Correspondence

Clinical Problem-Solving: Refusing HIV Testing

N Engl J Med 1998; 338:1544-1546May 21, 1998

Article

To the Editor:

The Clinical Problem-Solving case “A Hidden Agenda” (Jan. 1 issue)1 demonstrates the challenges involved in diagnosing an illness in a patient who denies having risk factors for human immunodeficiency virus (HIV) infection and refuses testing. We recognize the importance of confidentiality in the doctor–patient relationship, but we also acknowledge that there are times when this must be weighed against the risks to others that could be minimized or averted by disclosure. This is the lesson of the famous Tarasoff case, in which it was held that a psychologist had a duty to warn a murder victim and her family that his client had threatened to kill her.2 Although information related to HIV status is generally regarded as highly confidential, even states that give it statutory protection recognize that there can be a need for disclosure to persons at risk. New York's statute,3 for example, states that the physician should counsel the infected person about the importance of informing contacts. If the physician reasonably believes the infected person will not do so, then the physician may make the disclosure after informing the infected person and giving the infected person the option of having the disclosure made by an officer of the health department.

David R. Lambert, M.D.
Jane Greenlaw, J.D.
University of Rochester School of Medicine and Dentistry, Rochester, NY 14642

3 References

1
Landor M. A hidden agenda. N Engl J Med 1998;338:46-50
Full Text | Web of Science | Medline

2
Tarasoff v. Regents of U. of California, 131 Cal. Rptr. 14, 551 P.2d 334 (1976).

3
New York State Confidentiality Law, N.Y. Pub. Health Law Art. 27-F (McKinney 1997).

To the Editor:

The patient described in “A Hidden Agenda” had a wife and four-year-old and seven-year-old children who were apparently healthy. Late in the hospitalization he admitted to having been aware of his HIV seropositivity for eight years, which means that he had knowingly exposed his wife and children to this infection. Was there any attempt to ascertain whether his family was aware of his HIV status or had been tested for the virus?

Recently, officials in the state of New York arrested and pressed charges against a person who had knowingly exposed other people to HIV without informing them. And this case is not the only one in which jurisdictions have attempted to hold individuals accountable for such behavior.

I would like to know how this patient's physicians dealt with this issue and what, if any, law would have protected his family or partners. How does one protect the patient's right to privacy while protecting innocent people from a deadly disease? This is a troubling and serious ethical dilemma.

BobbieJean Sweitzer, M.D.
Massachusetts General Hospital, Boston, MA 02114

To the Editor:

Landor acknowledges that the patient's denial of risk factors for HIV and refusal to undergo definitive testing resulted in an unnecessarily long hospital stay and expensive evaluation. This workup was done despite the physician's belief and indirect evidence that the patient had HIV infection.

Does a physician have an obligation to provide nonemergency care to a patient who refuses a medically diagnostic test? I doubt that a physician who refuses to treat a patient because the patient does not consent to a key test violates state antidiscrimination statutes if the basis for the refusal is not the patient's HIV status itself but the physician's belief that he or she cannot appropriately treat the patient without that knowledge.

James Svenson, M.D.
University of Wisconsin, Madison, WI 53792

To the Editor:

I do not believe that knowing the patient's HIV status would have altered the evaluation in any appreciable way, because his CD4+ cell count was relatively high at 592 per cubic millimeter (26 percent). With this count the patient was not particularly predisposed to opportunistic infections. He should have been evaluated in the same manner as an HIV-negative patient. If he had had a CD4+ cell count below 200 per cubic millimeter, an entirely different set of diagnostic possibilities would have become more likely and prompted a search for such things as cytomegalovirus, Mycobacterium avium complex infection, extrapulmonary tuberculosis, and cryptococcus. In that situation a workup for a fever of unexplained origin would be altered by the knowledge of the patient's HIV status, but not in the one under discussion. Looking for these diagnoses in an HIV-positive patient with a high CD4+ cell count is generally inappropriate and not cost effective. I do not believe that “paradoxically, the patient may in this case have benefited from hiding his HIV seropositivity.” It was not paradoxical; this patient's HIV-positive status should not have altered the workup for his fever of unexplained origin.

Stephen J. Gluckman, M.D.
Hospital of the University of Pennsylvania, Philadelphia, PA 19104

To the Editor:

. . . If the law requires informed consent for HIV testing, why not require it for measurements of T-lymphocyte subgroups? Why not for differential white-cell counts (in an appropriate clinical setting, a very low lymphocyte count strongly suggests HIV infection)? Why not for an inspection of the tongue (hairy leukoplakia is virtually diagnostic of HIV infection1)? In our view, as the case report demonstrates, requiring informed consent only for HIV testing not only provokes absurd diagnostic difficulties, but also contributes to the psychological burden and stigma associated with HIV infection.

Bernardino Roca, M.D.
Enrique Simón, M.D.
Hospital General of Castellón, E-12004 Castellón, Spain

1 References

1
Wiebe CB, Epstein JB. An atlas of HIV-associated oral lesions: a new classification and diagnostic criteria. J Can Dent Assoc 1997;63:288-9, 292
Medline

Author/Editor Response

Dr. Landor replies:

To the Editor: Dr. Gluckman questions the relevance of HIV status to the workup for a fever of unexplained origin in patients with high CD4+ cell counts. The diagnosis of a fever of unexplained origin mandates a comprehensive, step-by-step evaluation that includes occult fungal and parasitic infections.1 The presence of seropositivity for HIV, however, dictates a rearrangement of the list of differential diagnoses so that mycobacterial infections, non-Hodgkin's lymphoma, and unsuppressed HIV infection itself become the first targets for investigation. The case under discussion actually represents the unusual situation in which, paradoxically, disclosure of HIV seropositivity would have been more damaging than beneficial, since it would have delayed the detection of the patient's underlying, asymptomatic cancer.

Dr. Svenson expresses concern about the undue utilization of medical resources for a patient who refuses a medically indicated diagnostic test. Drs. Roca and Simón find that the legal requirement of informed consent for HIV testing does not always protect the patient's privacy, because HIV infection can often be detected without serologic testing. To the extent that they propose eliminating the requirement, I think they are misguided, because the advantage of respecting an individual's right to autonomy over health care decisions far outweighs the disadvantages highlighted in these letters. In the case of HIV, that right is critical because of the overwhelming stigma that continues to color this infection and because of the enormous personal, social, and medical consequences of a diagnosis of HIV.

Without the informed-consent requirement, many persons would avoid the health care system out of fear of being tested without their consent. Moreover, this fear can seriously damage the physician–patient relationship. The stigma of HIV infection threatens the social and psychological integrity of those infected and must be addressed. It has been my experience that whenever mutual trust was carefully established and absolute respect was demonstrated for the variety of fears (“My family will discover my bisexuality”), misinformation (“Herbal extracts cure AIDS”), and misgivings (“Zidovudine kills blacks”) that patients may have, the patients almost invariably overcame their reluctance either to be tested for HIV or to use antiretroviral medications.

Lambert and Greenlaw and Sweitzer point out the apparent conflict between patient confidentiality and public health. The state of New York attempts to reconcile that conflict by permitting a physician to notify a sexual or needle-sharing partner of a seropositive patient without the patient's consent when the physician believes that the disclosure is medically appropriate, the partner is at high risk for HIV infection, and the patient will not inform the partner after being counseled to do so (Friedman S, Legal Action Center: personal communication). Alternatively, the physician may notify public health officials so that they can notify the partner.2 A physician has no legal obligation, however, to identify or locate any contact.3 For many patients it is difficult to think about notification of partners immediately after the infection is diagnosed. In my experience, developing a trusting relationship with the patient and offering to become personally involved usually lead to direct notification of partners by the patients. In the exceptional cases in which this approach failed, as eventually happened with the patient under discussion, the state's anonymous partner-notification program was used.

Michael Landor, M.D.
69-20 Main St., Flushing, NY 11367

3 References