POST-EXPOSURE PROPHYLAXIS

Currently, post-exposure prophylaxis (PEP) is available almost exclusively to persons who may have been exposed to HIV in the health-care occupational setting. The ethical and legal acceptability of this practice is now being challenged on the grounds that all persons who may have been exposed to HIV should have equal access to preventive treatment, regardless of the source, the setting (for example, occupational/non-occupational), or the nature of the potential exposure (for example, the route of transmission and the severity of exposure). We reprint here the most relevant sections of an article originally written for the "National Conference on HIV Post-Exposure Prophylaxis in the Non-Occupational Setting: Decision-Making in the Face of Uncertainty," held in Toronto on 23-24 October 1998, and since published in AIDS & Public Policy Journal 13; 3: 106-133. The article first calls into question the underlying assumption that PEP for HIV is a sufficiently safe and efficacious preventive treatment, pointing out that PEP for possible exposure to HIV, regardless of the source and nature of exposure, is most appropriately viewed as a non-validated practice requiring further research. It then analyzes various questions relating to funding, access, informed choice, confidentiality, and liability for PEP. It concludes by suggesting that, while maintaining the status quo - according to which PEP is available mostly to persons with possible health-care occupational exposures to HIV - does not appear to be an option, simply amending current policies to cover PEP for all possible exposures to HIV should be avoided. It suggests a number of questions that policy-makers should consider when working to improve current policies.

Expanding Access to Post-Exposure Prophylaxis: Ethical and Legal Issues

This article examines, from a Canadian perspective, the legal and ethical issues relating to post-exposure prophylaxis (PEP) for possible non-occupational exposure to HIV. Of primary concern is the risk of HIV transmission through sexual contact (for example, victims of sexual assault, individuals who engage in consensual unprotected anal or vaginal intercourse) or penetration of the skin (for example, injection of drugs, tattooing, and so on using unclean equipment).

The focus of this article is on issues relating to funding, access, informed choice, confidentiality, and liability for PEP. First, however, the differences between research, non-validated practice, and standard practice are reviewed and the underlying assumption that PEP for HIV is a sufficiently safe and efficacious preventive treatment is called into question. ⁽¹⁾

Research, Non-Validated Practice, or Standard Practice?

Currently PEP is available almost exclusively to persons who may have been exposed to HIV in the health-care occupational setting. The ethical and legal acceptability of this practice is now being challenged on the grounds that all persons who may have been exposed to HIV should have equal access to preventive treatment, regardless of the source, the setting (for example, occupational/non-occupational), or the nature of the potential exposure (for example, the route of transmission and the severity of exposure). In researching this issue, it quickly became apparent that the debate about expanded access to PEP rests in part on a critical assumption made by some: namely, that the use of PEP for possible occupational exposures to HIV is a sufficiently safe and effective preventive treatment. This assumption informs the claim about unjust discrimination in denying PEP to persons potentially exposed to HIV on the basis of a morally irrelevant consideration (for example, the setting in which exposure occurred). This assumption is controversial.

Anecdotal experience, a retrospective case-control study of PEP for health-care workers, ⁽²⁾ the findings from the ACTG 076 study on the effects of zidovudine (ZDV, AZT) on perinatal HIV transmission, ⁽³⁾ and some animal studies ⁽⁴⁾ suggest that zidovudine may be an efficacious preventive treatment for persons potentially exposed to HIV. On the basis of this limited data, some physicians seeking to provide care for their patients have been prescribing an antiretroviral drug regimen (a monotherapy regimen or a combination regimen) ⁽⁵⁾ they believe to be an efficacious prophylaxis. ⁽⁶⁾ Indeed, according to some, this is now the standard of care for health-care workers (for example, physicians, nurses, laboratory technicians) who have possible occupational exposures to HIV. ⁽⁷⁾

This practice and other factors have led governments and others to ask questions about funding, access, liability, and other issues relating to PEP for all persons possibly exposed to HIV, regardless of the setting in which the exposure may have occurred. Answering these questions in a responsible manner, however, requires serious deliberation about a prior question, namely whether PEP is research, non-validated practice, or standard practice.

The term "research" refers to a class of activities designed to develop or contribute to generalizable knowledge. Generalizable knowledge consists of theories, principles or relationships (or the accumulation of data on which they may be based) that can be corroborated by accepted scientific observation and inference.

The "practice" of medicine ... refers to a class of activities designed solely to enhance the well-being of an individual patient or client. The purpose of medical ... practice is to provide diagnosis, preventive treatment, or therapy. The customary standard for routine and accepted practice is a reasonable expectation of success. The absence of validation or precision on which to base such expectation, however, does not in and of itself define the activity in question as research.

... A practice might be nonvalidated because it is new; i.e., it has not been tested sufficiently often or sufficiently well to permit satisfactory prediction of its safety or efficacy in a patient population ... [or] because in the course of its use in the practice of medicine there arises some legitimate cause to question previously held assumptions about its safety and efficacy. It might be that the practice was never validated adequately in the first place ... or ... a question has been raised of a previously unknown serious toxicity. ⁽⁸⁾

In addition to differences in knowledge about safety and efficacy among research, non-validated practice, and standard practice, there are also important differences with respect to objectives, target populations, and professional consensus. The last of these features is particularly relevant:

with clinical practice there is a "professional consensus" as to the therapeutic merits of the treatment. With clinical research ... there is honest professional disagreement about the relative therapeutic merits of alternative interventions. The aim of the research is to resolve this dispute.... [W]ith non-validated practice there is honest belief on the part of some members of the profession [a respectable minority] about the therapeutic merits of the new drug, device or procedure. In time, this may give rise to honest professional disagreement. ⁽⁹⁾

In sum, despite a shared therapeutic intention, "non-validated practices" are by definition distinct from "practice," owing to the absence of sufficient reliable data about safety and efficacy and the lack of consensus among practitioners possessing the relevant expertise about the therapeutic merits of the intervention (that is, the intervention has "neither been accepted nor discredited by the expert clinical community"). ⁽¹⁰⁾ This is not to deny that much of current "practice" is in fact non-validated, but rather to highlight this as a danger to be avoided. ⁽¹¹⁾ Ideally, an intervention should not move from the realm of "non-validated practice" to "practice" without validation.

Because PEP is currently provided in a treatment context to health-care workers with potential occupational exposure to HIV, and not in a randomized controlled trial, ⁽¹²⁾ there is the tendency to presume that it is a safe and efficacious preventive treatment. In fact, however, data regarding the safety of PEP are incomplete, and definitive evidence of its efficacy is lacking, as is professional consensus regarding its therapeutic merits.⁽¹³⁾Attention to these facts is important, as history reminds us that patients can be harmed by well-meaning physicians. Consider the following notorious example. In the 1950s antibiotics were routinely administered to premature infants who were believed to be at risk of bacterial sepsis, until a study by Burns et al ⁽¹⁴⁾ concluded that administering penicillin and streptomycin gave no benefit to these infants and might "not be harmless." ⁽¹⁵⁾ The effect of a third antibiotic was more chilling; the groups of infants given chloramphenicol had "significantly higher" ⁽¹⁶⁾ mortality rates, and the researchers concluded that "chloramphenicol in the dosage used must have been responsible for the increase in mortality observed". ⁽¹⁷⁾

The dangers in following clinical instincts and impressions, and in extrapolating from animal studies or from experience with different patient populations, can be significant. The fact that zidovudine is effective in treating acute HIV infection and in limiting perinatal HIV transmission does not necessarily mean that zidovudine alone or in combination with other drugs is a safe and efficacious prophylaxis - encouraging animal studies on the pathophysiology of HIV infection ⁽¹⁸⁾ and some current professional opinion notwithstanding.

The most recent Public Health Service Guidelines published by the Centers for Disease Control and Prevention (CDC) in the US frankly acknowledge the limitations of the available scientific evidence on PEP for possible exposure to HIV:

The limitations of all of these studies must be considered when reviewing evidence of PEP efficacy. The extent to which data from animal studies can be extrapolated to humans is largely unknown, and the exposure route from mother-to-infant HIV transmission is not similar to occupational exposures; therefore these findings may not reflect a similar mechanism of ZDV prophylaxis.... Although the results of the retrospective case-control study of HCWs suggest PEP efficacy, the limitations of that study include the small number of cases studied and the use of cases and controls from different cohorts. ⁽¹⁹⁾

Additional potential limitations with the study involving health- care workers include possible reporting bias and ascertainment bias. ⁽²⁰⁾ More generally (and perhaps most important), a retrospective case-control study is not an optimal study design for evaluating the efficacy of PEP. With this type of study there is a higher risk of error, and so a lower level of evidence on the basis of which to make policy recommendations with confidence. ⁽²¹⁾

Taken together, this all suggests that PEP for possible exposure to HIV (regardless of the source and nature of exposure) is most appropriately viewed as a non-validated practice that "should ... be made the object of formal research at an early stage in order to determine whether [it is] safe and effective." ⁽²²⁾ In research, the "gold standard" is the randomized controlled trial because of the high level of confidence researchers can have in the validity of the results (absence of error); "it is the most effective and efficient means available to pursue the central purpose of biomedical research in the field of therapeutic innovation ... [namely,] to develop therapies that will accomplish the goals of curing or preventing diseases or of ameliorating their manifestations." ⁽²³⁾ And, more precisely, when there is no accepted therapy or intervention, the "gold standard" is the placebo-controlled randomized trial. ⁽²⁴⁾

For both scientific and ethical reasons, many believe that it is not possible to complete a placebo-controlled randomized trial of PEP. ⁽²⁵⁾ As the standards for introducing non-validated practices change, so too does the ability of researchers to collect data about safety and efficacy. This does not mean, however, that research on PEP is precluded. Alternate study designs with an open arm, or possibly using unequal randomization schemes ⁽²⁶⁾ and comparing different antiretroviral drug regimens, may be an option, so that research can proceed without interfering with the basic prophylactic objectives. ⁽²⁷⁾

Viewing PEP for possible exposure to HIV as a non-validated practice requiring further research means that discussions about PEP are not straightforward. While there are well-established legal and ethical frameworks for "practice" (that is, diagnosis, preventive treatment, and therapy) and "research," generally there is no analogous well-established framework for "non-validated practice," and so, in considering each of the issues addressed in this paper, one must ultimately determine whether PEP for potential exposures to HIV should be treated more like standard practice or research, and then apply the corresponding framework. To assist those responsible for making such determinations, background information on the legal and ethical aspects of funding, access, informed choice, confidentiality, and liability for PEP is provided within both a research and practice framework. Policy-makers and health-care providers will then be in a position to make separate determinations, for each of these issues, as to whether PEP should be treated as a research or therapeutic manoeuvre.

[Note: the section of the article on funding, focusing on direct drug costs, is not reprinted here]

Access

While funding and access are often considered together, they are separate topics. For the purposes of the discussion that follows, it is assumed that funding for PEP is available, either from government, a pharmaceutical company, private insurance, or the individual. The question then is, "if an individual has possibly been exposed to HIV, should she/he have access to PEP, and if so, on what basis?"

One account of the history of PEP for health-care workers tells the poignant story of an employee at the National Institutes of Health who became infected with HIV and subsequently died. In response,

... the center decided to augment risk-reduction efforts by offering postexposure chemoprophylaxis ... to any staff members who were occupationally exposed to HIV.... Although some scientific evidence supported this decision, the most compelling reason behind the center's decision to offer chemoprophylaxis was non-scientific.... [to] send a message of worker advocacy and be seen as empowering exposed workers. ⁽²⁸⁾

Similar non-scientific reasoning presumably influenced decision-makers at research and health-care centres around the world, and soon there developed the common practice of offering PEP to health-care workers who might have been exposed to HIV. Because PEP was available to health-care workers, the limited data that emerged regarding the safety and efficacy of PEP applied narrowly to this population, and this helped to entrench the practice of limiting PEP to health-care workers. Now, with the increasing perception that PEP may benefit persons potentially exposed to HIV, there is reason to critically examine current practice, and specifically to review the criteria and context for access to PEP. ⁽²⁹⁾

In Canada there is a fundamental commitment to the principle of justice, and in health care an important aspect of this principle is the fair distribution of the potential benefits of treatment and research. As documented by the National Forum on Health: "Equality of access was one of the most important values consistently advocated. Canadians should have equal opportunity to achieve health and well-being and to receive health services according to their needs." A similar commitment can be found in the newly revised Canadian research guidelines which recognize the benefits of participation in research and prohibit the use of discriminatory inclusion criteria: "researchers shall not exclude prospective or actual research subjects on the basis of such attributes as culture, religion, race, mental or physical disability, sexual orientation, ethnicity, sex or age, unless there is a valid reason for doing so." ⁽³⁰⁾

While several jurisdictions in Canada, including Alberta, Yukon, and PEI, have policies that facilitate access to PEP for possible exposure to HIV outside the health-care occupational context, this is not uniformly the case across Canada. ⁽³¹⁾ The current practice in some jurisdictions of limiting PEP to health-care workers appears to be inconsistent with the ethical principle that access should not be denied to either treatment or research on discriminatory grounds. Is this inconsistency ethically defensible? ⁽³²⁾ And, from a legal perspective, does the inconsistency amount to discrimination that could be challenged under the Canadian Charter of Rights and Freedoms, ⁽³³⁾ or human rights legislation?

Access/inclusion criteria

Possible access/inclusion criteria for PEP are many and include: the setting in which the potential exposure occurs; severity of possible exposure; time elapsed between possible exposure to HIV and the start of PEP; and current HIV status of the exposed person and of the source. Each of these is discussed in detail below.

The setting in which potential exposure occurs

Arguments in defence of the status quo suggest that equal access to PEP for potential non-occupational exposures to HIV, particularly sex- and drug-related exposures, may:

• erode established prevention messages and indirectly encourage high-risk behaviours, such as unsafe sexual practices, because of the perceived availability of a "morning after pill"; ⁽³⁴⁾ and
• increase the risk of creating a drug-resistant strain of HIV if PEP is unsuccessful. ⁽³⁵⁾

In response to the first of these points, advocates of expanding PEP to all possible exposures counter that the potential problem of increased unsafe behaviour is highly speculative and unfounded, given how onerous the PEP regimen is. Further, any such risk could be "minimized through individual counselling and carefully worded public health messages." ⁽³⁶⁾ From a more general perspective, it is also argued that denying PEP to persons with potential non-occupational exposures to HIV is unfair and moralistic. For example, there can be inadvertent possible sexual exposure among those committed to safer sexual practices and unwanted possible sexual exposure among those who are victims of assault. Further, there is reason to question the underlying assumption that access to PEP should be denied when physicians or others assign personal responsibility for potential HIV infection. This is inconsistent with current practice in the Canadian health-care context - persons "responsible" for their ski accidents, their liver damage, or their lung cancer are not denied access to research, non-validated practice, or standard practice on the basis of some failure to promote/secure their health and well-being.

As for the concern about creating a drug-resistant HIV, this is not an argument against expanded access to PEP, since the risk (such as it is) applies regardless of whether the potential exposure is occupational or non-occupational. For example, if PEP is unsuccessful because of suboptimal compliance with or failure to complete the prescribed drug regimen, the route of transmission is scarcely relevant. Secondly, the true incidence and significance of creating drug-resistant HIV strains is unknown. Arguably, the risk of inducing and selecting for drug- resistant viral strains with short-term prophylaxis is not very high; some believe, however, that it may be increased when there is repeated exposure to HIV. This may be a more likely occurrence with possible non-occupational exposures to HIV that are the result of certain lifestyle choices.

The severity of potential exposure

Currently, access to PEP is not solely determined by the setting in which possible exposure occurs. Another important factor is the severity of the potential exposure. In recent years there have been efforts to assess the level of risk, and, on this basis, to determine which health-care workers with possible exposure to HIV should have access to PEP. One proposal is that PEP be:

1. Recommended for massive exposure (transfusion),

2. Endorsed for dramatic exposure (deep intramuscular/intravenous injection),

3. Available for probable exposure (mucosal/subcutaneous needle stick), and

4. Discouraged for doubtful exposure (low risk fluids). ⁽³⁷⁾

Another proposed algorithm suggests:

1. When the risk of HIV transmission is high ("highest risk" and "increased risk" of percutaneous exposure to a larger volume of blood and/or blood containing high titer of HIV), PEP should be recommended;

2. When the risk is lower but non-negligible (no increased risk of exposure to larger volume of blood or blood with high titer of HIV, or exposure to fluids containing visible blood or other potentially infectious fluids or tissue), PEP should be offered; and

3. When it is negligible (exposure to other bodily fluids), PEP is not justified. ⁽³⁸⁾

Most recently, the CDC has developed a decision tree to assist in determining when it is appropriate to proceed with PEP for health-care workers with potential exposure to HIV - the approach suggested is considerably more sophisticated than others previously available. It seeks to identify more clearly situations in which PEP may not be warranted because the potential side effects and toxicity of PEP outweigh the risk of HIV transmission, and distinguishes these situations from those where it is appropriate to:

1. Consider a basic regimen;
2. Recommend a basic regimen, and
3. Recommend an expanded regimen. ⁽³⁹⁾

Presumably determinations regarding access to PEP for possible non-occupational exposure to HIV would also require an assessment of the severity of the potential exposure. Clearly, not all such exposures will warrant PEP (for example, exposure to urine). As with the approach to access to PEP for health-care workers, all available information on risk of transmission in other contexts would have to be carefully considered in order to determine how to distinguish legitimately between non-occupational exposures where PEP is not warranted and exposures where PEP should either be considered or recommended. ⁽⁴⁰⁾

Time elapsed between possible exposure to HIV and the start of PEP

Animal studies suggest that PEP is not efficacious if started later than 24 to 36 hours after exposure to HIV. The time after which there is no benefit from PEP for humans, however, remains undefined. ⁽⁴¹⁾ According to some, the general consensus is that PEP should be started within 72 hours after potential exposure. ⁽⁴²⁾ But are there differences for infections that are transmitted sexually rather than percutaneously? Should the time frame for initiating PEP vary depending upon the source, the setting, or the nature of exposure? If PEP cannot be started within the short "window period," should access be denied on this basis? These are important questions if it happens that persons with potential non-occupational exposures to HIV are not likely to present within 72 hours after potential exposure. ⁽⁴³⁾

Current HIV status of the exposed person and the source

Because the window period for initiating PEP appears to be 72 hours, it is argued that HIV status should not be an access criterion, though consent to HIV testing should be sought - information from the testing may be relevant to the person potentially exposed. If the person does not test positive, that will be a relevant factor to consider in deciding whether to accept the potential long-term consequences of PEP in healthy individuals; if he or she is already HIV positive, then presumably treatment would be more appropriate than prophylaxis.

A question for further debate and study, however, is whether consent to HIV testing should be an access criterion. This issue, though not irrelevant in the occupational setting, is perhaps more acute with certain potential non-occupational exposures, where there may be a greater likelihood of multiple exposures to HIV, and thus a greater likelihood that the person may already be HIV-positive. With the emerging availability of accessible, rapid HIV testing, the legal and ethical implications of this issue cannot be ignored.

Legal aspects

Current practice is that access to PEP is largely (though not exclusively) determined by the setting in which the potential exposure occurred. The key legal issue therefore is whether a person who is denied access to PEP solely because the potential exposure was non-occupational could use the Canadian Charter of Rights and Freedoms or human rights legislation to challenge the denial of PEP as unjustifiable discrimination or infringement of the security of the person.

Discrimination: The Canadian Charter of Rights and Freedoms

For the Charter to be applicable at all, the denial of access would have to be based on a government policy, rather than simply on the decision of an individual physician, since the Charter has been held to apply to government, but not private activity. ⁽⁴⁴⁾ Then, in order to prove a violation of s 15(1) of the Charter, it would have to be shown that the denial was based on one of the grounds listed in that section (race, national or ethnic origin, colour, religion, sex, age, or mental or physical disability) or on a ground that the court would consider "analogous" (for instance, sexual orientation, although not listed in s 15(1), has been accepted as an analogous ground). ⁽⁴⁵⁾ It would be difficult to argue that a distinction between those potentially exposed in the workplace and all others potentially exposed to HIV is in itself a violation of s 15(1). However, if it could be shown that access was being denied on the basis of assumed HIV status, then this might be characterized as unequal treatment based on disability. ⁽⁴⁶⁾ Similarly, if it could be shown that a significant number of those being denied access were, for instance, drug addicts or gay men, then the grounds of disability or sexual orientation might be used. However, not all denials of PEP would necessarily fit within a listed or analogous ground under section 15(1). Furthermore, even if a court held that s 15(1) had been violated, the government would then have the opportunity to try to show, under s 1 of the Charter, that the restriction being challenged was a reasonable limit that could be "demonstrably justified in a free and democratic society." ⁽⁴⁷⁾

Discrimination: human rights

This legislation applies to both government and private activity. All human rights statutes in Canada ⁽⁴⁸⁾ prohibit discrimination in the provision of services customarily available to the public; the argument would therefore be that PEP for possible exposure to HIV is such a service, and that the refusal to provide PEP to a particular individual was based on a ground protected by the relevant human rights act. The Canadian Human Rights Act, ⁽⁴⁹⁾ for instance, prohibits discrimination based on "race, national or ethnic origin, colour, religion, age, sex, sexual orientation, marital status, family status, disability and conviction for which a pardon has been granted." Again, the issue would be whether the grounds on which PEP was denied would fit within one of the listed categories.

Security of the person

Another possibility might be to argue that s 7 of the Charter, which states, "everyone has the right to life, liberty and the security of the person and the right not to be deprived thereof except in accordance with the principles of fundamental justice," requires the government to ensure access to medical treatments. In the 1997 case Sanders v Ontario (Ministry of Community and Social Services), ⁽⁵⁰⁾ the applicant sought a declaration that the s 7 rights of a psychiatric patient had been infringed by the Crown's refusal to allow her to undergo a particular treatment which had been recommended by three psychiatrists. The court dismissed the application, stating, "Should the courts dictate to the Crown what treatments will be offered in their various facilities? Both philosophically and practically, the Crown's discretion in this regard ought not to be interfered with except in the clearest and most serious cases." ⁽⁵¹⁾ In reaching this conclusion, the court was clearly influenced by the fact that the proposed intervention was "controversial" rather than a "generally accepted treatment." ⁽⁵²⁾

To  consider

If PEP is provided in a therapeutic context, there is the danger highlighted earlier of having a non-validated practice move into the realm of practice without validation. To avoid this, PEP would have to be provided in a research context where uncertainty is acknowledged and managed, there being a plan to systematically gather knowledge/evidence about safety and efficacy.

As discussed in the introduction, the "gold standard" of a randomized controlled trial may not be an appropriate approach for research on PEP for possible exposure to HIV; however, this does not mean that attempts to gather further data on the safety and efficacy of PEP should be abandoned. Also, this does not mean that access to PEP on an "off protocol," "expanded access" or "parallel track" basis, for compassionate ⁽⁵³⁾ or other reasons (for example, ineligibility for research participation, completion of enrollment in the protocol), is precluded. In closing:

The right to medical treatment does not encompass every drug or intervention that a patient [or physician-researcher] considers therapeutically worthwhile, on whatever evidence he or she has found convincing. Until the therapeutic advantage of an innovation has been demonstrated to the satisfaction of the community of expert practitioners, an innovation is no medical treatment, and so is not covered by a patient's right to access to medical treatments. Consistent with the above, however, we may add that to the extent that a patient's desire to receive [and a physician-researcher's desire to provide] an innovation may be satisfied without jeopardizing the conduct of a clinical trial, then, consistent with good clinical judgment, the desire of the patient [physician-researcher] should be allowed to prevail. ⁽⁵⁴⁾

Irrespective of whether PEP for potential exposure to HIV is provided in a research or treatment context, ethical principles would dictate that access should not be based on discriminatory distinctions. It is unclear, however, whether legal arguments for expanded access, based on the Charter and human rights legislation, would in all cases be successful.

[Note: the sections of the article on informed choice, confidentiality, and liability, are not reprinted here]

Closing Remarks

Although there is limited information regarding the safety and efficacy of PEP for possible exposures to HIV, the fact remains that PEP is currently available in a treatment context, with or without (partial) government funding, mostly to persons with possible health-care occupational exposures to HIV. As maintaining this status quo does not appear to be an option, the question before policy-makers is how to best modify the current approach to PEP.

One approach would be to introduce or amend provincial/ territorial policies to cover PEP for all possible exposures to HIV, striving for as much consistency as possible. This approach is intuitively appealing because it appears to be both simple and fair. The problem with this approach, however, is that it rests on two questionable assumptions: first, that existing policies are ideal, but for the provisions that limit access based on the setting in which a potential exposure occurs; and second, that there should be a uniform approach to all possible exposures to HIV. In our view, there is room to improve current policies, and in so doing to carefully consider possible differences among various potential exposures to HIV. For example, outside the health-care occupational setting, the following factors may be relevant: delayed recognition of possible exposure, lack of information regarding the HIV status of the possible source of exposure, inability to follow the prescribed regimen, and risk of repeated exposures during a period of prophylaxis. These factors may lessen the efficacy of PEP, or perhaps even increase the possibility of harm from PEP, and so must be considered in policy development.

The challenge brought forward by those who have been denied access to PEP provides policy-makers with a unique opportunity to step back and consider anew, in light of currently available data, the following questions:

• How safe and efficacious is PEP for possible exposures to HIV in the health-care occupational setting?
• How safe and efficacious is PEP for possible exposures to HIV in the non-occupational setting?
• What research is necessary in order to develop further knowledge about the benefits and harms of PEP?
• What ethical principles should inform government decision- making in this area?
• What are the Crown's legal responsibilities, in light of human rights legislation, the Canadian Charter of Rights and Freedoms, and tort law?
• How should PEP for possible exposure to HIV fit within an overall national strategy of HIV/AIDS prevention and treatment?
• What ethically and legally justifiable criteria should be developed to distinguish between possible exposures where PEP is not warranted and exposures where PEP should either be considered or recommended?
• What ethically and legally justifiable principles should form the basis for decisions regarding the funding of PEP?
• What procedures should be developed to ensure that individuals offered PEP are provided with sufficient information to make an informed choice, and that confidentiality is properly maintained?

- Françoise Baylis & Diana Ginn

Françoise Baylis, PhD, is an Associate Professor in the Office of Bioethics Education and Research, Faculty of Medicine, and in the Department of Philosophy at Dalhousie University, Halifax. She can be reached at <Francoise.baylis@dal.ca>. Diana Ginn, BA, LLM, is an Assistant Professor in the Faculty of Law at Dalhousie University. She can be reached at <Diana.ginn@dal.ca>.

The authors extend special thanks to The NAMES Project/Le Projet des NOMS - Canada for sponsoring this project, and to Dr Scott Halperin, IWK Grace Health Centre, for comments on earlier drafts of this paper. Thanks are also owed to Monica McQueen, Jan Sutherland, and Tania Trepanier for able research assistance.

Funding for this publication was provided by Health Canada. The views expressed herein are solely those of the authors and do not necessarily reflect the official policy of the Minister or Health Canada.

This document is also available on the Internet at the Office of Bioethics Education and Research website: <www.medicine.dal.ca/bioethics>.

1. Terminological precision and conceptual clarity are imperative to the goal of sound reasoning. "The correct description of an action is critical for ethical [and legal] evaluation. A kiss may be just a kiss; but, depending upon the context in which bestowed it may also be a mark of affection, the fulfilment of a contract for services, or as in the Godfather saga, the pronouncement of the death penalty." B. Freedman, A. Fuks and C. Weijer, "Demarcating Research and Treatment: A Systematic Approach for the Analysis of the Ethics of Clinical Research," Clinical Research 40 (1992): 653-60, p. 653.

2. D.M. Cardo, et al., "A case-controlled study of HIV seroconversion in health care workers after percutaneous exposure," New England Journal of Medicine 337 (1997): 1485-1490. See also, CDC Cooperative Needlestick Surveillance Group, "Case control study of HIV seroconversion in health care workers after percutaneous exposure to HIV infected blood--France, United Kingdom and United States, January 1994 - August 1994," Morbidity & Mortality Weekly Report 44 (1995): 929-33.

3. R.S. Sperling, et al., "Maternal viral load, zidovudine treatment, and the risk of transmission of human immunodeficiency virus type 1 from mother to infant," New England Journal of Medicine 335 (1996): 1621-1629; E.M. Connor, et al., "Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment," New England Journal of Medicine (1994): 1173-1180.

4. R.J. Black, "Animal studies of prophylaxis," American Journal of Medicine 102, Suppl. 5B (1997): 39-44.

5. The most recent guidelines from the Centers for Disease Control and Prevention (CDC) for PEP recommend a two drug regimen in certain circumstances, namely zidovudine (AZT) and lamivudine (3TC), with the addition of a protease inhibitor (indinavir or nelfinavir) in certain other circumstances, where an increased risk for transmission exists: CDC, "Public Health Service Guidelines for the Management of Health-Care Worker Exposure to HIV and Recommendations for Postexposure Prophylaxis," Morbidity & Mortality Weekly Report 47, Suppl. No. RR-7, (May 15, 1998): 1-14, p. 1.

6. D. Shelton, "Not a 'morning after' pill: But demand spurs look at HIV antiretroviral use," American Medical News 40 ( 11 August 1997): 1 (from the American Medical Association website: <www.ama-assn.org/public/journals/amnews>.

7. Centres for Disease Control, "Update: provisional Public Health Service recommendations for chemoprophylaxis after occupational exposure to HIV," Morbidity & Mortality Weekly Report 45 (1996): 468-472; J.L. Gerberding, "Prophylaxis for occupational exposure to HIV," Annals of Internal Medicine 125 (1996): 497-501.

8. R.J. Levine, Ethics and Regulation of Clinical Research, 2d ed. (New Haven: Yale University Press, 1988): 3-4. These definitions are based on those first articulated in: National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research, DHEW Pub. No (OS) 78-0012 (1978): 1-20, p. 2-4.

9. F. Baylis, "Assisted Reproductive Technologies: Informed Choice," in New Reproductive Technologies: Ethical Aspects, vol. 1 of the research studies of the Royal Commission on New Reproductive Technologies (Ottawa: Minister of Supply and Services Canada, 1993): 47-107, p. 52-53.

10. B. Freedman, "Nonvalidated therapies and HIV disease," Hastings Center Report 19, no. 3 (1989): 14-20, p. 14.

11. "Some of the treatments used in everyday clinical practice, and that have acquired the reputation of being standard and established therapies, have never been subjected to the methodological evaluation given to treatments in a controlled clinical trial. The reputation of being a standard or established therapy is often based on little more than clinical rumor, anecdotal evidence, or fashion. In the past, a number of such "standard" treatments [...] had to be abandoned because they were eventually shown, under careful study, to have been either useless or definitely harmful." D.J. Roy, J.R. Williams and B.M. Dickens, Bioethics in Canada (Toronto: Prentice Hall, 1994): 319.

12. A prospective double-blind placebo-controlled trial of post-exposure prophylaxis among health care workers in the U.S. was discontinued because of limited enrollment and the large numbers needed to assess a reduction in the risk of transmission. See S.W. LaFon, S.N. Lehrman and D.W. Barry, "Prophylactically administered Retrovir in health care workers potentially exposed to the human immunodeficiency virus," Journal of Infectious Diseases 158 (Special Notice, 1988): 503. S.W. LaFon, et al., "A double-blind, placebo-controlled study of the safety and efficacy of Retrovir (zidovudine, ZDV) as a chemoprophylactic agent in health care workers exposed to HIV," in American Society for Microbiology, Program and Abstracts of the 30th Interscience Conference on Antimicrobial Agents and Chemotherapy, (Washington D.C.: American Society for Microbiology, 1990): 167, abstract no.489.

13. D. Henderson, "Postexposure treatment of HIV--Taking some risks for safety's sake," New England Journal of Medicine 337 (1997): 1542-1543. See also D.N. Fish, "Prophylaxis of HIV Infection Following Occupational Exposure," Annals of Pharmacotherapy 27 (1993): 1243-56.

14. L.E. Burns, J.E. Hodgman, and A.L. Cass, "Fatal Circulatory Collapse In Premature Infants Receiving Chloramphenicol," New England Journal of Medicine 261, no. 26 (1959): 1318-1321.

15. Ibid., 1321.

16. Ibid.

17. Ibid.

18. For example, L.N. Martin, et al., "Effects of initiation of 3'-azido, 3'-deoxythymidine (zidovudine) treatment at different times after infection of rhesus monkeys with simian immunodeficiency virus," Journal of Infectious Disease 168 (1993): 825-35.

19. Centres for Disease Control, "Public Health Service Guidelines," see note 5 above, p. 7.

20. For a more complete discussion of the limitations of the retrospective case-control study see the Editor's note in CDC Cooperative Needlestick Surveillance Group, "Case Control Study... " note 2 above: 932. Ascertainment bias refers to possible bias in the way that subjective information is gathered. As regards the study in question, the times at which some relevant information were collected varied between the two arms of the study. It has been noted that "ascertainment bias may have affected some data, particularly subjective variables such as severity of injury, because information for control-HCWs was obtained prospectively soon after exposure but information for most case-HCWs was obtained after seroconversion," (p. 932).

21. David Sackett provides a useful table that correlates levels of evidence and grades of recommendation. The higher the level of evidence, the higher the grade of the recommendation. See D. Sackett, "A Science for the Art of Consensus," Journal of National Cancer Institute 89, no. 14 (1997): 1003-05, p. 1004.

22. National Commission, The Belmont Report, see note 8 above, p. 3. According to Susan Buchbinder, "several studies are in progress or being planned to clarify the benefits and drawbacks of PEP for people exposed non-occupationally, so that health care providers can make sound recommendations", S. Buchbinder, "Avoiding Infection After HIV Exposure," Scientific American 279 (1998): 104-105. The Centers for Disease Control has established a "HIV Postexposure Prophylaxis (PEP) Registry" which, with the consent of patients, will "collect information about the safety, tolerability, and outcome of taking antiviral drugs for postexposure treatment." Identifying information will not be collected. From the <AIDSminingco.com> web site.) It should be noted, however, that this registry is intended only for health care workers with possible occupational exposure to HIV (according to the website) so the registry will not provide direct data on the safety or efficacy of PEP in other contexts.

23. R. Levine, "Comment on Richard Royall's 'Ethics and Statistics in Randomized Clinical Trials'," Statistical Science 6 (1991): 71-74, p. 72.

24. For Canadian guidelines on placebo-controlled studies, see The Medical Research Council of Canada, The Natural Sciences and Engineering Research Council of Canada, and The Social Sciences and Humanities Research Council of Canada, Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (Ottawa: Public Works and Government Services, 1998): p. 7.4.

25. J. Gerberding, "Is Antiretroviral Treatment after Percutaneous HIV Exposure Justified?" Annals of Internal Medicine 118, no. 12 (1993): 979-80.

26. "Mathematically rigorous data-dependent designs for studies can, in some cases, provide firm information on the effectiveness of a therapy while exposing fewer patients to suboptimal treatment than a conventional trial would. During a data dependent study, the health of the participants is continually analyzed. The accumulating evidence about the different treatments modifies the odds used to select which therapy the next participant will receive." T Beardsley, "Coping with HIV's Ethical Dilemmas," Scientific American 279 (1998): 106-107. [An open arm study design allows subjects and researchers to know what drug is being tested. In an unequal randomization scheme, a large group of subjects is randomized to a drug, and a smaller group receives another drug or placebo - ED].

27. R.J. Levine, Ethics and Regulation of Clinical Research, note 8 above, p. 4.

28. D.K. Henderson, "Postexposure,"note 13 above, p. 1542.

29. National Forum on Health, Canada Health Action: Building on the Legacy, vol. II (Ottawa: Minister of Public Works and Government Services, 1997): 6.

30. Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (1998): p. 5.2 Clearly, no one can be guaranteed access to pharmaceutical research in order to gain access to a particular drug, for a variety of reasons; the drug in question may not be the subject of research at that time, or the person may not the meet the inclusion criteria, or researchers may have all the research participants they need at a given time. The point is that if research is being carried out, and a person would be otherwise eligible, he or she should not be denied access on discriminatory grounds.

31. Federal - Health Canada "An Integrated Protocol to Manage Health Care Workers Exposed to Blood Borne Pathogens"; Manitoba - Manitoba Department of Health, Public Health Branch, "Guidelines for Managing Occupational Exposure to Blood/Body Fluids" / "Counselling Guidelines for Clients with Accidental Exposure to Blood or Body Fluids"; Yukon - Yukon Department of Health, "Blood-borne Exposure Management Protocol" ["this protocol is for all persons living or travelling in the Yukon who are accidentally exposed to blood-borne pathogens"]; Prince Edward Island - Prince Edward Island Department of Health and Social Services "Guidelines for the Management of Persons Exposed to Blood Borne Pathogens in An Occupational Or Community Setting"; Alberta - Alberta Ministry of Health "Health Standards for Non-Occupational Community Post-exposure Follow-up and Prophylaxis of Blood Borne Pathogens".

The distinction between those possibly exposed in a health care occupational setting, and those possibly exposed in other contexts appears to be the policy in England as well. See "After Sex Cocktail: A Gay man's struggle through the maze of post-exposure prophylaxis", from the website <AIDSminingco.com>.

32. For a variety of perspectives on whether the use of PEP should be expanded, see E. Zold et al., "Forum on Using Anti-HIV Drugs Soon After Sexual or Drug-Use Exposure" San Francisco Bay Area Reporter, 13 February 1997, as it appeared on the website: <http://hivinsite.ucsf.edu>; M.H. Katz and J.L. Gerberding, "Post Exposure Treatment of People Exposed to the Human Immunodeficiency Virus through Sexual Contact or Injection-Drug Use," (Sounding Board) New England Journal of Medicine 336, no. 15 (April 10, 1997): 1097-100; D.K. Henderson, "Postexposure," see note 13 above; and C.C.J. Carpenter et al., "Antiretroviral Therapy for HIV Infection in 1996: Recommendations of an International Panel," Journal of the American Medical Association 276, no. 2 (1996): 146-54, p. 152.

33. Canadian Charter of Rights and Freedoms, Part I of Constitution Act, 1982, being Sch. B of Canada Act, 1982, 1982, c. 11(U.K.).

34. "Can HIV be stopped within first 72 hours?" OUTReach (newspaper), June 1997, San Francisco AIDS Foundation, from their website: <www.sfaf.org>.

35. For a discussion of issues relating to drug resistant strains of HIV see M. Wainberg and G. Friedland, "Public Health Implications of Antiretroviral Therapy and HIV Drug Resistance," Journal of the American Medical Association 279, no. 24 (1998): 1977-83; O.J.Cohen and A.S. Fauci, "Transmission of Multidrug-Resistant Human Immunodeficiency Virus - The Wake-Up Call," New England Journal of Medicine 339, no. 5 (30 July 1998): 341-43; and Katz and Gerberding, "Post Exposure," note 32 above. See also note 6 above.

36. M.H. Katz and J.L. Gerberding, note 32 above, p. 1098.

37. D. Fish, "Prophylaxis," see note 13 above, p. 1252.

38. "Provisional Public Health Service recommendations for chemoprophylaxis after occupational exposure to HIV, by type of exposure and source material--1996," (table) in M.E. Chamberland and D.M. Bell, "Human Immunodeficiency Virus Infection", in J.V. Bennett and P.S. Brachman, eds., Hospital Infections, 4th ed. (Philadelphia: Lippincott-Raven Publishers, 1998): 665-87, p. 678.

39. Centers for Disease Control, "Public Health Service Guidelines," see note 5 above, p. 7.

40. Criteria for offering PEP to the survivors of sexual assault are described by G. Opio, R. Torres and R. Alvalle, "Post-sexual exposure prophylaxis (PSEP) with HAART after sexual assault," (paper delivered at the 12th World AIDS Conference, 28 June-July 3, 1998, Geneva), conference abstract no. 250/33174.

41. Centres for Disease Control, "Public Health Services Guidelines," see note 5 above, p. 18; Chamberland and Bell, see note 38 above.

42. Buchbinder, "Avoiding HIV Infection," see note 22 above.

43. Results from one study suggest that, in consensual sexual relations, persons potentially exposed to HIV are unlikely to present within 72 hours, in part because they are unaware of their partner's HIV infection. This suggests that the value of PEP for potential non-occupational exposure to HIV, if proven effective, may nonetheless be very limited. P. Sudrel, G. Schockmel, C. Fagard et al., "Post HIV exposure prophylaxis: Who may benefit?" (paper delivered at the 12^th World AIDS Conference, 28 June-3 July 1998, Geneva), conference abstract no. 33188.

44. Canadian Charter of Rights and Freedoms, s. 32; see also McKinney et al v. University of Guelph, [1990] 3 S.C.R. 229-449; Eldridge v. British Columbia (A.-G.), [1997] 3 S.C.R. 624. Note that there it must also be shown that there has been a denial of equality before or under the law, or a denial of the equal protection or benefit of the law. However, given the discussion of this issue in McKinney, it seems likely that a government policy would be found to meet the criterion of "law".

45. Egan v. Canada, [1995] 2 S.C.R. 513-626.

46. The argument might go as follows: you think I am HIV positive and until I prove to you I am not, you are refusing me PEP; thus I am being refused PEP because of a perception that I have a certain illness or medical condition. This constitutes a denial on the basis of disability.

47. Canadian Charter of Rights and Freedoms, s. 1.

48. Canadian Human Rights Act, R.S.C. 1985, c. H-6, s.5; Human Rights, Citizenship and Multiculturalism Act, R.S.A. 1980, c. H-11.7, s. 3; Human Rights Code, R.S.B.C. 1996, c. 210, s. 8; Human Rights Code, R.S.M. 1987, c. H175, s. 3; Human Rights Code, R.S.N. 1990, c. H-14, s. 6; Human Rights Act, R.S.N.B. 1973, c. H-11, s. 5; Human Rights Act, R.S.N.S. 1989, c. 214, s. 5(1)(a); Human Rights Code, R.S.O. 1990, c. H.19, s. 1; Human Rights Act, R.S.P.E.I. 1988, c. H-12, s. 2; Saskatchewan Human Rights Code, S.S. 1979, c. S-24.1, s. 12; Human Rights Act, S.Y. 1987, c. 3, s. 8(a).

49. Canadian Human Rights Act, R.S.C. 1985, c. H-6.

50. Sanders v. Ontario (Ministry of Community and Social Services) (1997) O.J. No. 552-56 (O.C.J. Gen. Div.) (From QL).

51. Ibid., at para. 7.

52. Ibid., at para. 8.

53. For a discussion of compassionate access, see O. Madore and W. Murray, National Roundtable on Compassionate Access to Experimental Therapies, (an unpublished report prepared for the House of Commons Sub-committee on HIV/AIDS, 10 January 1996).

54. B. Freedman, "Nonvalidated Therapies," see note 10 above, p. 18.