Monday, March 21, 2011

The "Back-Door CD4" and Its Ethics, or Lack Thereof

Every year, virtually every first-year medical student gets introduced to medical ethics by learning about the quirky religious beliefs of a small Christian sect: Jehovah’s Witnesses. Based on their interpretation of a passage from Leviticus, Jehovah’s Witnesses consider blood transfusion to be against God’s word, and will thus not accept them, even if their life depends on it. Medical students are introduced to this situation as the classic example of how an American physician is supposed to behave: it don’t matter what your personal beliefs are with respect to the Witnesses, only that you accept their right to their beliefs and, if need be, protect them, defend them, and do whatever else is necessary to let them decide what to do with their bodies. We call that concept informed consent, and it’s the basis of most serious discussions about life-saving medical care in this country. (Of course, I'm talking about adults here; with kids, the situation gets stickier, and with adolescents, stickier still.)

In theory, informed consent is the process by which patients take the reins and make all the genuinely important medical decisions for themselves, while docs serve as something like advisors. Needless to say I’m not talking about the minutiae of medical decision making, like whether to switch a patient from amlodipine to atenolol, but rather the stuff that most patients worry about when they (or their loved ones) walk through the hospital doors: do I want to be resuscitated? do I want to have “everything” done for me? can I refuse some procedures the doctor recommends? These are precisely the kinds of questions where we, as physicians, have an obligation to help patients and their families figure out what they want—that is, not impose on them what we want—and guide them as best we can, even if, indeed, especially if they make decisions that we find shortsighted or wrongheaded or both.

That’s the theory. Fortunately, in terms of practice, most of the time I think we physicians do a pretty admirable job of supporting our patients. You would be hard pressed to find a doctor who thinks it’s acceptable to override the beliefs of a Jehovah’s Witness and force-feed them blood to save their lives. That said, while the core of that philosophy is wholly adopted by the profession, precisely where the boundaries lay can be contentious. And last week I was again reminded that we don’t all agree on what constitutes honoring our patients’ wishes.

It happened when I was hearing about a case of a patient in the Intensive Care Unit fit for an episode of House, M.D. The patient was older without being elderly, and had a respiratory illness that had defied diagnosis despite the best intentions of cardiologists, pulmonologists, infectious disease specialists, and a few other medical professionals to boot. The ID docs, though unsure of what was going on, thought that this was most likely some infection seen in the setting of underlying AIDS, and I, being an ID guy, shared their point of view while hearing the details. Like my colleagues, the question to which I wanted to know the answer was: what was the result of the HIV test? But the patient had adamantly refused HIV testing. And that’s within his right: a doctor can order most blood tests without having to discuss them with a patient, but an HIV test—just like a blood transfusion—is special, and requires a signed form saying that the patient agrees to it. In this case, where a diagnosis of HIV infection might be helpful (more on this later), such a refusal can be maddening. But that is how the rules are set up right now, and once the patient says no, then that’s all she wrote.

However, in this case, that wasn’t all she wrote. HIV is a virus that infects, and destroys, a special kind of white blood cell called a CD4 cell. In general, the further your CD4 cell count drops, the more you are at risk of being infected by the weird organisms that are the sine qua non of AIDS, things like toxoplasmosis, cryptococcosis, penicillium marneffei and a host of other parasites, bacteria and fungi that people with healthy immune systems never develop. A person with a healthy immune system typically has a CD4 count that runs from 500 to 1500, give or take; the definition of AIDS is someone with HIV infection and a CD4 count less than 200.

One of the loopholes of informed consent for HIV testing is that it does not cover CD4 counts.

You can see where this is going. Some doc tried to do an end run around the refusal and checked the patient’s CD4 count. Surprise! It was low, less than 200, although the pattern of the CD4 cells didn’t really look like AIDS (skipping some technical detail here). So then what do you do? You have gotten no closer to the diagnosis, and you have put yourself into the uncomfortable situation where you may be tempted to take action. Sometimes the treatment for some of those weird “opportunistic infections” in AIDS is to just give a person medications for HIV, but you wouldn’t give those meds to a patient without a diagnosis of HIV. In this situation, that wasn’t the case, but what if it were? Would you throw antiretrovirals at the person because his CD4 count was suggestive of HIV—even though the patient unambiguously refused the test? To me, this smells exactly like giving blood to a Jehovah’s Witness: doing what I call a “Back-Door CD4” might seem clever, but it pretty obviously violates the spirit of the patient’s wishes, if not the letter.

Some would argue that the very need to consent people for HIV is outdated. HIV consent was established at a time when the diagnosis was severely stigmatizing, and “positives” could lose jobs, insurance premiums and in general face ostracism from their communities; it also happened at a time when treatment wasn’t exactly effective. Today, the latter is definitely not true, and as I tell my HIV patients all the time, there is no reason to suppose that, should they take their meds every day faithfully, they should live as long as anyone else. Whether the former is true I am a touch skeptical, but I acknowledge that the level of stigma nowhere near approaches where it was twenty years ago. Thus, proponents of the Back-Door CD4 would say that the time for HIV consent has come and gone. Way back in 2004 a doc named LA Jansen wrote in the Journal of Medical Ethics and supposed that the Back-Door CD4 was a form of “conscientious subversion,” something akin to conscientious objection, where a physician acknowledged the existing legal landscape but did his or her own thing based on their personal ethics.

Jansen dresses the term up in calling it “conscientious subversion”; I prefer using simpler language and think of it as a bad idea. By my compass, patients have rights to refuse tests, medications, procedures, and any attempt at thwarting those desires defeats not only the entire point of informed consent, it belies the idea that we are advocates for our patients. Don’t get cute, is what I’d say: we are better than most professions at standing by those we serve. Let’s not mess things up by thinking we know better than they do.


PS—the Jehovah’s Witness example has, for me, not been entirely academic. Twice I have cared for Witnesses who were in situations where transfusion was definitely worth considering, and in one of the cases it was pretty clearly indicated. Like most of my profession, it never entered my mind to try to push the idea on them once I learned of their religious beliefs. For more on Jehovah’s Witnesses and their philosophy behind their refusal of blood transfusions, see here, here and/or here.


  1. Joseph P Arpaia, MDMarch 26, 2011 at 12:07 AM

    I disagree here. Since there may be other conditions which can cause a low CD4 level, as may be happening with this patient, the tests are not equivalent. With the low CD4 result and the refusal for an HIV test your task is to rule out other causes of a low CD4 level.

  2. But any of those conditions require an HIV test to rule out the most common cause of CD4 depletion, so I'd argue that you've turned the situation upside-down. It's possible that this guy had idiopathic CD4 lymphopenia, an extremely rare disorder, but you still have to rule out HIV in order to make the diagnosis. And regardless, I think it's safe to say in this case that the physicians pretty clearly intended to subvert the patient's wishes viz. refusing to consent for the HIV test. This isn't the first time I've seen some ICU physicians at my medical center do this.