Testing and Screening Controversies

The question of whether to screen inmates for HIV antibodies has arisen with particular urgency in the prison setting, and considerable resources have been expended in legislative debate and in court challenges of testing and screening practices. Prison administrators have been under pressure from legislators, city and county officials, correctional officers, and inmates themselves to conduct mandatory screening of all inmates and to identify seropositive inmates. Many actors in the prison drama have asserted a ''need to know" information about HIV serostatus.

As in the outside world, proposals for mandatory, identity-linked screening of prisoners have raised questions related to the accuracy of test results and the appropriate balance between resources expended for screening and other educational activities. Critics of mandatory HIV antibody screening of prisoners have been skeptical of the public health, clinical, or behavioral justification for such programs, although the advent of efficacious early therapeutic intervention is muting some of the criticisms. Critics of mandatory screening have also pointed to the special difficulties in keeping health information confidential in a prison setting. In some instances, HIV-positive inmates have been forced to disclose their status to parole boards, family members, sexual partners, and in at least one state (Alabama), potential employers (Freeman, 1991).

A number of prison systems, as noted, have conducted blinded, anonymous seroprevalence studies from which data are available only in the aggregate and individual inmates are not told the results (Hammett and Dubler, 1990). Informed consent is generally not sought, and inmates are often not told a study is being conducted. Most HIV seroprevalence and transmission studies in prisons have used portions of blood drawn from all inmates for independent purposes, such as entry or annual physicals. Some commentators have raised ethical concerns about blinded seroprevalence studies (Bayer, Lumey, and Wan, 1990). Under the design of blinded studies, seropositive individuals cannot be apprised of their status, a practical concern now that there is evidence of efficacious treatments that can be administered before symptoms develop. Others have suggested that blinded epidemiologic studies "risk stigmatizing the entire [prison] population" (Hammett and Dubler, 1990:496).

In some states, prison officials have had to rethink their approach to identity-linked HIV antibody screening when the results began to come in. A few states (e.g., New Mexico and South Dakota) abandoned their programs when they found low seropositivity rates. Other states (e.g., Arkansas and Texas) screen on the basis of apparent risk factors, testing intravenous drug users, prostitutes, self-identified gay men, and those with clinical symptoms of sexually transmitted diseases, tuberculosis, or HIV infection. According to Jan Diamond, a physician at the California Medical Facility at Vacaville State Prison (quoted in Smith, 1991:29):

The prison administration knows they do not have the capacity to handle everyone who is infected, so they are not eager to find out. For a long time the California legislature wanted to institute mandatory testing of prisoners, followed by quarantine of the infected, ostensibly to stop transmission. But when AZT became accepted therapy for treating asymptomatics, they quickly figured out how much it would cost to really know who had the virus and they dropped their push.

Most prison systems offer HIV antibody testing on a voluntary basis. In Oregon's program, which is characterized by aggressive educational efforts and individualized risk assessments, two-thirds of the inmates volunteered to be tested in a study conducted from September 1987 to January 1988 (Andrus et al., 1989). In a sample of 977 newly incarcerated inmates, only 1.2 percent (12) were HIV positive, despite the fact that 63 percent (611) of the sample reported having engaged in risk behaviors. For each inmate who had engaged in a risk behavior and seroconverted, 53 had not, which underscores the potential of education and counseling in preventing HIV transmission.

Transmission Within Prisons

Very little data are available on how many prisoners become infected while in prison. Even when retesting is conducted among prisoners who were seronegative at intake, the "window" period of seroconversion (during which HIV-infected individuals test negative with available screening methods) makes it impossible to tell whether those who later test positive actually became infected while in prison. Impressions from fragmentary data gathered in Maryland and Nevada suggest that transmission rates among inmates while in prison may be quite low (Horsburgh et al., 1990). In an analysis of AIDS cases in New York State and Florida, only a small percentage of the inmates had been incarcerated for more than 2 years prior to their diagnosis in prison, which makes it unlikely that the majority became infected while in prison. Another transmission study is being conducted among male Illinois prison inmates, using blinded blood samples collected at two sequential annual physicals (Hammett and Dubler, 1990), but no data are yet available.

Data from the Federal Bureau of Prisons based on time-interval tests indicate extremely small rates of seroconversion while in prison. Of approximately 98,000 HIV antibody tests, 14 previously seronegative inmates were positive on retesting. All of the 14 inmates seroconverted within the first 6 months of incarceration, which suggests they had become infected prior to imprisonment. Moreover, seroprevalence rates for releasees from federal prisons continue to be lower than for incoming inmates, which also suggests little, if any, seroconversion within federal prisons. This is not to say that transmission of the virus never occurs in correctional settings. The fact that inmates acquire sexually transmitted diseases with incubation periods of days or weeks is evidence that sex occurs in jails, and the sexual transmission of HIV is possible.⁶

Prison sex is a particularly sensitive issue, and it has received more attention since the AIDS epidemic began (A Federal Prisoner, 1991). Both in the popular imagination and in actuality, sex in prisons includes violent, forcible rape.⁷ Prison sex may also be "consensual," although consent is always suspect in the prison context (Lockwood, 1980; Propper, 1981). Submission to sexual overtures may involve outright intimidation or bargaining in exchange for protection, commissary items, or other favors. New inmates and openly gay or effeminate inmates may be particularly vulnerable to forced sex. The prevalence of all types of sexual activity most likely varies from institution to institution. There have been few studies of sexual activity in prisons. The most frequently cited involve anecdotal evidence collected prior to the AIDS epidemic (see studies reviewed in Hammett and Moini, 1988:55-56).

Perhaps surprisingly, however, it is concern about the spread of HIV by seropositive prisoners through other than sexual means that has preoccupied judges and correctional staff. Much of the focus has been on modes of transmission highly unlikely to transmit the virus, such as by casual contact or through food. There have been 40 or more criminal cases involving assaultive behavior by HIV-positive inmates, (e.g., biting, spitting, or throwing the contents of slop buckets at prison officials), but there has not yet been a case documented in the medical literature of HIV transmission to a correctional worker or inmate in such a manner. In one case a prisoner sued correctional officials for failure to isolate an HIV-positive inmate who bit him, claiming that failure to do so violated his Eighth Amendment rights to be free from cruel and unusual punishment (Cameron v. Metcuz, 705 F. Supp. 454 (N.D. Ind. 1989)). In another instance, an HIV-positive inmate saved up his saliva in a bucket to pour over a particularly despised guard. Although relatively few in number, the cases involving assaultive behavior by an HIV-positive inmate have tested the ability of the courts to sort out scientific data about transmission risks from fears and prejudices (Burris, 1990).

Other potential mechanisms of disease spread are unique to prison culture. Because razor blades are difficult to acquire in prisons, inmates commonly share them and thus risk exposure to contaminated blood. Tattooing and ear piercing are also common among inmates, who are resourceful in finding sources of pigment and instruments for such purposes. Instruments used in tattooing can be a mode of disease transmission when not sterilized between use. There is also "spitback methadone," whereby a prisoner swallows and regurgitates methadone for later use.

Some commentators have urged that greater attention be given to the prison as a setting for counseling for intravenous drug users, especially those who are young and might otherwise be unlikely to seek drug abuse treatment for several years (Lampinen, Brewer, and Raba, 1991). Drug use continues within prisons, and ingenious methods and the cooperation of guards and visitors are used to smuggle drugs into the prison (Dash, 1990).

HIV Education and Prevention

Most prison and jail systems have instituted some kind of educational program for inmates and staff concerning HIV prevention. The nature and quality of these efforts vary considerably from jurisdiction to jurisdiction. Many local jails rely almost exclusively on written or audiovisual materials. Most state and federal prisons, on the other hand, conduct AIDS education sessions led by trained staff members.

Prisoners have special needs with regard to AIDS education. Materials must be geared toward the sensitivities of racial and ethnic minorities. Moreover, many prisoners are illiterate or do not speak English, and less than one-fourth of correctional systems have made special provisions for AIDS education in this regard. As one solution, the National Commission on AIDS (1991) recommended strengthening the role of community-based organizations in providing educational and support services for inmates and their families. One of the most promising solutions is the use of prisoners as peer educators, although only a few systems report having adopted such programs. One of the earliest and most successful such programs, initiated by the prisoners themselves, is at Bedford Hills prison for women in New York State. Women in prison have special needs for targeted AIDS education that have often been overlooked (Viadro and Earp, 1991), and peer educators can be expected to be sensitive to those needs and know how to communicate with women about them.

With regard to prevention, one specific aspect that continues to vex prison administrators is the propriety of distributing condoms to inmates. As noted, an undeniable, if difficult to quantify, amount of sexual activity takes place within prisons (van Hoeven, Rooney, and Joseph, 1990). Many inmates are aware of the risk of HIV and other sexually transmitted diseases, as evidenced by an illicit market for plastic wrap and similar items that can provide barrier protection when used in sexual activity. Nevertheless, the distribution of condoms continues to be resisted by prison officials, who are unwilling to admit the extent of same-gender sexual contact among prisoners or to appear to condone such behavior among prisoners. Fears have also been voiced that condoms might be used as weapons or as containers to smuggle drugs. According to the National Commission on AIDS (1991), however, condom distribution has not disrupted prison operations in the few systems that distribute them.

Vermont was the first state in which the prison system distributed condoms. Mississippi, Philadelphia, San Francisco County, and New York City are among the few other jurisdictions that allow condom distribution (Hammett et al., 1989). As one proponent graphically explained (Rooney, 1990:63-64):

What we did [to convince officials to allow the distribution of condoms] was to prove to Corrections that sexual intercourse did take place in a correctional setting by proving to them a substantial number of cases of acquired gonorrhea, oral, urethral, and rectal.

When condoms were first introduced in New York City jails, uniquely colored condoms were used so they could be distinguished from commercially available condoms in the event they were used to smuggle drugs from visitors. Such fears about condom distribution have not been realized, however. The way in which condoms are made available has much to do with whether they are actually used. Distributing condoms under the "medical model" has proven palatable to some correctional officials who otherwise resisted allowing their use. Making them available only by prescription or only through the prison pharmacy, however, may significantly limit access and use. In Philadelphia, condoms are available to inmates on intake into the facility, at AIDS education sessions, and at "medication call." A report by the U.S. Conference of Mayors (1989:5, 8) described how the process works:

Depending on the jail, condoms are placed in shoe boxes or buckets and left for residents to voluntarily pick up at the sick call dispensary, thus allowing for non-personal disbursal to avoid embarrassment or identification with unallowed sexual activity. … During education sessions, condoms are enclosed in a packet of information so as to reduce residents' anxiety about being seen taking condoms.

Housing and Segregation

Prior to the AIDS epidemic, prison officials often faced questions about where to house specific types of prisoners. Many prisons, for example, have special units for sex offenders or inmates with a propensity for violence; those convicted of capital offenses are commonly housed together on a "death row." Some prisons designate units for openly gay or effeminate inmates, and transvestites to protect them from predatory behavior. For 15 years a central Florida jail segregated homosexuals and forced them to wear pink arm bands (Associated Press, 1989).

Since the AIDS epidemic began, many jurisdictions have been faced with decisions about the advisability of segregating asymptomatic HIV-positive prisoners or those with AIDS. Segregation decisions have been justified on the grounds of inmate security, the possible risk of transmission of HIV, or availability of specialized health services. In some aspects the prison debate mirrors concerns of health care providers about the creation of AIDS-dedicated hospital wards or medical facilities (see Chapter 3). Is the creation of an AIDS-dedicated prison unit a way of delivering health care more efficiently by those with specialized training, a way of protecting the health of inmates whose immune systems are compromised, or merely an administrative convenience, which will exacerbate the stigma that attaches to AIDS?

A number of potential harms are inherent in blanket segregation of HIV-infected inmates. Isolating HIV-positive prisoners labels them in the eyes of all other inmates and staff and may put them at greater risk of assault and discrimination.⁸ Segregation often limits prisoners' access to a wide range of prison activities, such as religious services, visitation, and drug treatment programs (e.g., Alcoholics Anonymous or Narcotics Anonymous). It also limits access to libraries, educational and recreational facilities, and work: many jurisdictions exclude known HIV-positive prisoners from food service positions, despite the lack of evidence of any danger in this regard.⁹ In some prisons, segregation of inmates with HIV disease has resulted in harrowing conditions, some of which have been the subject of journalistic exposés and court challenges.¹⁰ Furthermore, the segregation of HIV-positive prisoners may give a false sense of security about the risk of HIV transmission, however. Because of the window period for seroconversion, even widespread screening programs are unlikely to identify all HIV-positive entrants, and this HIV transmission may still be a possibility within the general prison population. As one official reported (Maisonet, 1990:96-97):

[with segregation of HIV-positive inmates] the inmates themselves believe that they are now safe. My greatest problem now is having putatively heterosexual men continually solicit sexual favors from our effeminate male homosexuals. … Segregated housing has created a myth that we don't have to worry about HIV in the general [prison] population. … Most inmates who are HIV positive have not been identified and are still involving themselves in high risk behavior.

HIV status can overwhelm a wide range of relevant considerations in decisions about where to house prisoners. According to Catherine A. Hanssens, of the New Jersey Public Advocate's Office (National Commission on AIDS, 1991:23):

In the New Jersey prison system, AIDS is the great equalizer; a prisoner's AIDS diagnosis substitutes for the system of classification based on the offense, prior record and incarcerations, institutional behavior, staff evaluations and similar factors by which all other offenders, including those with other types of chronic illnesses, are judged. Only one other group of prisoners in the New Jersey state system is subject to automatic segregation without periodic review for consideration of return to the general population—those under a sentence of death.

At least 20 state prisons segregate all prisoners with AIDS; 8 segregate those with AIDS-related complex; and 6 segregate inmates who are HIV positive but not symptomatic (Moini and Hammett, 1990). The courts have thus far rejected efforts to either compel or stop the segregation of prisoners with HIV disease. Judges have viewed isolation and segregation as matters of prison administration rather than public health and have shown a "marked propensity" to "uphold the administrative discretion of corrections officials unless their conduct is arbitrary or capricious" (Gostin, Porter, and Sandomire, 1990:18).

The trend in state and federal prisons has been away from blanket policies that segregate prisoners solely on the basis of HIV status. In a majority of state systems, individualized clinical or behavioral assessments are the basis for segregation decisions. In city and county jails, segregation of all AIDS cases is only slightly more prevalent than individualized determinations, 46 and 43 percent, respectively (Moini and Hammett, 1990).

Visitation Policies

In states with many prisoners with HIV disease, prison administrators have also had to reexamine their policies related to visitation, including conjugal visitations. Prior to the HIV/AIDS epidemic, correctional administrators advanced a number of justifications for conjugal visits: conjugal visits provide for sexual and emotional release; some believe that they reduce the level of homosexual activity in prison, although there is no definitive evidence in that regard; and conjugal visits may help provide for a smoother release back into society by helping to maintain or reestablish family ties. Their attractiveness to prisoners means that they are a significant privilege and their granting or denial can be a reward or punishment (Bates, 1989). In many states, inmates with HIV disease have been refused not only conjugal visits with spouses but also visits with parents, siblings, or children. Gay male and lesbian prisoners are never allowed conjugal visits with their lovers.

Prisoners' Health and Access to Care

It is difficult to gauge the impact of HIV disease on the health of the prison population, in part because of the lack of epidemiologic data related to prisoners' health status. Most studies tend to focus instead on the adequacy of health services (Hammett and Dubler, 1990). The prison population, as noted, is drawn disproportionately from minority groups and the urban poor, whose overall health status has been declining in recent years. Studies of the health of prisoners prior to the AIDS epidemic cited a number of problems related to poverty, drug use, and lack of access to care that had imperiled inmates' health before they were incarcerated (Barton, 1974; Marini, Bridges, and Sheard, 1978). One study summed up the current situation by stating that "prisoners now arrive at lock up sicker than at any time in the last 50 years" (Shenson, Dubler, and Michaels, 1990:655).

A report from the Montefiore Medical Center (1990), which provides medical and mental health care services under contract to Rikers Island, a New York City jail, recently called for more systematic attention to gathering data on prisoners' health status. The report did, however, record its impressions of the health status of prisoners under the care of the Montefiore Medical Center (1990:3):

Because many prisoners come from the most disadvantaged sectors of our population, they reflect the epidemiology of these communities. Recent trauma is often encountered in the clinical setting. … HIV infection, venereal diseases, hypertension, substance abuse, asthma, abnormal liver function tests, dental cavities and missing teeth are frequently seen (p. 2).

The report goes on to speculate about the impact of HIV (p. 3):

A high prevalence of HIV infection has dramatically altered the nature of routine medical care. … Generalized lymphadenopathy, oral candidiasis, herpes zoster, and seborrheic dermatitis have increased in frequency and must now be evaluated in the context of HIV-related superinfections such as tuberculosis, pneumocystis pneumonia, cryptococcal meningitis, and cryptosporidial infection. Provision of medical therapies directly related to HIV infection, such as zidovudine, immunizations, and pneumocystis prophylaxis, require significant allocations of medical, nursing, and pharmacy services.

One barometer of the imperiled health status of inmates has been the resurgence of tuberculosis in correctional facilities (Braun et al., 1989; Snider and Hutton, 1989). Unlike HIV, tuberculosis can be spread through the air, and poor sanitary conditions and prison overcrowding contribute to its spread. HIV is a major risk factor for the development of tuberculosis among individuals infected with the tubercle bacillus. Intravenous drug users appear to be at higher risk for tuberculosis than other HIV risk groups. One study of a cohort of 260 HIV-positive intravenous drug users in Baltimore sought to assess the relative risk of purified protein derivative (PPD) tuberculin positivity: age, receiving public assistance, history of arrest, and duration of drug use were associated with increased relative odds of being PPD positive. The study illustrates the difficulties in sorting out the discrete impact of incarceration. Although the association between arrest and PPD positivity suggests that transmission of tuberculosis infection may be occurring in prisons, the authors' multivariate analysis actually suggests that "previous arrest is just another marker for an impoverished lifestyle, extended time using drugs, and other risk behavior" (Graham et al., 1992:373). A Centers for Disease Control (1989) survey of 29 states in 1984-1985 found inmates to be three times as likely to develop tuberculosis as age-matched controls who were not incarcerated. In 1989, all of the 70 cases of active tuberculosis identified in New York State prisons were among prisoners who were also HIV positive (National Commission on AIDS, 1991).

Prison health care embodies some distinct paradoxes. The goals of medicine and corrections are frequently at cross-purposes. Medicine seeks to heal, extend life, and relieve suffering. In furtherance of these goals, a high value is placed on the quality of the doctor-patient relationship and the confidentiality that helps to foster it. Prisons exist to confine, punish, and perhaps, rehabilitate. Many aspects of the correctional enterprise impinge on health care delivery. Moreover, according to Prout and Ross (1988:130): "Access to medical care is a loaded emotional issue for inmates who have a lot of time to think about their own physical and emotional states, who crave opportunities to test the system, and who feel isolated and alienated from normal, safe human contact."

Health workers may also find the prison system difficult to negotiate. Physicians must cede to prison administrators some of the authority and control they have in most health care settings. The milieu in which prison health care is delivered is also generally dreary. The preeminence of security concerns means that health care workers, together with the inmates, are isolated from the community while they work. These factors combine to make it difficult to recruit and retain quality medical care staff, yet dedicated and professional medical staff make a major difference in the care of HIV disease within prisons. In a few prison settings, prisoners with AIDS have been offered state-of-the-art drug therapies because individual physicians have been diligent in seeking out information from knowledgeable colleagues and pressing institutional administrators for support.

Prisons have historically been backwaters of health care, yet inmates are virtually the only group with a constitutional right to health care.¹¹ In 1976 the Supreme Court held that "deliberate indifference" to the serious medical needs of inmates violates the Eighth Amendment to the Constitution that bars cruel and unusual punishment (Estelle v. Gamble, 429 U.S. 97 (1976)). The Court reasoned that because prisoners are unable to seek care anywhere else while in custody, to fail to provide them care would be "unnecessary and wanton" infliction of pain. The "deliberate indifference" standard articulated by the Supreme Court and fleshed out in subsequent case law is not a very demanding one. The establishment of a constitutional right to at least some level of health care has been a useful tool in litigation, but advocates for better prison health care still have an uphill battle. Burris (1990:3, 14) provided the following interpretation of the law in the context of AIDS-related health care:

For the most part, the law provides prisoners, as such and as people with HIV, with negative rights—rights not to be abused. The legal system is more open to a claim that a particular medical procedure was improperly denied than that inmates are entitled to the most effective medical care possible, more open to a claim that an inmate has suffered discrimination than a claim that a prison ought to be educating staff and prisoners about HIV. …

Prison medical care is not required to be very good, and many prisons live down to that low standard. AIDS patients do not get very good treatment, but neither do heart patients or back patients. When conditions are bad enough, a general attack on medical care, or care of people with AIDS, may have a better chance of success than a single inmate's complaint, but such a case requires an enormous investment in collecting and presenting the factual evidence. On the other hand, a suit for a specific treatment known to be effective, like AZT, may be easier to conduct than a global challenge, but a judge who does not see the systemic failures in care is more likely to indulge what a prison will likely claim is an isolated failure.

Many state prisons and local jails do not have the facilities to treat sick prisoners. Often the highest level of care available is infirmary care. New York State has a total of 36 beds available for treating prisoners with AIDS; an estimated 1,200 inmates have symptoms of HIV. In New York City, the jails have access to six skilled-care nursing beds, which are always filled. According to prisoners' rights advocates, the lack of prison health facilities leads to "chaining, like dogs, sick and debilitated prisoners to their hospital beds in regular civilian wards—a practice called 'outposting'—or in being shuttled between hospital and prison infirmary—a practice referred to by City officials as 'ping-ponging'" (Wiseman, 1990:8).

The lack of appropriate, state-of-the-art health care has taken its toll on inmates' lives, at least according to some observers. A study conducted by the Correctional Association of New York found that the median time between diagnosis and death was 159 days for intravenous-drug-using prisoners with AIDS, compared with 318 days for nonprisoner intravenous drug users. Of the cases reviewed, 25 percent were not diagnosed as AIDS until autopsy (Dubler, Bergmann, and Frankel, 1990:368, n. 24). One New York study found a mortality rate of 22 percent for prisoners with a first bout of Pneumocystis carinii pneumonia (PCP), the most common cause of AIDS deaths, compared with 8 percent for patients in the community. The key recommendation of the Presidential Commission on the Human Immunodeficiency Virus Epidemic (1988:135) in the area of corrections was that ''care and treatment available to HIV-infected inmates in correctional facilities should be equal to that available to HIV-infected individuals in the general community." The gulf between what is generally available to HIV-infected prisoners and what is available on the outside, however, appears to be widening with the advent of early intervention to treat HIV disease.

According to National Institute of Justice surveys (whose data, as noted, are admittedly incomplete and reflect official policy, not necessarily actual practice), by November 1989 (Hammett and Dubler, 1990:489) "less than one-third of correctional systems were providing AZT to all HIV seropositive inmates with CD4 … counts below 500," although all systems with 50 or more cumulative AIDS cases reportedly met the Public Health Service's standard for providing for AZT. PCP prophylaxis was not as readily available (Hammett and Dubler, 1990:489):

Just over half (58 percent) of state/federal prison systems and less than half (39 percent) of city/county jail systems … had policies in compliance with these PHS standards. … Less than half of the correctional systems with more than 50 cumulative AIDS cases (four of 11 prison systems and two of five jail systems) meet new standards for aerosolized pentamidine.

In evaluating prisoners' access to care and treatment for HIV, it is important to keep in mind a critical feature of prison health care. Health care for prisoners is supported by state and federal corrections budgets. Prisoners are not eligible for conventional public entitlement health care programs, such as Medicaid or Medicare. In most instances, funds for HIV care in prisons have had to come from corrections health care budgets, and they have remained static as prison populations have skyrocketed. In New York State, for example, two-thirds of the correctional system's health care budget of approximately $100 million is earmarked for HIV care. In New Jersey, the state supreme court stopped the practice of giving hospitalized prisoners emergency releases in order to avoid putting the burden of paying for health care on public hospital programs (Saint Barnabas Medical Center v. Essex County, 111 N.J. 67 (1988)).

One trend of interest—although not directly related to the HIV/AIDS epidemic—is the "privatization" of prisons. Increasingly, states are contracting with private enterprise to build, staff, and operate prisons, including providing security, food, transportation, and medical care. A number of models of prison health care delivery have been developed with various combinations of public and private services. In some cases, local public health departments or nearby university hospitals are involved in delivering prison health care. In others, private companies, operating in a manner similar to health maintenance organizations, provide health care services under a fixed, annual contract. Incentives to keep costs down may have a particularly harsh impact on HIV-infected prisoners and make it especially difficult for asymptomatic, HIV-positive prisoners to obtain adequate care.

Access to Experimental Treatments

AIDS tests the limits of prison health care because treatments tend to be expensive and difficult to deliver. Often the only treatments possible are experimental and are available only through clinical trials or expanded access programs (see Chapter 4). Prisoners confront formidable barriers in trying to gain access to drugs on clinical trials, and prisoners have access to experimental drugs in early stages of development in only a handful of corrections systems.

Historically, prisoners have been favored research subjects because of their accessibility and the limited expense incurred by their participation. The practice dates back to ancient Rome, where poisonous substances were tested on prisoners. In this century, the Nazi medical experiments on prisoners of war spawned new codes of medical ethics when what had transpired in the name of science came to light at the Nuremberg trials. In the decades following World War II, prisoners in the United States were frequently the subject of medical research, sometimes without their knowledge. Exposés in the 1970s brought to public attention some of the practices involved in testing drugs and vaccines on prisoners, who in journalistic parlance, were "cheaper than chimpanzees" (Mitford, 1973). Many observers believed prisoners' consent to the research to be highly suspect. Simple inducements of decent food, clean sheets, and medical care could mean a great deal to inmates who were otherwise deprived of many necessities and amenities. A stipend of a few dollars a day for participating in research might be considered coercive when compared with the 10 to 25 cents an hour typically earned for prison work.

A new era in medical research began in the 1960s and 1970s with the heightened concern for the rights and welfare of patients and research subjects (National Commission on AIDS, 1991; see Chapter 4). In 1978 federal regulations were adopted to address such issues as informed consent and voluntariness in participation, subject selection, confidentiality, and independent review related to federally funded research (45 C.F.R. 46.301-306 (1978)). In 1983 those regulations were amended to include special provisions that made it difficult to conduct clinical research with prisoners as subjects (45 C.F.R. 46.306 (a)(2)(d)(1983)). Not all prisoners welcomed the benevolent paternalism of the regulators, however. Many prisoners had willingly accepted the risk of participation in research, welcoming both the stipends and the diversion from the daily routine. Moreover, much of the nontherapeutic drug testing in prisoners, involving initial toxicity studies of new drugs, was of minimal risk (Schroeder, 1983). Some prisoners' rights groups asked that inmates be allowed to continue to participate, but their arguments failed to win the day with prison officials, physicians, and regulators.

One lasting impact of the AIDS epidemic on prisons may be a new attitude concerning the participation of prisoners in clinical and epidemiologic research, both on HIV and other diseases of concern in correctional settings. Prior to the AIDS epidemic, most prison medical experimentation involved nontherapeutic research—toxicity assessments and studies of treatments for diseases the prisoners did not have. Today, regulations promulgated to protect prisoners from overreaching and abuse in nontherapeutic research are now perceived as obstructing their access to needed medical care. With HIV disease, the line between research and treatment has never been more blurred.

A number of policy groups have been convened in recent years to address the role of prisoners in clinical and epidemiologic research on AIDS. A group convened by the Division of Law and Ethics in the Department of Epidemiology and Social Medicine at Montefiore Medical Center in New York City concluded that a reasonable interpretation of federal research regulations currently in effect would allow participation of prisoners in clinical trials so long as use of a placebo is not part of the study design (Hammett and Dubler, 1990; Dubler and Sidel, 1991). With this understanding, most AIDS research protocols would be open to prisoners. Other groups, such as the one that met under the auspices of the Washington, D.C.-based AIDS Action Foundation in 1990, have concluded in principle that "with the proper safeguards in place—voluntary decision making, confidentiality, and protections against abuse—prisoners should be permitted access to Phase II and III trials" (Hammett and Dubler, 1990:492).