Archive
Conflicts of interest in HIV medicine: The 2012 revised DHHS HIV treatment guidelines and what’s wrong with expert opinion
The most recent revision of the DHHS guidelines on the use of antiretrovirals in HIV infected adults and adolescents now recommends starting therapy at a CD4 lymphocyte count greater than 500/ mm3.,
For those with greater than 500 CD4 lymphocytes the recommendation is only supported by expert opinion – the opinions of the experts on the DHHS panel. Almost all of the non-governmental researchers on the panel have financial arrangements with entities that stand to gain from the decisions they make. There are plenty of other experts who are not members of the DHHS panel who are not so certain that starting treatment above 500 CD4 lymphocytes will confer a net benefit to the patient..
This particular recommendation is unlike those made for individuals with lower CD4 numbers where more reliable evidence from clinical trials clearly demonstrates a benefit to the patient
Evidence based medicine has brought us a long way from the days when clinical decisions were based on authority and tradition (“expert opinion”); it attempts to use the best available evidence on which to base clinical recommendations. The term “best available evidence “means that not all types of evidence are of equal quality. There are several systems that grade the relative strengths of evidence derived from different sources. All agree that evidence provided by randomized controlled clinical trials is of the highest quality and therefore the most reliable. Applied to HIV medicine, a strong recommendation that antiviral treatment be initiated at 350 or fewer CD4s can be confidently made because the evidence of substantial benefit is derived from a randomized controlled clinical trial.
At the other end of the scale rating the quality of evidence, is evidence based on “expert opinion”. This may not even be a marginal improvement on the bad old days when the doctor knew best; when there was no need to justify a recommendation other than by the authority of the doctor or by tradition.
The rating of the recommendation that people with more than 500 CD4 lymphocytes start treatment, according to the system used by DHHS is B III. It’s a moderate recommendation supported only by the opinion of experts.
But when expert opinion is the basis for a recommendation, this does not even mean that the opinion represents a consensus of all experts. It only represents the opinion of those experts chosen by the organization making the recommendation.
Making a recommendation based solely on expert opinion is particularly troublesome when the means exist to obtain evidence of the highest quality. The START trial that directly addresses the question of when it’s best to begin antiviral treatment is enrolling, and one must wonder why the panel did not defer making a recommendation concerning individuals with greater than 500 CD4 lymphocytes until the trial results become available. This is even more puzzling as individuals who have waited to start at CD4 numbers between 350 to 500 have in general done very well, so waiting to make a recommendation for some years until the START trial results are available seems to be a much more reasonable and prudent option than jumping the gun and making a recommendation based on evidence of the weakest quality.
Bur when we come to look at the associations of the experts on the DHHS panel, a recommendation based on expert opinion is even more problematic. We note that almost all of the non-governmental researchers have financial arrangements with entities that can benefit from the decisions they make. Some of these arrangements are quite extensive.
A conflict of interest becomes particularly troublesome when it’s only the opinion of the expert that supports a recommendation. Since people with greater than 500 CD4 lymphocytes represent a huge proportion of the HIV infected population, treating them will have an impact on expanding the market for antiviral drugs. With greater efforts to encourage testing, greater numbers of individuals with higher CD4 numbers will be identified, and now recommended to receive lifelong treatment with expensive and potentially toxic drugs whose benefit has not yet been proven to outweigh their harms.
The conflicts of interest of panel members are duly noted in the DHHS financial disclosures.
Early AIDS activists performed a great service for all individuals who must deal with illness, in asserting their right to make informed decisions concerning their care, and that the decisions are made free from coercion. Withholding information and supplying misinformation are forms of coercion.
Although the guidelines ask physicians to inform patients with high CD4 numbers that evidence for benefit is not conclusive I think it’s safe to conclude that individuals with greater than 500 CD4s will not always, and may only rarely informed be informed of this important caveat. As to informing patients of the conflicts of interest noted above, this isn’t even a consideration. They are also unlikely to be told that the recommendation that they start treatment is based on the opinion of certain experts only, and that there are other experts with a different opinion. In fact, the DHHS guidelines may be the only ones in the world to make this recommendation.
Undoubtedly the DHHS panel members believe that people with higher CD4 numbers will receive a net benefit from treatment. But the recommendations would have greater authority if the non-governmental researchers on the panel were better balanced with respect to members who had no financial arrangements with entities that stand to benefit from their decisions; in fact many would agree that such conflicts of interest should be a disqualification for panel membership.
The recommendations also refer to the prevention benefit of treatment. The greatest prevention benefit will result from the treatment of individuals with lower CD4 numbers who will have the highest viral loads. These individuals need treatment. On this point there is no doubt or debate. For those with higher CD 4 numbers, not known at this time to benefit from treatment, the prevention benefit is likely to be much lower as their viral loads will also, on average be much lower than those with more advanced HIV disease.
Providing treatment to everybody who needs it to stay alive should surely be our first priority. It is here that treatment will also have its greatest prevention benefit.
Conflicts of interest are of course common among those making treatment recommendations. However HIV medicine seems to be unique in that these conflicts of interest, which may be among the most egregious, seem to go almost completely unnoticed. In every other field of clinical medicine they occasion extensive discussion. The apparent indifference to conflict of interest issues and the influence drug marketing practices in HIV medicine is unfortunate, as precedents in that field may go unnoticed but will have implications for other fields of clinical medicine. The rapid approval of zidovudine by the FDA in 1987 may be such an example.
Two years ago in a tribute to Michael Callen I responded to similar recommendations to treat all HIV infected individuals irrespective of CD4 numbers.
I cannot express my reservations more clearly than with the words I used then:
I miss Michael Callen. He was my patient when AIDS began, but soon became my collaborator and friend.
For a time, Michael and Richard Berkowitz, another patient collaborator, were able to work out of an office adjoining my practice on W 12th street in New York City. It was in this setting that Michael and Richard learned about the medical aspects of this new disease and participated in the creation of some of the earliest organized community responses to the epidemic.
Michael and Richard helped in the formation of the AIDS Medical Foundation; they wrote the very first publication to recommend condom use by gay men. Michael played a role in the first attempt to protect the confidentiality of people with AIDS, and he helped to create both the Community Research Initiative and the PWA Health group.
A thread running through all of these endeavours is the notion of self empowerment. This extends beyond the belief that individuals who are fighting a disease should actively participate with their doctors in making decisions about the care they receive. Empowerment also means the inclusion of affected individuals at all levels of the response to the disease, from research to the provision of services.
The Community Research Initiative was sponsored by the PWA Coalition of which Michael was President. This is the very embodiment of self empowerment. It is people with a disease sponsoring research into that disease themselves and not waiting for some benevolent institution to come to the rescue.
Michael understood that his interests and priorities as a person living with AIDS might sometimes be at odds with those of some scientists conducting research into this new disease. He knew very well that he was living in a world that was still capable of cruel and discriminatory behavior towards him. Who better to protect the interests of those who had the most to lose than people living with AIDS themselves?
Self empowerment found expression in the Denver Principles. Michael and Richard were both signatories to this historic document. Michael played a major role in crafting the words of the Denver Principles.
Almost thirty years later these Principles remain as important as when they were first articulated.
One of the Denver principles asserts the right to obtain full explanations of all medical procedures and risks.
I wish Michael Callen were here today to bring attention to the violation of this right.
This is happening with little protest in places like San Francisco where antiviral medications are now recommended for healthier HIV positive individuals for whom the benefits of treatment have not been shown to outweigh the risks.
As always, you can’t beat the truth, and the truth is that for people with more than 350 CD4 lymphocytes, the best time to start treatment is not known. This may seem surprising as potent antiretroviral drugs have been available for fifteen years.
We have not yet done the kind of study that would most reliably provide the information those HIV positive individuals with higher CD4 numbers and their doctors need to make the best decisions about when to start treatment.
With information provided by a properly designed and conducted prospective randomized trial, we could know with confidence when in the course of HIV infection the benefits of treatment absolutely outweigh the risks.
Some feel that a decision can be made with less reliable information. But surely all would agree that a decision to start treatment or to defer it must always be an informed one voluntarily made by the individual considering treatment.
It is here that the principle asserting the right to a full explanation of the risks of medical interventions is being violated.
The San Francisco Department of Public Health in advising all HIV infected individuals to receive treatment is in effect telling them that at all stages of HIV disease the benefits of treatment outweigh the risks. This may be so, but apart from those with 350 or fewer CD4 lymphocytes, we just do not have the most reliable evidence to support this contention.
People with higher CD4 numbers have the right to know not only what evidence there is that immediate treatment will have a net benefit compared to deferring it, but also the quality of that evidence. They surely should also be made aware that experts hold differing opinions on whether treatment should begin immediately or be deferred.
A physician in San Francisco who recommended that all HIV infected individuals should start treatment immediately was reported to have said:
“If I’m wrong, we’ll start people [on treatment] a couple years earlier than we otherwise would. But if I’m right and we don’t start early, there’s no going back,”
Others who are concerned about drug side effects might feel that more may be at stake for HIV positive individuals with higher CD4 numbers. This also includes the possibility that fewer options may be available when treatment is definitely known to be needed.
This doctor is also reported to have said:
“The old paradigm was that drugs are toxic so we should wait as long as possible. The new paradigm is that while today’s drugs are not totally benign, they are less toxic than the virus.”
“The” paradigm? Is it not misleading to give an impression that his views on drug toxicities represent a consensus?
How on earth can the longer term toxicities of the newer drugs be known?
Just a few days ago it was reported that AZT and 3TC based therapies produced a metabolic abnormality called hyperhomocysteinemia. This is a condition associated with vascular abnormalities including a greatly higher risk of heart attacks. We have been prescribing AZT and 3TC for about twenty years, so what information does the San Francisco doctor have that gives him such confidence that the drugs in use for only a few years are less toxic?
Empowerment means that HIV positive individuals make their own decisions to start or to defer treatment. They have the right to clear and honest information to enable them to make this decision. Those with higher CD4 counts have the right to know that there still is uncertainty about when it is best to start treatment.
The views of the San Francisco Department of Public health and those who share them are just opinions; healthier HIV positive individuals should also know that these opinions are not held by all experts. Respect for the autonomy of healthier HIV positive individuals requires that opposing views on when it’s best to start treatment be presented together with the evidence supporting these views, so those who have most at stake can decide for themselves.
There will continue to be opposing views on when it’s best to start antiviral therapy as long as the question has not been put to the test.
The best way to resolve uncertainty in clinical medicine is by conducting prospective randomized trials. A properly designed and conducted trial could reliably and safely answer the question of whether, on average, immediate or deferred treatment is better or worse or makes no difference.
HIV positive individuals deserve the most reliable information to inform them in making treatment decisions. The START trial is a randomized prospective study that directly asks the question about the best time to start antiviral medications. We could really finally know what’s best, and no longer rely on opinions based on data of inferior quality.
Is an immediate or deferred initiation of treatment better or worse, or does it make no difference? If knowledge is power a demand to complete the START trial is the embodiment of the self empowerment of HIV positive individuals. The very antithesis of self empowerment is to allow researchers to persuade us with evidence of inferior quality, such as their personal opinions, presented as if there were a consensus, or with the results of embarrassingly uninterpretable studies such as NA-ACCORD so often used to justify earlier starts to treatment. NA-ACCORD was not a prospective randomized trial. It was a review of a large number of medical records. Such retrospective observational studies are beset with interpretative difficulties because subjects are not randomly assigned to receive one or another kind of intervention, in this case, to start treatment early or to defer it. We don’t know why a particular course of action was chosen. The reasons why decisions were made to start treatment early or to defer it may have determined the outcome rather than the time treatment was started.
In situations where prospective randomized trials cannot be conducted for whatever reasons, then we have to do the best with data of inferior quality. But fortunately this is not the case with HIV infection.
I miss Michael Callen. He would have reminded us that HIV positive individuals must demand that the best evidence be obtained to inform their treatment choices.
.
Individualization of HIV therapy
Why treatment of HIV infection must be individualized.
HIV disease is usually a progressive disease. That is, it has a starting point; the time of infection. The disease then progresses, and without treatment will generally end fatally. There are some very fortunate HIV infected individuals who are able to control viral replication and remain disease free. But for most, HIV disease does progress. But, for each individual, the rate at which it progresses varies widely. Disease progression is reflected in the fall in the numbers of CD4 lymphocytes.
So any single CD4 count measurement is really a point on a descending curve, one that does not necessarily proceed in a straight line, and falls at widely differing rates in different individuals.
Recommendations for the treatment of HIV infected individuals are issued periodically by DHHS and bodies such as the International AIDS Society. These recommendations, particularly those concerning when to start antiviral treatment, have always included a particular CD4 count as a signal to start or to consider starting antiviral treatment.
All individuals with a CD4 count of less than 200 should be on therapy. They are in great danger of acquiring a possibly fatal opportunistic infection and evidence derived from clinical studies makes it absolutely clear that antiretroviral treatment is life saving.
But what about people with higher CD4 counts? Here there is uncertainty about when in the course of HIV infection it is best to start treatment. Of course, if the drugs were completely harmless (including cost) it might be less important to have an answer to this question. However the drugs can have significant adverse effects, some of which only become evident after years of use. For people with fewer than 200 CD4 lymphocytes, the benefit of antiviral treatment overwhelmingly outweighs the risks.
For others, a very mixed group, with CD4 cells anywhere from 200 to over 1000, and each with a different rate of disease progression, we cannot, with any security, make a “one size fits all” recommendation as to when it is best to start treatment.
The best way to resolve clinical uncertainty remains randomized prospective clinical trials. By now we might already have obtained reliable evidence as to whether, on average, it is best for infected individuals with more than 200 CD4 lymphocytes, and who have no symptoms, to start antiviral treatment immediately, or to defer it. (A suggestion made in 1997 when the first guidelines were issued: http://aidsperspective.net/articles/guidelines1.pdf )
The current recommendations, regarding people with greater than 200 CD4 lymphocytes, and who are without symptoms, propose a CD4 count of 350 as a point to start treatment ( many believe this number should be 500). This recommendation is made for all individuals – it is a one size fits all approach[1]. This kind of approach is appropriate for some aspects of treatment; for others it is very wrong[2].
Perhaps the most important example of a recommendation, where its application across the board is problematic, is that which deals with the time when antiretroviral treatment should be started in individuals with greater than 200 CD4 lymphocytes. This recommendation specifies a specific CD4 count at which to start. As noted, for individuals with a CD4 count below 200, there is no doubt that they will benefit from therapy. For others who have no symptoms, specifying a CD4 count for all is mistaken. It is here that individualization is necessary.
The reason is that no two HIV infected people are the same with respect to the rate of disease progression. During the early years of the epidemic, before antiretroviral treatment was introduced, we soon noted that the CD4 count declined at different rates in different people, and not necessarily in a straight line. As noted, at one extreme, there were the few fortunate individuals in whom there seemed to be no disease progression, at the other there were the few people whose CD4 cells fell very rapidly after infection, and who did not survive for more than 2-3 years, but most fitted somewhere between these extremes .
To illustrate this I have considered four possible situations. This is a picture of the possible rates of CD4 decline in four different individuals. . It is true that these pictures are constructs, but they do accurately reflect the observed variability in disease progression; real examples showing this variability would be easily found in my medical records, and of course in those of other physicians during the period between 1981 and about 1993.
The dip in CD4 cells following infection is usually seen when there is an opportunity to observe this. CD4 cells then rebound to a level called the set point, which will be different in relation to the pre infection level in different people. From then on it declines, but at a very variable rate, and can remain steady for varying periods before declining, again at varying rates.
Look at where three of them (A ,B and C) reach a count of 450 CD4 lymphocytes; A (an unusual rapid progressor) gets there in about one year, B in about 3 years, C in 7 years, and D, who is a fortunate non progressor is nowhere close after 18 years.
The arguments for starting early are not only to forestall reaching the dangerous level of 200 CD4 lymphocytes. The continuous deterioration of the immune system and diminished chances of recovery at lower counts are also arguments for an earlier start. There is also the possibility that there is a greater incidence of cancer, – other than lymphoma and Kaposi’s sarcoma, at higher CD4 counts in HIV infected people. If this is so then it remains to be shown how frequently these events occur and whether antiviral therapy can avert them.
Treatment itself, particularly if extended over many years, is not without risks, some of which cannot even be completely known yet, particularly with the newer antiviral agents. We have to do the best we can in making a risk benefit assessment. In order to do this we should attempt to obtain information on the rate of disease progress in any one individual. This may not be entirely possible, as the rate of disease progression in any one individual may not be steady; it may accelerate or slow down. But it is possible to obtain a good, if not perfect, picture of the course of HIV disease in any one person.
How might we obtain some information about a given individual’s rate of disease progression? Apart from obvious exceptions, and in people below 200 CD4 cells, there are no emergencies in HIV medicine. For each person we generally will have time to observe the CD4 count and viral load over a period of 6 to 12 months and obtain some idea of the rate of progress. A rapid fall in CD4 count might result in a decision to start in less than six months of observation. Or a consistent fall in CD4 count might lead to a decision to start treatment at CD4 numbers higher than even 500. This is far from perfect, as changes in CD4 cell numbers do not necessarily follow a straight line. But it is far better than basing a decision on a snapshot – which is what the experts are telling us to do.
Individualization involves more than considering the rate of disease progression. There are other factors, such as associated diseases, domestic and social circumstances such as a lack of housing, as well as mental health issues, and many other considerations that are involved in individualization. Observing people also provides the time to establish a doctor patient relationship and for the physician to become familiar with the patients particular circumstances.
The natural history of untreated HIV disease is relevant to the “when to start treatment” issue and will be the topic of the next post.
[1] Evidence supporting the recommendation is derived in part from retrospective observations. The reasons why these are unreliable guides are outlined in the previous post. It is critical to as far as possible, eliminate bias in study designs because this increases the probability that a particular outcome can be interpreted as indeed resulting from a particular intervention. In this case it would be that improved survival is due to an earlier start of antiviral therapy and that the medications mediate the effect – and not for example, from simply being under the supervision of a physician. Retrospective observations, that is, looking back at information already gathered cannot be free of confounding factors as described in the previous post. In a prospective study people would be randomly assigned to receive immediate treatment or to defer it. This will give us the most reliable answer to the question of which approach is better on average.
[2]Examples of measures that should be taken in the treatment of every HIV infected person, irrespective of the rate of disease progression are the types of tests that are performed on the initial assessment of an infected person. For example, the initial assessment of an HIV infected person should always include not only CD4 counts and HIV viral load measurements, but also tests for hepatitis, toxoplasmosis, and many other investigations. Another example of an intervention that is appropriate for categories of infected people is treatment to prevent Pneumocystis pneumonia in people with less than 200 CD4 cells. And of course, people in this category must always be offered antiretroviral therapy.
AIDS and MINORITIES
February 24th 2009
AIDS and Minorities
In the US, African Americans constitute 12% of the population, yet almost half of the total number of AIDS cases in the country occurs among them. This disaster is only now being generally recognized, with the leadership of the Federal AIDS response finally turning their attention to this tragedy, at least publicly.
This is a tragedy that has been developing in full view for more than twenty years. One only has to look at statistics provided by the Centers for Disease Control (CDC) since the epidemic started to know, as early as 1987 that without intervention a preventable disease was inexorably moving into African American communities.
Firstly, take a look at the situation in 2006 (data from the Kaiser family Foundation).
Now take a look at this picture that clearly tells a horrible story that words cannot match.
In the light of this devastating evidence how is one to understand the comments of Dr Fauci, who can be regarded as the head of the Federal scientific and medical response to AIDS?
He noted that these “shocking statistics would be tragic anywhere but are particularly inexcusable in a wealthy country such as the United States.”
His complete statement can be seen here: http://www3.niaid.nih.gov/about/directors/news/BAAID_09.htm
Look at the above picture again. Cases in African Americans started to exceed those in white Americans in 1994, but the trends were quite evident long before this. We knew in 1988-1989 what was coming. So, one must wonder why it took all of twenty years to announce only in 2009, that these statistics are shocking.
A preventable tragedy was taking shape in full view of the Federal AIDS leadership, who rather than fund a vigorous prevention education campaign directed towards those most in danger, instead chose to support a wasteful, vacuous untargeted prevention education program in the form of “America Responds to AIDS”.
It is not only the federal AIDS leadership that failed to respond to warning signals flashing brightly right in front of them. In the early days of the epidemic there was a vigorous and exemplary community activist response. This was a terrific example of people dealing with a deadly disease taking action on their own behalf, fighting for the best medical and scientific response and against the all too frequent shameful stigmatization of HIV infected individuals.
The flowering of AIDS activism in the late 1980s and early 1990s achieved a great deal. All people dealing with serious illness have benefitted from the precedent that was set. Yet, in recognizing this achievement, we must also wonder why many of these experienced advocates, who no doubt were aware of the demographic trends shown above, seemed generally less willing to at least try to avert the disaster threatening their fellow citizens? Of course some tried, and maybe were overwhelmed by massive indifference.
Whatever the reasons, the advocacy of US activists abroad, particularly in Southern Africa, proved to be more effective than anything they were able to achieve in their own country for their fellow African American citizens. There are also other groups where AIDS has been, and continues to be a growing problem, but have been relatively neglected.
Not for profit organizations, that raise funds to help and advocate for people with AIDS were also aware of what was developing, but if there were any efforts devoted to preventing a calamity visibly descending on the largest community at greatest risk, these clearly were of little benefit.
What is almost, but not quite, as shocking as the neglect of a disaster developing in front of our eyes, is the complete absence of any sense of contrition, let alone a simple acknowledgment of failure, on the part of those who might have helped to prevent it.
If justice includes the notion of equity then prevention resources should be distributed in proportion to the needs of different communities. This clearly has not happened and it might be helpful to give some thought as to why this has been the case.
We need to know and admit what it is that we did unsuccessfully, or failed to do, to stop a preventable disaster, so that we can get it right in the future.
Instead we have a wringing of hands by those who could have done something, but did nothing, to stop this.
A similar article was posted on a health related web site in December of last year. There was no response to the issue I raised.
This post can be seen here.
http://Aidsperspective.net/articles/AIDS_AND_MINORITIES.pdf
Take yet another look at the picture showing the changes in the proportion of AIDS cases according to racial/ethnic group over the past 20 years. It tells the whole story; no comment is really needed.
Aidsperspective 