prevalence rates and the relatively low fertility rates—also an effect of the HIV/AIDS epidemic—population growth in Botswana, South Africa, and Zimbabwe will range from −0.1% to 0.3%. Growth rates for these countries would have been an estimated 1.1% to 2.3% in the absence of the epidemic.

Population growth of several other African countries, including Malawi, Namibia, Swaziland, and Zambia, will be near zero as a result of AIDS.

"The take-home message is how we continue to underestimate this epidemic—the scope of it, how rapidly it's moved from the urban to the rural areas, the rapidity of the rise of the epidemic," said Paul DeLay, MD, chief of USAID's HIV/AIDS division. "And most significantly, we've underestimated the severity and how high prevalence can get in the general adult population. Five years ago, no one here could have estimated that we would see countries with a national prevalence of over 35%."

The study comes on the heels of the latest report from the Joint United Nations Programme on HIV/AIDS (UNAIDS), released a week before the conference commenced. The UNAIDS report notes that nearly 36% of the adult population of Botswana is living with HIV/AIDS. AIDS is now the number one cause of death in Africa and the fourth highest globally. Seven countries—Botswana, Lesotho, Namibia, South Africa, Swaziland, Zambia, and Zimbabwe—now have estimated HIV prevalence rates of 20% or greater.

"It is really a sobering report—shocking even for me, working in this area for such a long time," said Peter Piot, MD, executive director of UNAIDS, referring to his organization's report.

"The problem is going to get much worse before it gets better," said epidemiologist Roy M. Anderson, PhD, of the University of Oxford, England, who presented an analysis of the successes and failures of various interventions to limit the spread of HIV. "We don't know how much worse it will get—that depends on what happens now."

Hard-won gains in life expectancies in many countries in sub-Saharan Africa are withering as a result of the AIDS epidemic. "In Botswana, the life expectancy is now 39 instead of 71," said Stanecki. In Zimbabwe, life expectancy has dropped from 70 to 38. Four other countries in sub-Saharan Africa—Malawi, Mozambique, Rwanda, and Zambia—have had life expectancies slashed a decade or more by HIV/AIDS to less than 40 years of age.

Even more shocking are the projected drops in life expectancies in the next decade. By 2010, at a time when life expectancies in the absence of AIDS would have been expected to reach about 70 years, many countries in southern Africa will have expected life spans in their populations plummet to about 30 years of age. "These are levels probably not seen since the beginning of the 20th century," noted Stanecki.

For a variety of reasons, the HIV/AIDS epidemic in sub-Saharan Africa is taking a heavier toll on women than on men. Girls and women are infected at higher rates and at younger ages—resulting in even lower average life expectancies for women, whose mortality rates will peak during the approximate ages of 30 to 34 compared with 40 to 44 for men.

This discrepancy could contribute to a vicious cycle, Stanecki said. If current trends continue, by 2020 there will be more men than women between the ages of 18 and 44, which in turn may cause men to seek even younger partners. Studies indicate that older men are infecting younger women and as those women go on to infect other partners, even higher HIV infection rates may result.

Although antiretroviral medications have dramatically improved survival in the United States and Europe, the costly drugs are not an option for most Africans. Although five pharmaceutical companies recently promised to lower the cost of the drugs for developing countries, the cost may still be too high for many. Even if cost was not an issue, poor countries lack the health system infrastructure needed to adequately administer complex drug regimens.

The results of a new project announced at the conference may provide some insight into the degree of difficulty involved in overcoming such obstacles. The Bill & Melinda Gates Foundation, Merck and Co, and the Republic of Botswana have established the Botswana Comprehensive HIV/AIDS Partnership, an initiative to improve the overall state of HIV/AIDS care in the country. The Gates Foundation will spend $50 million to help Botswana build the kind of infrastructure needed to offer HIV prevention and treatment services, and Merck has said it will match the Gates funding, helping with the development and management of the program and, in large part, by providing free medications for people with HIV and AIDS.

But with no preventive vaccine on the horizon and with antiretroviral therapy and other medications currently beyond the reach of the vast majority of Africans living with HIV and AIDS, the best tool at hand with which to intervene in the pandemic is prevention efforts targeting behaviors that put people at risk for the infection.

Some sub-Saharan countries have maintained relatively low rates of infection because of early preventive efforts to educate people about the disease and about reducing the risk of infection via condom use and other means. "Senegal started out in the '80s with a strong STD control program as well as a strong political commitment to [containing] the epidemic, and HIV prevalence has remained low in that country," said Stanecki.

Similar commitment to prevention efforts has helped Uganda decrease rates of new HIV infections in urban settings by half, said DeLay. But while such steps are urgently needed and likely will ultimately make a difference, for countries that already have high prevalence rates, the short-term impact of the epidemic will be devastating.

"Even though we assume that by 2010 there will be intervention programs in place and behavioral change in place that will be resulting in lower AIDS mortality, it will take these countries a long time to recover from the current levels of HIV prevalence," Stanecki told her somber listeners.