Currently, it is estimated that there are more than 1 million persons in the United States who are living with HIV/AIDS.1 Estimates of HIV incidence in the United States suggest that incidence has been relatively stable since the early 1990s, at approximately 40,000 new HIV infections yearly.2 But increases in HIV/AIDS diagnoses among men who have sex with men (MSM) between 2000 and 20043,4 raise the possibility that HIV incidence may be increasing-at least in some US populations.

Also troublesome is the realization that these epidemiologic data derive from a time in the American HIV epidemic when there exists a fairly broad consensus in the public health community that, with appropriate focus and effort, it should be possible to substantially reduce new HIV infection rates-even in the absence of a vaccine or curative treatment.5

There are those who have interpreted absent declines in US HIV incidence to mean that prevention efforts have failed. Others argue that the effectiveness of US domestic HIV prevention efforts is indeed substantiated, by looking at the difference between observed HIV incidence and estimated HIV incidence in the absence of publicly funded prevention efforts.6,7 Yet, another perspective on the state of the current American HIV epidemic might be found within the ongoing discussion of the organization of America's healthcare system.

The premise of this exposition is that the outcome, good or bad, of efforts to prevent the spread of HIV infection in the United States is inextricably linked to larger, call them systemic, factors that characterize our national healthcare system. In setting out this analysis, salient features of the American healthcare system will be described-for the benefit of readers who may not have a background in public vadjust health and those who are unfamiliar with the context of healthcare in the United States.

In the United States, systems for personal-level healthcare (ie, medical care) and systems for population-level healthcare (ie, public health) do not typically work together to support one another.8 An expert panel of the US Institute of Medicine of the National Academies suggested that the "American fascination with technology, science, and medical interventions and a relatively poor understanding of the determinants of health" have led to a "the lower status, fewer resources and limited influence of public health."9 One substantial manifestation of the dichotomy between medicine and public health is found in the distribution of US spending on healthcare, the highest, per capita, in the world. More than 95 percent of the US health budget is directed toward personal healthcare and biomedical research, with only 1 percent to 2 percent directed toward prevention of disease.10

This "disconnect" between American medicine and public health represents an important impediment to reducing HIV incidence in the United States. Consider the example of early diagnosis of HIV infection. Somewhere between 252,000 and 312,000 persons in the United States are infected with HIV and unaware of it.1 Not only is failure to diagnose HIV infection, early on, suboptimal from a medical perspective, but also from a public health perspective, late diagnoses of HIV infection represent missed opportunities to prevent new infections. Undiagnosed persons may be spreading HIV infection to partners-all the while unaware of their serostatus. A recent meta-analysis demonstrated that high-risk sexual behavior is markedly lower in HIV-positive persons who are aware of their infection status than in HIV-positive persons who are unaware of their serostatus11-strong proof of the significant prevention benefits of early HIV diagnosis.

In an analysis of more than 3,000 persons with HIV/AIDS diagnosed in 16 US health sites between 2000 and 2003, nearly half (45%) of the persons with AIDS had their first positive HIV test 1 year or less before their diagnosis of AIDS-hence, representing "late testers" who were infected for years before they were first diagnosed.12 Parenthetically, the problem of late diagnosis of HIV infection has also been noted in the United Kingdom13-although not the focus of this commentary.

On the basis of data from two, population-based surveys of US adults conducted in 2002, only 10 percent to 12 percent of persons aged 18 to 64 years reported being tested for HIV infection in the preceding 12 months.14 The most frequent source of HIV testing was a medical office setting (ie, private doctor or health maintenance organization), accounting for 43.5 percent of all HIV testing in the preceding 12 months.14 Hospitals, emergency departments, and outpatient clinics were cited as testing locales by 22.4 percent of the respondents and public sources (eg, health departments, sexually transmitted disease clinics, family planning clinics, drug treatment clinics) by 23.6 percent of respondents.14 Thus, in the United States, most voluntary HIV testing takes place in office- and hospital-based physician practices.

We know that when healthcare providers recommend HIV testing to their patients, acceptance and uptake are increased.15,16 As one might surmise, persons who do not have access to healthcare are less likely to be tested for HIV infection. A study of 1,643 heterosexual men and women from high-HIV prevalence neighborhoods in Philadelphia showed that for both men and women, going to the doctor was the strongest correlate of having been tested for HIV infection.17 Yet, evidence suggests that even when patients have good access to medical care, they may experience late diagnosis of HIV infection.18,19

Reasons for failing to diagnose HIV infection in a timely manner among those who have access to medical care are multifaceted. Individuals may be hesitant to reveal highly personal behaviors-or may be unaware of their risks; providers may not employ valid and reliable screening approaches-or shy away from offering HIV testing because of misperceptions that extensive "counseling" is required.20,21 And, in keeping with our emphasis on systems, we must recognize that the current configuration of US healthcare delivery, itself, may embody inadvertent disincentives to the early diagnosis of HIV infection. Of note, this commentary will not explicitly address another important system that influences early diagnosis of HIV infection in the United States-namely, the various mechanisms that provide health insurance coverage and ensure, or hinder, access to care.

Back to systems of healthcare delivery, concerns about widespread disincentives to offering prevention services, not just HIV screening, in medical care settings have been raised by others who note that the current US medical system is geared to respond to symptomatic presentation and acute conditions.22 They observe that many prevention activities like behavioral counseling and linkage to community resources are "outside the scope and culture of clinical medicine" and that providers may have "little or no training" in the requisite skills.22^(p583) Tellingly, many prevention services are not reimbursed in medical settings, and physicians who are reimbursed for their clinical services on a per capita basis may actually be provided with unintended incentives for not performing prevention activities.22^(p585)

Then there is the perennial question of time. Yarnall and her colleagues calculated that in order to fully satisfy all of the recommendations of the US Preventive Services Task Force, physicians, on average, would have to increase their working days by 7.4 hours!!23 Clearly, focusing only on provider behavior as the solution to improving the delivery of preventive services, like screening for HIV infection-without considering fundamental modifications to systems of care-is doomed to failure. Those advocating for improved care for patients with chronic diseases call for a "reengineering" of the healthcare system, including changes to client-level information systems, so that medical records can prompt providers to deliver needed clinical preventive services as well as other fundamental modifications to delivery system design-in particular, the redistribution of health promotion responsibilities across "teams" of healthcare providers, not placing the sole responsibility on busy physicians.22 Similar changes may well be required to improve the early diagnosis of HIV infection.

And, once a diagnosis of HIV infection is established, then what? Certainly, we must ensure that persons receive appropriate medical treatment; but what of their ongoing prevention needs?24 More people living longer with HIV means more opportunities for transmission. Here, too, systemic changes are called for. In clinical settings, we must develop and support brief, easily delivered interventions that stress the importance of preventing the further spread of HIV.25 As noted above, clinical encounters must be reconfigured so that the responsibility for delivering prevention messages does not rest with physicians alone. The smoking cessation literature pointedly demonstrates that successful interventions in medical practice settings involve physicians and nonphysicians alike, delivering complementary messages to patients on multiple occasions.26

Beyond invoking new paradigms within existing medical systems, we must also expand health delivery systems in nonmedical sectors. In the three decades since the onset of the AIDS epidemic, substantial advances have been made in the science of HIV prevention, and many of the most effective interventions are those delivered within community settings.27 Taking a page from the health promotion literature, we recognize that combined clinical and community interventions are more effective than either modality by itself.28 Increasingly, governmental organizations charged with containing the spread of HIV infection have come to realize that they must partner with community in "sharing public health governance."29 Thus, sound approaches to the early diagnosis of HIV infection and sustained efforts to prevent HIV transmission will require substantial community involvement.

Community-based organizations (CBOs) have an undeniable track record of responding to the threat of AIDS in the United States and elsewhere.30,31 Despite this illustrious history, it would be shortsighted to expect CBOs to scale up their efforts, let alone adapt to the changing circumstances of the US HIV/AIDS epidemic, without higher level systems to develop and sustain their capacities.32,33

Although some HIV/AIDS CBOs have banded together into coalitions and loose confederations, CBOs are generally independent, sometimes financially precarious, organizations that often compete with one another for scarce public resources. This, along with the reality that these organizations typically serve underprivileged clients with multiple, complex health and social needs, strongly suggests that CBOs face many inherent barriers to the de novo development of stable systems for the delivery of healthcare and preventive services.

A particular vulnerability lies in the faithful adoption, into practice, of prevention strategies that have been proven effective in rigorous scientific trials or controlled program evaluation settings.34 Successful transfer of proven prevention technologies into community settings requires systems for training, quality assurance, and ongoing evaluation. For the reasons cited above, most CBOs will be unable to develop these systems on their own.

Strategies that expand the capacity of individuals, organizations, and communities to implement, deliver, and sustain effective prevention interventions can pay off with handsome health dividends. Back to our example of early diagnosis of HIV infection, with proper resources and training, CBOs have been shown to be effective at using social network strategies to identify persons with undiagnosed HIV infection.35 After having been trained to provide rapid HIV testing, nine CBOs in seven US cities were able to recruit high-risk clients and diagnose those who were HIV seropositive at a rate that was five times the average prevalence rate reported by traditional, publicly funded HIV testing sites.35

Evidence suggests that failure to reduce HIV incidence in the United States may be less a reflection of our current prevention strategies than it is a manifestation of deficiencies in the systems through which we deliver, or intend to deliver, HIV prevention services. If we are serious about reducing HIV infections in the United States, we have to recognize and address the profound influence of systems of healthcare delivery, for both personal and population health, on HIV prevention goals and strategies. This will require progress in four critical domains: modifying systems of healthcare delivery to promote and incentivize early diagnosis of HIV infection; creating and sustaining opportunities for early diagnosis in nonmedical, community-based settings; improving systems of referral to treatment and prevention following HIV diagnosis; and strengthening and sustaining the capacity of governmental organizations and CBOs to deliver ongoing prevention services to those who are infected or at high risk for infection.

Reconfiguring healthcare systems to improve their ability to deliver effective HIV prevention services, including early diagnosis of HIV infection, will allow us to reach our national HIV prevention goals. And in so doing, these systemic modifications will also advance the delivery of other, vital clinical prevention services, thus avoiding untold economic, personal, and societal costs.

REFERENCES