A Proposed National Policy on Health Care Workers Living With HIV/AIDS and Other Blood-Borne Pathogens

A decade ago the Centers for Disease Control and Prevention (CDC) reported a cluster of 6 cases of transmission of human immunodeficiency virus (HIV) infection from a Florida dentist to his patients.^{1 - 5} These cases, in which the mode of transmission was never determined, had a direct effect on federal and state policy. In 1991, after a year-long national debate,⁵ the CDC recommended that health care workers (HCWs) infected with HIV or hepatitis B virus (HBV) (hepatitis B e antigen [HBeAg]–positive) should be reviewed by an expert panel and should inform patients of their serologic status before engaging in exposure-prone procedures.⁶ Later in 1991, Congress mandated states to adopt the CDC guidelines or their equivalent.⁷

Public health authorities, in the face of scientific uncertainty, took the path of caution.⁸ However, considerable evidence has emerged since 1991 suggesting that Congress and the CDC should reform national policy. The data demonstrate that risks of transmission of blood-borne pathogens in the health care setting are exceedingly low. Current policy, moreover, does not improve patient safety. At the same time, implementation of the CDC policy at the local level poses significant human rights burdens on HCWs. Consequently, national policy should be changed to end professional practice restrictions and the requirement to disclose an HCW's HIV or HBV infection status to patients.

After describing current CDC guidelines, this article presents evidence from the fields of epidemiology and law to demonstrate why revision of the national policy is desirable. While the focus is on HIV infection, much of the analysis is relevant to other blood-borne infections. The article concludes by proposing a new national policy that emphasizes patient safety by ensuring that infection control procedures are systematically implemented in health care settings. The policy, therefore, focuses on safer systems of practice rather than on excluding and stigmatizing infected HCWs.

In 1991, the CDC recommended that HCWs infected with HIV or HBV (HBeAg-positive) should not perform exposure-prone procedures unless they have "sought counsel from an expert review panel and [have] been advised under what circumstances, if any, they may continue to perform these procedures."⁹ Even if the panels permit them to practice, HCWs must still inform patients of their serologic status. The CDC defined an exposure-prone procedure to include "digital palpation of a needle tip in a body cavity or the simultaneous presence of the health care worker's fingers and a needle or other sharp instrument or object in a poorly visualized anatomic site." The CDC sought assistance in clarifying the definition of exposure-prone procedures, but professional medical and dental associations declined to specify such procedures.¹⁰

In October 1991, Congress enacted a statute requiring each state, as a condition of receiving Public Health Service funds, to certify that CDC guidelines "or their equivalent" had been instituted.⁷ During the following year, the CDC notified states that the agency would be flexible in accepting the certifications of equivalency. States could determine exposure-prone procedures "on a case-by-case basis, taking into consideration the specific procedure as well as the skill, technique, and possible impairment of the infected HCW."¹¹ As a result, variability exists in state law and practice. In particular, many states do not require disclosure to patients even though this is currently recommended by the CDC.¹²

In December 1995, the Presidential Advisory Council on HIV/AIDS asked President Clinton to instruct the CDC to review its guidelines that arbitrarily restrict HIV-infected HCWs and lead to discrimination. Later, professional organizations recommended a change in policy so that HCWs are not required to be tested, hospitals are not required to notify patients, and infected HCWs are not excluded from practice.^{13 - 14} The CDC is currently considering updating its guidelines for the management of infected HCWs. In law, the CDC is entitled to withdraw its 1991 recommendations without congressional approval and to replace them with new guidelines.

A change in CDC guidelines would significantly affect national policy for 2 reasons: First, as explained above, current federal legislation requires states to adopt CDC guidelines or their equivalent. Thus, changes in CDC guidelines should result in statewide reforms. Second, under state tort law, courts usually measure professional and hospital practices against a national standard of care. The CDC guidelines do not necessarily set that standard, but they are highly influential in judicial cases.¹⁵

It should be emphasized that the other branches of government may not necessarily agree with the new federal agency guidelines. The issue of infected HCWs is still politically charged, and Congress holds the ultimate power to set policy. Congress, therefore, could enact new legislation reinstating an exclusionary policy with respect to infected HCWs. Furthermore, the courts are not bound by CDC guidelines. Judges retain discretion in applying tort law and disability discrimination law. This could mean that the courts will continue to decide cases against the interests of infected HCWs. The result, of course, is that hospitals still might insist that infected HCWs refrain from practice or, at least, inform patients because of the hospital's fear of liability.

At the time of the 1991 guidelines, little evidence existed to assess the risk of HIV transmission from HCWs to patients. Since that time, data from reported cases, retrospective investigations, and national HIV/AIDS surveillance show that the risk is exceedingly remote. Since the beginning of the HIV epidemic, nearly 2 decades ago, few episodes of HIV transmission from HCWs to patients have been documented. In addition to the Florida dental cluster, a single case was reported in 1997 involving an orthopedic surgeon in France with advanced symptomatic but undiagnosed HIV disease.^{16 - 17} A further possible case of nurse-to-patient transmission in France was reported in 2000.¹⁸

Comprehensive retrospective investigations, particularly among physicians engaged in invasive procedures, have not identified additional cases.^{19 - 20} As of July 1999, the CDC had analyzed HIV test results for more than 22,000 patients of 63 HIV-infected HCWs, and no documented case of transmission had occurred.²¹ Similarly, in 1997, 1180 surgical patients of an HIV-infected obstetrician/gynecologist were tested in the United Kingdom; none were found to be HIV positive.²² Finally, state health department follow-up of reported cases of HIV or AIDS have failed to confirm additional cases of HCW-to-patient transmission.

There are theoretical reasons, in addition to the epidemiological data, to believe that the risk of transmission is exceedingly remote. New therapies can reduce HIV viral load in plasma to very low levels, and assays to measure viral load are available.²³ Consequently, if HIV-infected HCWs are in treatment, it ought to be possible to monitor their health status and ensure low viral loads, thus even further reducing risks to patients.²⁴

The likelihood of transmission of HBV is significantly greater than HIV. For example, the rate of seroconversion after accidental percutaneous exposure to HBV is approximately 12% to 17%, compared with 0.03% to 0.9% with HIV.²⁵ Despite the elevated risk, HCW-to-patient transmission of HBV also has been rare since the early 1990s. In 1992, a cluster of 19 cases of transmission occurred from an HbeAg-positive cardiothoracic surgeon in California who did not double-glove even after infection was diagnosed.²⁶ Since this cluster of cases, no episode of HCW-to-patient HBV transmission has been documented in the United States, although transmission has been documented in the United Kingdom.^{27 - 30} A cluster of HBV infections did occur in the United States between December 1995 and May 1996 among patients undergoing long-term hemodialysis, but the apparent mode of transmission was through the use of medical equipment shared among infected and uninfected patients.³¹

The 1991 guidelines did not address hepatitis C virus (HCV) because it had been only recently discovered, and there were few epidemiologic and laboratory data to support guidelines. In 1998, the CDC found the risk of HCV transmission by HCWs to be "very low."³² There is 1 documented cluster of cases of HCV transmission from a cardiothoracic surgeon in Spain to 5 patients between 1988 and 1993, but investigation of those cases did not identify factors that might have contributed to the transmission.³³ With that exception, reported transmission of HCV from HCWs to patients has been rare.^{34 - 35} Model-based calculations similarly suggest a low rate of HCV transmission from surgical staff to patients.³⁶

Currently, no federal policy exists regarding management of HCWs infected with HCV. However, the American College of Surgeons states that "Currently, there is no indication for surgeons to take special measures to protect their patients except during acute, symptomatic illness with hepatitis C."³⁷ The level of risk entailed in HCV transmission in the health care setting appears to be comparable with other blood-borne infections. Consequently, national policy should include strategies for the management of HCV as well as HIV and HBV.

The published data suggest that the risk to patients of contracting HIV infection in a health care setting is remote. While the data for HBV and HCV are not as clear and settled, transmission to patients appears to be rare. The risks of blood-borne disease transmission, moreover, are significantly lower than comparable risks faced by patients, such as the possibility of adverse outcomes from medical error.³⁸ Furthermore, if all transmissions from HCWs to patients were entirely prevented, the public health burden from the HIV epidemic would be reduced infinitesimally.³⁹

Sanctions against HIV-infected HCWs would be ethically permissible if they were necessary to avert a serious risk to patients or if the limitations on human rights were trivial. At present, however, the risks to patients are negligible, and the burdens on HIV-infected HCWs are disproportionately high. Qualified and experienced professionals face discrimination, invasion of privacy, and potential liability. An unintended consequence of CDC guidelines is that they have become a nationally recognized standard of care used by health care organizations and the courts to justify discrimination and invasion of privacy of infected HCWs.⁴⁰

The Americans With Disabilities Act (ADA) and similar state statutes⁴¹ proscribe employment discrimination against "qualified" persons with a disability. The Supreme Court has held that all stages of HIV disease, including asymptomatic infection, are covered disabilities under the ADA.⁴² Persons are qualified if they are capable of performing the essential performance functions of the job. Notably, "qualification" standards include a requirement that a person does not pose a "direct threat" to the health or safety of others.⁴³ The ADA, however, requires reasonable accommodations if they would enable persons with disabilities to perform their job safely (ie, without posing a significant risk) and effectively.⁴⁴ Thus, if reasonable accommodations could help HCWs practice safely, health care organizations must provide those accommodations.

The critical questions, therefore, are whether HIV-infected HCWs pose a significant risk to patients and, if so, whether there are reasonable accommodations to reduce that risk. The Supreme Court assesses "significant risk" by its nature (the mode of transmission), probability (the likelihood of transmission), severity (the magnitude of the harm should the infection be transmitted), and duration (the length of time the person remains infectious).⁴⁵ The US Equal Employment Opportunity Commission states that "an employer . . . is not permitted to deny an employment opportunity to a person with a disability merely because of a slightly increased risk."⁴⁶

Despite this clear language, the courts have held that because HIV infection carries grave consequences, even a theoretical risk of transmission can be sufficient to justify discrimination. The courts have had little difficulty finding that HIV-infected HCWs pose a significant risk,⁴⁷ often resolving the issue through summary judgment without medical facts and expert testimony.⁴⁸ These courts have relied on the severity of harm should the risk materialize. One court allowed a hospital to reassign a surgical assistant to the procurement department while recognizing the "small" risk: "The duration of infection is perpetual. And the virus inevitably leads to the fatal disease AIDS."⁴⁹

This kind of reasoning perverts the significant risk standard because it means that individuals can be discriminated against in the absence of an appreciable public threat. A federal court of appeals, for example, recently upheld the segregation of HIV-infected inmates, arguing that they pose a threat to the prison population.⁵⁰ Because significant risk is a product of 4 criteria (nature, probability, severity, and duration), if any criterion demonstrates a very low chance of serious harm, then the risk should be "insignificant."⁵¹ Notably, courts use CDC guidelines to support their position: "We must decide in light of CDC's recommendation that hospitals may bar HIV-positive surgeons from performing [exposure-prone] procedures. . . ."⁵²

Even if HIV-infected HCWs pose a significant risk, hospitals have a duty to provide reasonable accommodations, such as infection control, training, and a leave of absence for treatment. If HIV-infected HCWs were to receive this kind of support, the great majority would be able to function effectively and safely.

Despite the potential of the ADA to safeguard the rights of workers, the courts and administrative agencies, with few exceptions,^{53 - 54} have upheld discrimination against HIV-infected HCWs. The judiciary has permitted decisions to exclude HCWs from medical⁵² and nursing⁵⁵ practice. In response to concern about the Florida dental case, one court upheld the exclusion of an HIV-infected student from a dental school.⁵⁶ Despite the legal standard that the employer bears the burden of proving that the practice poses a significant risk of harm, in reality, the courts have shifted the burden to HCWs to show they pose virtually no risk.⁵⁷ The courts have approved discriminatory action not only for exposure-prone procedures but also for less invasive procedures.^{58 - 59} Notably, some courts have upheld restrictions despite expert review panel advice that the HCW should continue to practice.^{52 ,56} These are cases in which public health authorities have complied with the ADA's mandate for individualized determinations, and the judiciary has nonetheless declined to protect HCWs' rights.

The CDC requirement to disclose an HCW's HIV-status to patients before performing exposure-prone procedures has become common practice. Patient notification is supported by cultural, ethical, and legal reasoning. American cultural thought strongly supports a patient's "right to know." Patients want to be informed and to make their own decisions whether they want to incur risks. Similarly, infected physicians may feel they owe an ethical duty to patients and the wider public to disclose their serologic status. Physicians may choose to "put patients first" by disclosing their serologic status, even if the risk of transmission is very low. Finally, the legal doctrine of informed consent appears to support disclosure. In many states, the law requires physicians to disclose risks that "reasonable" patients would want to know; and many patients do want to know if their physician is infected with HIV.

Despite these significant arguments, the law should not require HCWs to disclose their HIV status to patients. A legal obligation to notify patients is an invasion of the HCW's privacy. It is important to recall that the HCW also is a patient who is being treated for a stigmatizing disease. Forced disclosure of this sensitive information may be socially embarrassing and may harm the HCW's livelihood and professional reputation.⁶⁰ Moreover, the invasion of privacy is extensive because patients have no duty of confidentiality and may spread the information to other patients and the media.⁶¹

The doctrine of informed consent, moreover, should not require HIV-infected HCWs to disclose their HIV status to patients. First, informed consent requires disclosure of "material" risks, not remote risks. The risk of HIV transmission from HCW to patient is too low to meet the legal standard for disclosure. Moreover, if the risk were significant, the logical remedy would be to restrict the HCW's right to practice, not to notify the patient. Public health authorities would not permit a clinician to practice knowing he/she poses a meaningful risk, even if patients were notified. Second, informed consent usually requires disclosure of risks entailed in the medical procedure, not those posed by the HCW. Health care workers ordinarily are not required to inform patients about their personal health status or disability (whether temporary or permanent). For example, HCWs who have depression, have had insufficient sleep, or have had alcoholic beverages at a social occasion in the hours preceding a patient encounter are not ordinarily required to disclose these factors to patients. In each case, the HCW may pose a risk that is the same as, or greater than, the risk of HIV transmission. Yet, provided the HCWs are considered competent and reasonably safe, they are entitled to practice without a duty to disclose.

State courts routinely affirm hospital decisions to disclose the HCW's HIV status to patients. Relying on CDC guidelines, these courts require compulsory disclosure as part of the informed consent process.⁶² Some courts even permit disclosure for the purposes of retrospective investigations.⁶³ In Faya v Almaraz,⁶⁴ for example, an HIV-infected oncologic surgeon performed a partial mastectomy on 2 occasions. When the patients read that the physician had died of AIDS, they filed a lawsuit alleging emotional distress. The court allowed damages for the period between the time that the patients learned of the physician's HIV status and their conclusively testing negative for HIV. A similar result occurred in other court decisions.^{65 - 66} In these kind of cases, hospitals and HCWs face liability even though patients experienced no tangible harm or infection.⁶⁷ This litigation, moreover, is not restricted to HIV, but could apply equally to other infectious diseases, such as HBV, HCV, or tuberculosis.

Certainly, claims of emotional distress sometimes fail because the patient cannot prove actual exposure to a blood-borne pathogen as a result of a physical injury.⁶⁸ However, even if litigation is unsuccessful, it can influence hospitals to restrict HCW rights more than they think may be justified in the public interest. Hospitals, fearing potential liability, may assume it is less costly to mandate patient notification or restrict an HCW's practice than to face litigation. The problem is exacerbated by malpractice liability concerns if insurers urge "screening out" infected HCWs rather than incur liability risks.

The CDC does not recommend HIV testing in the absence of behavioral risk factors. Nor does the agency recommend mandatory testing. Nevertheless, hospitals occasionally require HCWs to be tested against their will, and courts have approved policies of this kind. For example, a federal court of appeals upheld a hospital's decision to fire a licensed practical nurse who was suspected of being HIV infected but refused to be tested. The court reasoned that the hospital was prevented from "deciding what, if any, measures were necessary to protect the health of patients."⁵⁵ Courts have held that HIV testing can be "job related and consistent with business necessity."⁶⁹ In one case, a training school obtained a medical assistant's blood sample purportedly to test for rubella but instead performed an HIV test.⁷⁰ The court found an invasion of privacy, but the case shows the lengths to which some facilities will go to identify HIV-infected HCWs. If the standard of care remains that HCWs must notify patients and possibly discontinue practice, then health care organizations and the courts may believe that compulsory testing is an appropriate response.

Health care workers living with HIV/AIDS face loss of their livelihood, professional status, and self-image. Health care workers have been terminated, forced to resign, reassigned, denied rotations, or not permitted to continue their education.⁷¹ Many HCWs, because of the fear of discrimination, do not seek diagnosis and treatment because they have greater legal protection if they can honestly say that they did not know their serologic status. Some pay for treatment out of their own pockets to avoid filing claims with their health insurance carriers. Others travel outside their geographic areas to protect their anonymity or avoid treatment altogether.⁷²

Current policy and practice adversely affect not only the professional and personal interests of HCWs but also may not be in the public interest. Punitive policies may provide disincentives for HCWs to determine their HIV status or to disclose it to their employers or personal physicians. Consequently, the CDC policy's restrictive approach may reduce HCWs' opportunity for therapy, thereby increasing risks to patients.

Current national policy offers no discernible risk reduction for patients—very few, if any, new blood-borne infections are prevented. At the same time, the policy imposes human rights burdens, undermines efforts to retain experienced professionals, and poses liability risks. Consequently, a new national policy should focus on structural changes to make the health care workplace safer for both patients and HCWs rather than on identification and management of infected HCWs.^{73 - 74} Certainly, patient safety remains the most important part of any national policy, but it ought to be possible to ensure patient safety while not undermining the HCW's interests in autonomy, privacy, and livelihood. The following 5 recommendations should ensure that patients receive care in a safe environment, while treating HCWs with respect and dignity.

Health care organizations should be responsible for planning, implementing, enforcing, and evaluating effective strategies for prevention of blood-borne pathogen transmission. Prevention programs should include policies and procedures for (1) standard (universal) precautions (eg, barrier equipment, such as gloves, gowns, and face protection)^{75 - 76} ; (2) reprocessing patient-care equipment (eg, cleaning, disinfecting, and sterilizing) and cleaning environmental surfaces; (3) infection control training (in professional schools and in the health care setting); (4) prevention and management of infectious conditions (eg, HBV vaccination); (5) injury and exposure prevention during surgical, obstetric, and dental procedures, such as less invasive alternatives to conventional interventions (eg, laparoscopy), practice changes (eg, "no touch" techniques and double-gloving), and safer instruments (eg, blunted suture needles and use of staples instead of sutures); and (6) surveillance for transmission of blood-borne pathogens (eg, reporting and evaluating parenteral injuries).

Health care workers, as professionals, have ethical responsibilities to promote their own health and well being and to ensure patient safety. Health care workers should learn their serologic status through testing with informed consent. Infected HCWs should seek medical care and treatment, including ongoing monitoring of viral load, as well as evaluations of physical and mental health status. Health care workers should be responsible for notifying patients and hospitals in all cases of significant exposure to the HCW's blood.

Identifying infected physicians, requiring expert review panels, and adopting special restrictions for exposure-prone procedures have served to single-out and stigmatize HCWs. The term exposure-prone procedure has not been clearly defined, so the courts have erred on the side of restricting HCW practice. Expert review panels, moreover, are constituted differently in each state, so there are multiple standards of care; even when panels permit practice, health care organizations sometimes disregard their advice. The requirement to disclose an HCW's serologic status to patients has rendered expert review panels virtually irrelevant since few patients are willing to be treated by an infected HCW. Review panels, therefore, should be replaced by expert consultants who would be available for advising about HCWs' health and patients' safety but not empowered to restrict HCW practice.

Health care workers may feel morally obliged to voluntarily notify patients of their infection status, but the law should not require disclosure of very low-level risks. In particular, HIV is a highly personal and sometimes stigmatic health condition that usually has little relevance to patient safety. Because notification represents an invasion of privacy and may result in loss of livelihood, it should not be legally mandated.

Public health authorities owe a public duty to ensure patient safety. Consequently, health care organizations and/or public health authorities should issue practice restrictions if an HCW (1) has a physical or mental impairment that affects his/her professional judgment or practice; (2) has exudative lesions or weeping dermatitis; (3) has a history of poor infection control technique and practice; or (4) has had an incident of transmission of a blood-borne pathogen. In these kinds of cases, the HCW cannot practice competently and safely. Consequently, health care organizations and public health authorities owe a primary duty to safeguard patient safety.

The health care system must be made as safe as possible for both patients and their health care providers. The HIV epidemic has driven the government and the public toward achieving near-zero risk. Such a standard is not achievable and, in the process, existing national policy has imposed human rights burdens, driven qualified HCWs from the medical profession, and posed unfair liability risks. A new national policy, focused on management of the workplace environment and injury prevention, would achieve high levels of patient safety without discrimination and invasion of HCW privacy.