AIDS Denialism and Public Health Practice
AIDS and Behavior
Published online: 8 January 2010
Pride Chigwedere1 and M. Essex1
Harvard School of Public Health AIDS Initiative and Department of Immunology and Infectious Diseases, Harvard School of Public Health, FXB 402, 651 Huntington Ave, Boston, MA 02115, USA
"When AIDS denialism infiltrates public health practice, the consequences are tragic"..."from 2000 to 2005 at least 330,000 South Africans died prematurely and 35,000 babies were infected"
"Questioning whether HIV causes AIDS and the safety of using antiretroviral drugs (ARVs), the South African government led by former president Thabo Mbeki......some South Africans could have benefited, however, small the benefit, had Mbeki not obstructed drug use. ......we concluded that from 2000 to 2005 at least 330,000 South Africans died prematurely and 35,000 babies were infected with HIV as a result of Mbeki's policies......Mbeki implemented negligent policies that led to the premature death of hundreds of thousands......Mbeki withdrew support from clinics that had started using ARVs, restricted use of donated ARVs, obstructed Global Fund grants, and generally delayed implementing a national ARV program. Two independent studies have estimated that Mbeki's policies led to at least 330,000 premature deaths.....Duesberg..... moves from the claim that AIDS does not exist to a multiplicity of causes even if it means creating a different cause for different geographies, different time periods, and different demographic groups, and without producing a shred of evidence [28, 86, 87]. This is what is called denialism-"the rejection of objective reality to sustain a flawed, hurtful, and ultimately dangerous belief system" .......Contrary to what Duesberg suggests, there are unanimous data (all trials conducted in Africa published by December 2006) to demonstrate the usefulness of ARVs in PMTCT in Africa, and other groups have arrived at the same conclusions for ARV use in PMTCT generally....extensive clinical trials data demonstrate the efficacy of ARV drug combinations in treating AIDS.....When AIDS denialism enters public health practice, the consequences are tragic. The implications start in honest science but extend to the need for accountability and, perhaps, public health reform......AIDS denialists are dangerous to the general population; many have been persuaded into risky behaviors, ineffective alternative remedies, and other harmful actions"
"The science behind Mbeki was Duesberg and other denialists. Duesberg is still arguing for AIDS denialism and defending Mbeki and the policies that led to more than 330,000 deaths. By any reasonable standard, this requires some form of accountability. Seth Kalichman has likened the AIDS denialists to the Holocaust deniers  and Edwin Cameron likened letting AIDS patients die without medications to those who silently enabled the evils of Nazi Germany and apartheid South Africa to go unchecked . John Moore and Nathan Geffen have called for AIDS denialists to be put on trial [136, 137] and Mark Wainberg has argued that denialists should be charged with public endangerment and "people like Peter Duesberg belong in jail"
We recently published a paper estimating the human cost of not using antiretroviral drugs in South Africa . Questioning whether HIV causes AIDS and the safety of using antiretroviral drugs (ARVs), the South African government led by former president Thabo Mbeki withdrew government support from Gauteng clinics that had begun using zidovudine (ZDV or AZT) for preventing mother-to-child transmission of HIV (PMTCT) in 1999, restricted the use of nevirapine donated free of charge by Boehringer Ingelheim in 2000, obstructed the acquisition of grants for AIDS treatment from the Global Fund in 2002, and generally delayed implementing a national ARV treatment program until 2004. By considering the decreasing costs of ARVs, the increasing availability of international resources to fight AIDS, and comparing South Africa to neighboring Botswana and Namibia, we conservatively estimated the number of AIDS patients that could have received ARVs for treatment or PMTCT. Factoring in the efficacy of ARVs, we concluded that from 2000 to 2005 at least 330,000 South Africans died prematurely and 35,000 babies were infected with HIV as a result of Mbeki's policies. Independently and using a different model, Nattrass arrived at similar estimates .
Duesberg and colleagues published a critique of the study in the Journal Medical Hypotheses which was subsequently retracted by the publisher pending an investigation of the quality and global health implications of the paper . Peter Duesberg is the most well known AIDS denialist who was part of President Mbeki's commission tasked to determine whether HIV causes AIDS in 2000, and he has recently received attention from a mainstream magazine  and a whistleblower award for his AIDS denialist writings.1 Consistent with earlier writings, Duesberg and colleagues:
1) Deny that HIV causes AIDS; that instead, it is a harmless passenger virus;
2) Deny that ARV drugs are useful, and therefore Mbeki's decisions could not have harmed anyone;
3) Deny that hundreds of thousands of South Africans have died from AIDS, and thus it does not make sense to attribute 330,000 deaths to Mbeki .
We choose to respond to the issues raised above for two reasons: first, some readers may be hoodwinked by Duesberg's dishonest arguments and think that there is a genuine debate in light of the surge in denialist coverage, and second, to emphasize the grave implications of AIDS denialism for public health practice.
Does HIV Cause AIDS?
Duesberg has been denying that HIV causes AIDS for more than 20 years .President Mbeki joined the debate in 1999 initially by questioning whether AZT was safe for use by pregnant women , and then joined the denialists by questioning whether HIV was the "real" cause of AIDS as a way of broadening the debate from the usefulness of AZT to the usefulness of all antiretroviral drugs in fighting the AIDS epidemic, since they all target HIV [7, 8]. He then appointed Duesberg and others to a commission to examine whether HIV causes AIDS . Whether HIV causes AIDS is therefore at the very center of the policies implemented by Mbeki.
The evidence that HIV causes AIDS has been available for over 20 years.2 Careful epidemiological studies showing that individuals with a new, severe immunosuppressive disease clustered among homosexual men [11-13], intravenous drug users , female sexual contacts of drug users , hemophiliacs [16, 17], other recipients of blood transfusion products [18, 19], and newborn babies  suggested that the cause was an infectious agent transmitted by body fluids . Early suggestions that illicit drugs or immune reactions to sperm were the cause [22, 23] could not explain all the patient groups affected by the immunosuppression. Serological studies then suggested that the causative agent was likely to be a retrovirus [24, 25], and this was confirmed by isolation and culture of the retrovirus from infected patients [26, 27]. Diagnostic assays were developed and much larger studies were then possible to identify HIV-infected persons using the presence of HIV antibodies, antigens, viral nucleic acids and virus, and to compare them to uninfected persons in longitudinal studies to learn the virology, immunology, pathology, and clinical and population features of the disease.
HIV meets several standards of epidemiologic causality . HIV has satisfied Koch's postulates, the traditional standard of infectious disease causation. To satisfy Koch's postulates, one has to isolate the infectious agent from diseased animals, culture it in the lab, inoculate the agent into healthy animals which then develop disease, and reisolate the same infectious agent . The difficulty in fulfilling the postulates was because HIV does not cause disease in animals other than humans  and it is unethical to infect healthy persons with HIV just to satisfy Koch's guidelines. However, the postulates were satisfied when the HIV virus was isolated from AIDS patients, cultured in vitro, and upon accidental inoculation into previously uninfected lab workers who subsequently developed AIDS, the exact laboratory HIV clone was re-isolated from the patients [31, 32]. Using a causal model developed for chronic disease, HIV satisfies all of Sir Bradford Hill's guidelines for assessing causality:  numerous studies comparing infected and non-infected persons have shown that AIDS develops only in those infected with HIV (very strong association, consistency and specificity) [34, 35]; follow-up cohorts have shown that the time relationship is that HIV infection always precedes AIDS (temporality) [36, 37]; higher level of virus as measured by viral load correlates with and predicts severity of disease (biological gradient) [38, 39] ;treatment that suppresses virus leads to clinical improvement (experiment) [40, 41]; there is an almost unique pathophysiological mechanism of how HIV leads to AIDS through the loss of CD4 lymphocytes (specificity and plausibility) [42, 43]; and numerous studies on HIV-1, HIV-2, SIV, SHIV and other viruses satisfy the coherence and analogy guidelines .
The above data have been presented and debated over the last 25 years. Duesberg's response has been to ignore or deny the data that does not support his position, and to cherry-pick statements from studies and present them out of context to suggest that the evidence for HIV causation is unconvincing. His early argument was that HIV had not satisfied Koch's postulates for infectious disease causation, and he also indicated several aspects of the pathogenesis that were not understood then . However, when lab workers accidentally inoculated themselves with the virus and satisfied the postulates, Duesberg refused to accept the data  and now conveniently does not discuss the postulates. Similarly, early on, Duesberg agreed that hemophiliacs were the best group to test whether HIV causes AIDS because most of them did not have the drug use exposures that Duesberg considered causes, and both HIV-positive and HIV-negative hemophiliacs had received transfusions, hence foreign-protein contaminants . When Darby and colleagues published mortality data in the complete UK population of 6,278 hemophiliacs showing that those with HIV had 10 times the mortality of those without, with 85% of the deaths attributable to HIV , journal editors who had hoped this was an honest debate asked whether Duesberg was going to concede defeat . He did not. He just moved the goal posts and suggested that AZT was the cause of AIDS ; the approach that he had agreed to of using "hemophilia as the best test" was no longer relevant. While the other points raised by Duesberg pertain to pathogenesis and not causation , most of the mechanisms are understood today. Thus, molecular techniques were developed and it became possible to isolate and quantify free virus in plasma [49, 50]; the dynamics between virus and CD4 cells and how this relates to disease progression were unraveled [39, 51]; highly effective medications that work by suppressing virus were developed and are now in widespread use [52, 53]; and opportunistic infections similar to those in the US were reported from Africa and Asia [54, 55]. Duesberg has moved on from those arguments. One of his remaining arguments is that if there is no AIDS vaccine, which some predicted we would have soon after the discovery of HIV in 1984, then HIV does not cause AIDS . The same reasoning could of course be used to argue that Plasmodium falciparum does not cause malaria, as there is no malaria vaccine.
What therefore causes AIDS, in Duesberg's opinion? His answers are inconsistent and contradictory. On the one hand, he seems to argue that AIDS (the syndrome) does not exist at all, labeling it "a fabricated epidemic ," since all opportunistic infections that define it already existed before AIDS . On the other hand, he also concedes that AIDS exists and offers causes , and seems unbothered by posing mutually exclusive arguments at the same time. In his earlier writings, he accepted that there is statistical association between HIV and AIDS (although he argued this was insufficient for causation)  and even considered the HIV-antibody test as useful surrogate to identify patients at risk of AIDS;  today, he denies that and argues that HIV is a passenger virus with no relationship whatsoever to AIDS . In the same contradictory way, Duesberg has argued that HIV is not the cause of AIDS because "in most individuals suffering from AIDS, no virus particles can be found anywhere in the body" ; yet at about the same time that he published this, he was involved in a disagreement with other AIDS denialists who had challenged the very existence of HIV where he defended that "HIV has been isolated by the most rigorous method science has to offer ."
Duesberg clings to the early argument that AIDS is caused by use of recreational drugs , but as explained above, this hypothesis was discarded when AIDS was seen in patients that had never used drugs including hemophiliacs, transfusion recipients, babies, and some African populations. For hemophiliacs, he suggests that "foreign-protein contamination" through blood products is the cause , yet does not explain how AIDS from transfusion has virtually been eliminated just by incorporating the HIV test into blood screening . The strangest cause he proposes is that AIDS is caused by AZT and other antiretroviral drugs , even though AZT was only used after 1987 and used primarily on persons already with AIDS rather than healthy persons. To this, Duesberg replies that there was no AIDS in persons other than illicit drug users before 1987.3 In babies, he moves from arguing that there is no AIDS in babies  and HIV cannot cause AIDS in babies (as it would otherwise kill itself together with its host) , to arguing that there is immunosuppression in babies but it is different and characterized by B cell deficiency , then that babies with AIDS are born to drug-addicted mothers . Nevertheless, there are data showing that pediatric AIDS is real and has killed over 250,000 children per year since 1998 , that it has the same immunological profile of CD4 deficiency as in adults [20, 73], and that HIV-negative babies born to drug addicts do not get AIDS [74, 75]. What of Africa, the worst affected continent, which has comparatively much less recreational drug use and until this decade did not have ARVs in large supply?  Duesberg suggests that the cause is "protein malnutrition, poor sanitation and subsequent parasitic infections ." However, AIDS has affected the well-off and over-nourished Africans, not just the undernourished , and this raises the question why the same explanation does not apply to other less-developed countries outside Africa that do not have as much AIDS , or earlier time periods when poverty and the attendant sanitation and nutritional problems were not any less in Africa (and other places). Moreover, AIDS is a particular type of immunosuppression with selective depletion of CD4 lymphocytes , and neither homosexuality , illicit drugs , ARVs , blood transfusions , malnutrition , nor living in Africa  cause this.
In short, any explanation other than that HIV causes AIDS seems better to Duesberg-he therefore moves from the claim that AIDS does not exist to a multiplicity of causes even if it means creating a different cause for different geographies, different time periods, and different demographic groups, and without producing a shred of evidence [28, 86, 87]. This is what is called denialism-"the rejection of objective reality to sustain a flawed, hurtful, and ultimately dangerous belief system" .
Are ARVs Effective in PMTCT and AIDS Treatment?
Estimating the human cost of not using ARVs in South Africa rests on the efficacy of ARVs when used for PMTCT and AIDS treatment. Mbeki entered the AIDS debate by questioning whether AZT was safe and useful for pregnant women , and Duesberg argues this position for all ARVs .
There are two observations to make from the way Duesberg argues the case. First, he discusses how and when AZT was first discovered and its mechanism of action inhibiting DNA synthesis, then cites some anecdotal cases, and concludes that all ARV drugs are toxic and not useful . Mechanisms of action are interesting to scientists but this is the wrong evidence to evaluate for efficacy . If one were to ask how best to treat hypertension, for example, the answer does not come from the interesting neurobiology of the hypothalamic blood pressure control centers, the crystal structure of angiotensin, or how Captopril was initially discovered. The relevant standard of proof, the gold standard, is the clinical trial where the drug in question is compared to placebo (or alternative treatments) in a randomized controlled manner and a priori chosen outcomes analyzed . This is why the US Food and Drug Administration requires clinical trial data before licensing any new drug.4 By choosing mechanisms of action, Duesberg is using inappropriate evidence, but purposefully so as to obfuscate the argument.
After deciding on the standard of proof-which is the clinical trial-the second step is to agree on how to assess the results from many such trials done in different countries and populations. Duesberg's method is narration, where he ignores the data he dislikes, cherry-picks the statements he likes from different publications, and selectively interprets them to support his position, disregarding even the main conclusions of the studies .5 Narrative reviews, while very common and perhaps relatively less demanding to perform, have the drawback that it may be unclear whether all the relevant evidence has been used or the reviewer selected studies that support a desired conclusion, and whether the apportionment of weight to studies was based on objective criteria such as sample size . The relevant standard here is a meta-analysis, that is, a systematic review with statistical synthesis of all relevant available data . When a meta-analysis is performed well, there is an a priori protocol specifying the question asked, the databases to be searched for publications, justifiable inclusion and exclusion criteria, the data to be extracted from studies, the quality assessment score to be used for each study, and models for statistical analysis . For Duesberg to convince impartial readers that ARVs are useless or toxic when used for PMTCT and AIDS treatment, he has to produce a properly conducted meta-analysis (the objective standard for summarizing evidence) of clinical trials (the highest grade of evidence for assessing efficacy) where the drugs were used. Obviously, he cannot produce this because numerous clinical trials and meta-analyses have already been conducted and the evidence, as shown below, is unanimous in that the benefits of ARVs outweigh the side effects.
To quote an example from our work, we recently published "Efficacy of Antiretroviral Drugs in Reducing Mother-to-Child Transmission of HIV in Africa: A Meta-Analysis of Published Clinical Trials ." The question asked was how efficacious have ARVs been in PMTCT in Africa, first to generate an efficacy estimate directly relevant for policies on the continent that is worst affected by HIV/AIDS, and second, to pre-empt the debate on what is feasible in Africa (due to drug compliance, C-section rates, breastfeeding, late antenatal presentation, etc.) by considering only studies performed in Africa. The key result of this meta-analysis is that ARVs reduce mother-to-child transmission of HIV from 21% (combined placebo estimate) to 10.6% (combined ARVs estimate) at 4-6 weeks after birth. From all the studies that reported toxicity, ARV regimens for PMTCT are well tolerated by both the mothers and babies. The quantity of this evidence is 10 clinical trials with a combined sample size of over 7,000 HIV-infected pregnant women, and over 800 transmission endpoints. The type of evidence is high grade, that is, randomized clinical trials rather than observational, cross-sectional, or case reports . The Jadad  quality of the individual clinical trials is high. The efficacy of using ARVs versus placebo is 50%. Using the US Institute of Medicine categories of certainty in assessing evidence [97, 98], the evidence establishes that ARVs are efficacious in reducing MTCT in Africa, and the evidence favors rejection of the hypothesis that ARVs, in the doses used, are toxic to the mothers or babies. Example diagrams reproduced from the paper are given in Fig. 1, which shows the one-study-removed meta-analysis indicating that the results are not disproportionately driven by any one study, and Fig. 2, which shows the cumulative meta-analysis by publication date showing when the evidence was available; other figures and data (not reproduced here) show that we tested for publication bias, tested the effect of different statistical models of summarization, and assessed heterogeneity of effect estimates. Contrary to what Duesberg suggests, there are unanimous data (all trials conducted in Africa published by December 2006) to demonstrate the usefulness of ARVs in PMTCT in Africa, and other groups have arrived at the same conclusions for ARV use in PMTCT generally [99, 100].
Likewise, extensive clinical trials data demonstrate the efficacy of ARV drug combinations in treating AIDS [40, 53, 101, 102]. The results from use of drugs in combination were so dramatic that the term "HAART," for Highly Active Anti-Retroviral Therapy, was coined. Many systematic reviews have been conducted and updated by the Cochrane Collaboration and other groups, and the data are unanimous regarding efficacy [103-108]. In addition, data are now available from the use of ARVs at the program level in African countries and these support the efficacy observed in clinical trials [109-113]. Several studies have systematically reviewed the data just for developing countries and Africa [114, 115], and others compared low- and high-income countries . In short, if Duesberg wishes to demonstrate that certain ARVs are no better or worse than placebo or other treatments, he has to conduct a meta-analysis that considers all available evidence, rather than his approach of discussing the molecular biology of DNA chain termination and somehow inferring that ARVs are not beneficial.
Moreover, for Duesberg to totally discredit the paper on the human cost of not using ARVs, he has to argue that all ARVs are totally ineffective when used for AIDS treatment and PMTCT because if some ARVs are even marginally effective, then it means that some South Africans could have benefited, however, small the benefit, had Mbeki not obstructed drug use.
Population Growth and AIDS Deaths
The third of Duesberg and colleagues' arguments is that there is no evidence of large-scale deaths in South Africa, and therefore whatever policies Mbeki implemented, they did not lead to deaths. To support this, they present two arguments: one, that the population of South Africa increased over the last 30 years, and two, the statistics of reported AIDS deaths in South Africa .
Regarding the first argument, it is true that the population of South Africa increased over the last 30 years.6 The population in a country is determined by the balance between the number of live births, the total number of deaths, and net migration . Without doing an analysis of the above determinants, it not possible to use such aggregate population trend data to infer that the number of AIDS deaths was small. If this reasoning is sound, then it should be applicable to other countries and diseases as well. Is it logical to infer that AIDS deaths are few in any country that has increased its population over the last three decades? Similarly, is it logical to infer that there has been no increase in the number of persons dying of cardiovascular diseases and cancer or that the absolute numbers of death from these diseases are small in the US, whose population has increased over the last half century? This argument does not support Duesberg's assertions at all.
The second part of the argument quotes Statistics South Africa, which recorded an average of 12,000 deaths per year in South Africa between 1997 and 2006 . The shortfall is that these data are "Findings from Death Notification ." First, as explained by surveillance experts, "In resource-poor countries with underdeveloped health infrastructures, reports of AIDS or HIV cases are usually not complete enough to be considered reliable measures of the scope of the epidemic ". This simply means that the death notification system in South Africa had/has much underreporting. Indeed, the "former so-called independent homelands of Transkei, Boputhatswana, Venda and Ciskei (TBVC) were not included in the reporting system until 1994" when the reporting system began centralization, and a new death certificate was introduced in 1998 to improve reporting . The second shortfall is that of misclassification of deaths. AIDS patients die of the resulting opportunistic infections and cancers, and these immediate causes of death are often recorded without noting the underlying acquired immunodeficiency. According to the Medical Research Council (SA), up to 61% of HIV deaths are misclassified and the majority of them are recorded as tuberculosis and lower respiratory tract infections, which become the leading causes of death .7 It is apparent that Duesberg selected highly deficient statistics.
Yet, there are several sources for AIDS mortality data. UNAIDS has a Reference Group on Estimates, Modelling and Projections 8 that works collaboratively with officials in the partner countries and academia, and it regularly reviews information needs and new data, updates the methodology and assumptions used in estimating AIDS statistics [122, 123], and produces estimates that include the level of uncertainty around each estimate [124, 125]. The resulting statistics represent a fair picture of the epidemic [126, 127]. For South Africa, the reported deaths from AIDS range from 270,000 in year 2000 to 320,000 in 2005.9 These estimates are consistent with data from South Africa's Health Systems Trust,10 the Actuarial Society of South Africa AIDS and Demographic Model 2003 , and the South African Medical Research Council Burden of Disease estimates .
This high mortality is consistent with the demographic changes that have happened in South Africa. From 1995 to 2005, the absolute number of deaths rose from 321,000 per year to 780,000, a doubling of the crude death rate (per 1,000 population) from 8 to 16. The life expectancy at birth dropped from 64 years to 49 years, and the under-five mortality (per 1,000 population) rose from 57 to 69. 6 Community studies show that the increased mortality comes from the increasing death of young adults, that up to 50% of the deaths are attributable to HIV, and mortality in children born to HIV-positive mothers is two to five times that of children born to uninfected mothers11 [129, 130]. Indeed, demographers have shown that the population would have grown significantly more if there were no AIDS [119, 131].
The extreme end of the argument is to suppose that Duesberg's low statistics of about 12,000 AIDS deaths per year were correct-that would translate to a total of 72,000 deaths from 2000 to 2005, and this would still enable a calculation of the number of persons that could have been treated using ARVs had Mbeki not obstructed, 24,000 lives if we assume a third of them would have been treated. This is not a small number of people to let die because of AIDS denialism.
There are several implications to draw from this work. First is the translation of denialism into public health practice. One of Duesberg's first papers questioning whether HIV causes AIDS was published in the prestigious journal Science in 1988 . Some researchers initially took this as a genuine scientific debate but as Koch's postulates were fulfilled, randomized controlled trials demonstrated the high efficacy of ARV, there was much success in PMTCT, and studies elucidated the dynamics between virus and CD4 cells, Duesberg maintained his arguments and it became clearer that he was not just a dissident scientist but a denialist. When Mbeki took up the denialists' position in 2000, there was international outcry . Not only was he lending his ear to discredited scientists, but AIDS denialism was crossing into national health policy through a head of government. Participants at the 2000 International AIDS Conference in Durban (SA), news outlets, scientific journals, and the public were outraged and some went as far as saying that South Africa was tripping into anarchy, descending into an abyss [132, 133]. South Africa did descend into that abyss. Mbeki withdrew support from clinics that had started using ARVs, restricted use of donated ARVs, obstructed Global Fund grants, and generally delayed implementing a national ARV program. Two independent studies have estimated that Mbeki's policies led to at least 330,000 premature deaths [1, 2]. When AIDS denialism infiltrates public health practice, the consequences are tragic.
The second implication follows directly from the first and concerns accountability. Mbeki implemented negligent policies that led to the premature death of hundreds of thousands. His reasons, as stated by himself and health minister Tshabalala-Msimang, were that he questioned whether HIV causes AIDS and whether ARVs are safe, and neither ever publicly backed down from this thinking. The science behind Mbeki was Duesberg and other denialists. Duesberg is still arguing for AIDS denialism and defending Mbeki and the policies that led to more than 330,000 deaths. By any reasonable standard, this requires some form of accountability. Seth Kalichman has likened the AIDS denialists to the Holocaust deniers  and Edwin Cameron likened letting AIDS patients die without medications to those who silently enabled the evils of Nazi Germany and apartheid South Africa to go unchecked . John Moore and Nathan Geffen have called for AIDS denialists to be put on trial [136, 137] and Mark Wainberg has argued that denialists should be charged with public endangerment and "people like Peter Duesberg belong in jail ." Zachie Achmat has called for a commission of enquiry such as the Truth and Reconciliation Commission that was tasked with handling the apartheid era crimes.12 For how are South Africans ever going to trust their health system again? How can a modern government be penetrated by denialists to the extent of implementing policies that kill hundreds of thousands? William Makgoba suggested that impeding AIDS treatment was collaborating in committing genocide,13 and Wycliffe Muga has asked whether Mbeki's killing of 330,000 by obstructing life-saving medications is much different from Sudan's President al Bashir's killing a similar number in Darfur through obstructing humanitarian aid and militias . Is this not a crime against humanity? Does the International Criminal Court not have a role, for it was established to handle those cases where national courts may be unable or unwilling to prosecute?14 Whatever the most appropriate avenue is, what seems apparent is the need for accountability.
The third implication somewhat generalizes the argument. AIDS denialists are dangerous to the general population; many have been persuaded into risky behaviors, ineffective alternative remedies, and other harmful actions, although there is no easy way of evaluating how many . Similarly, denialists can impact public or national health policy and South Africa is one extremely tragic case. However, denialists seem ineffective against physicians as a group. The reason is that if an AIDS patient goes to a physician, and the physician decides not to treat, the physician is held for malpractice. The medical profession is practiced only by those who have earned defined credentials. The standards of practice are generally known and deviant practitioners are disciplined by the medical societies and deregistered by states. Moreover, the law of torts offers patients a private right of redress against negligent doctors. The above seem absent in public and global health. The practitioners are ill defined and there are no laws restricting practice to persons with specified credentials. The concept of standards of practice is not well developed, and there are no bodies tasked with self-regulation and discipline. The concept of public health malpractice has not yet been developed . Thus, at a general level, AIDS denialism in South Africa has also exposed the deficiencies of public health practice-it is open to unqualified practitioners, negligent policies go unchecked, and the consequences are tragic. How to rectify this is beyond the scope of this paper; here it suffices to point out the deficiencies of public health in terms of standards, practitioners, and accountability, as exposed by the South Africa example.
Last, Duesberg was able to publish his paper (which was later withdrawn) only because it was not reviewed by peers knowledgeable on the subject. Denialist writings require close scrutiny and peer review before being published in scientific journals, especially when they have the potential to impact public health practice.
When AIDS denialism enters public health practice, the consequences are tragic. The implications start in honest science but extend to the need for accountability and, perhaps, public health reform.
Acknowledgments We wish to acknowledge Professor George Seage who critiqued and helped refine the arguments on causation of AIDS in an earlier version of the argument.
Conflict of Interest We declare that we have no conflict of interest.