There is currently no public policy that provides guidance concerning whether and when physicians infected with hepatitis B virus (HBV), hepatitis C virus (HCV), and/or human immunodeficiency virus (HIV) can safely perform invasive procedures. A committee of experts in the fields of medicine, law, and biomedical ethics and 1 community member, aided by an advisory board, was established to produce recommendations for policy reform. An extensive literature review was conducted for these 3 infectious diseases, medicine, surgery, epidemiology, law, and bioethics to gather all relevant data. Special recommendations are made regarding the management of physicians who are infected with HIV, HBV, and/or HCV. This policy proposal includes a list of exposure-prone procedures and a decision chart that indicates under what conditions infected physicians can practice beyond the need for disclosure of their serological status.
The question of whether physicians infected with blood-borne pathogens should perform invasive procedures—and, if so, under what conditions—was raised by reports of the transmission of hepatitis B virus (HBV) [1–14], hepatitis C virus (HCV) [15–21], and HIV [22–24] from health care workers to patients during invasive procedures. Despite evidence of such transmissions, there is no uniform national policy that provides definitive guidance concerning whether and under what conditions infected physicians can practice. The lack of a uniform national policy means that patients' rights and the viral transmission risks patients assume do not depend on a national standard but on the state or even the institution where treatment is provided.
In response to the HIV and HBV transmission problem, the Centers for Disease Control and Prevention (CDC) issued Recommendations for Preventing Transmission of Human Immunodeficiency Virus and Hepatitis B Virus to Patients During Exposure-Prone Invasive Procedures [25]. (Although the CDC has stated for years now that it is in the process of revising the 1991 guidelines, proposed drafts of the recommendations are embargoed, and there is no projected publication date for revised guidelines as of this writing.) The CDC recommendations impose no restrictions on infected health care workers with regard to performance of specific invasive procedures. Instead, the CDC recommendations state that infected health care workers “should not perform exposure-prone procedures unless they have sought counsel from an expert panel and notify prospective patients of the healthcare worker's seropositivity prior to undergoing an exposure-prone procedure.” This expert panel should consist of health professionals from a variety of medical backgrounds. The CDC rejects mandatory testing (but encourages voluntary testing) of health care workers who perform invasive procedures. A 1998 document published by the CDC on prevention and control of HCV infection states that there are no recommendations that restrict the professional activities of health care workers with HCV infection [26]. The American College of Surgeons (ACS), the largest and arguably most influential professional organization to represent surgeons in the United States, recently made a statement about surgeons who are infected with HBV and/or HCV. Other than encouraging surgeons to know their immunization status, the ACS does not suggest any practice restrictions for infected surgeons [27].
Federal legislation requires states to adopt the CDC recommendations or their equivalents. The CDC recommendations define characteristics of exposure-prone procedures but do not identify specific procedures. This is left to state health departments to decide. Also left to the discretion of the states is the issue of disclosure. In some states, infected physicians must inform patients of their infection status before undertaking an invasive procedure. Others leave it to the discretion of the review panel to determine whether physicians must disclose their infection status. Of course, this is all dependent on voluntary reporting of infectious status.
There is a wide variation in the proposed management of infected physicians [28–38]. Most guidelines offer no, minimal, or fairly nonspecific guidance as to what constitute exposure-prone procedures. Meanwhile, the issue of disclosure remains controversial. Most organizations do not recognize a need for disclosure; some guidelines still favor either postexposure or preprocedure disclosure and/or identification of the source of infection. Our committee has attempted to move the issue beyond disclosure of the physician's serological status. Disclosure is an unnecessary, ineffective, and inappropriate remedy for guiding the practice of infected physicians. The significance of the potential for harm, we argue, is not mitigated nor remedied by disclosure of a physician's viro-serological status to a patient. Others have argued that the law should not require health care workers to disclose their HIV status (or HBV or HCV status), because this is an invasion of privacy [39]. Perhaps most importantly, serostatus disclosure, as proposed by the CDC, does not improve patient safety, as was previously argued by Closen [40, 41] and others. It has been duly noted that the acknowledged social risks of disclosure could be avoided by refraining from performing procedures that expose patients to the risk of transmission [42].
In those cases in which the risk of viral transmission is deemed to be negligible and in which procedures can thus be performed safely, a priori informing of patients of their physicians' serostatus is unnecessary and will only cause more harm in the form of patient anxiety and physician stigmatization. In the unlikely event that a “bleedback” injury does occur during such safe procedures and an infection becomes a theoretical possibility, a posteriori patient notification is in order for timely serological testing and prophylaxis. Identification of the source of the potential viral transmission is not appropriate, because it does not serve a medical purpose. This approach is in agreement with the official guidelines from both the UK Department of Health [43, 44] and the Canadian Medical Association [29]. In a recent publication, a European Consensus Group left disclosure of serostatus up to the health care worker performing exposure-prone procedures. The same group advocated limited practice restrictions for HBV-infected health care workers, but they could not reach a consensus on HCV-infected health care workers [45].
In 1999, one committee member (P.M.T.) and colleagues undertook a legal and policy analysis of this issue, which concluded that both legal and medical realities call for a straightforward policy that provides clear guidance. The current project is a continuation of this previous work [46]. Recognizing that an essential component of future policy development is a balanced approach that fosters representation and protection of all of the parties involved, a multidisciplinary national committee was established.
Members of this committee were selected by the principal investigators (P.M.T. and J.D.M.) in consultation with several experts in the relevant fields. The 2 meetings of the committee were semi-open (i.e., they were open to interested professionals in relevant fields of expertise). The committee comprised attorneys, biomedical ethicists, and physicians. One physician committee member was a plastic surgeon who has publicly disclosed that he was occupationally infected with HCV. An advisory board of faculty was also established to provide further guidance. Other committee members are listed at the end of the text.
A first draft of the consensus statement was a synthesis of specific information gathered and provided by the individual committee members, who presented their evidence during the first meeting of the committee on 10 November 2001. This version was discussed at the second committee meeting on 18 September 2002, during which consensus was achieved on the decision chart and on the issue of disclosure. There was no significant disagreement on the statement, and further comments were incorporated into a final draft. This final consensus statement incorporates all relevant evidence obtained by the literature search in conjunction with final consensus recommendations supported by all committee members.
During the period of 1972–1994, there were >375 patients with documented transmission of disease from 42 HBV-infected health care workers (primarily surgeons or dentists) [47]. Additional clusters of HBV transmission have occurred in Europe [48] and Canada, and an outbreak of 75 cases of HBV infection occurred in the Toronto, Canada, area [49].
Transmission of HBV to patients from HBV-infected surgeons who have hepatitis B e antigen (HBeAg) present, a marker for high infectivity, has been widely documented [2–5, 10, 50]. In addition, there has also been documentation that HBV-positive surgeons without detectable levels of HBeAg transmitted HBV to patients during invasive procedures [14]. It now appears that a subset of persons infected with HBeAg-negative mutants may be as infectious as those who are HBeAg positive [51, 52], and HBeAg-negative mutants have been associated with fulminant hepatitis and more-severe chronic hepatitis than is nonmutant HBV [53].
The surgical specialties involved in transmission have tended to be those with the highest rates of percutaneous injuries and recontact. These recognized cases probably represent an underestimation of the extent of transmission, because only approximately one-half of patients with acute HBV infection are symptomatic; isolated, sporadic cases may be more difficult to link with an health care worker; and completeness of surveillance may vary among jurisdictions. Although reporting is certainly not complete, both outbreaks and sporadic transmission of HBV from surgeon to patient appear to be relatively uncommon but still greatly exceed reports of HIV and/or HCV transmission. This amounts to evidence that transmission risk is low but does exist, given the estimated pool of infected surgeons. Practice restrictions seem appropriate and reasonable and will likely affect a steadily shrinking group of HBV-infected physicians.
To date, 5 health care workers are documented to have transmitted HCV infection to a total of 232 patients [15–21]. In one case in which transmission could be traced, the infection risk amounted to somewhere between 0% and 5%. Although the true frequency of infection is unknown, the risk of transmission is probably related to the nature of the procedure and to the viral load of the infected physician. Ross et al. [18] designed a formula to predict the transmission risk per procedure and over the course of a (hypothetical) career. Interpretation of these numbers requires some caution and scrutiny, because risks are indicated per procedure/per career but not per individual patient, which, from the patient's perspective, is the truly important issue.
HCV infection is a serious danger to both health care workers and patients alike. To date, to our knowledge, far more patients have been proven to have acquired HCV infection from a health care worker, compared with HIV infection. It is thus appropriate to set evidence-based limits to the practice of HCV infected physicians.
To date, there have been 8 confirmed cases of transmission of HIV infection from 3 health care workers (a dentist, an orthopedic surgeon, and a nurse) to patients. The risk of transmission of HIV from surgeons to patients is unknown, although several estimates have been made. To date, studies have not implicated HIV-infected surgeons in the transmission of disease to their patients, and the CDC, with use of modeling techniques, estimated that the average risk of sporadic HIV transmission from an HIV-infected surgeon to a patient during an invasive procedure was 2.4–24 episodes of transmission per million procedures [54]. HIV transmission from (asymptomatic) health care workers to patients using standard antiseptic technique is quite unlikely. In the ∼20 years that HIV/AIDS has been known and studied, only 8 patients have been proven to be infected by a health care worker, one of whom was treated under questionable antiseptic circumstances [23] and another whose treatment involved a surgeon with symptomatic AIDS [22].
Practice restrictions for HIV-infected physicians are prudent, but they should necessarily be less extensive than restrictions for HBV and HCV infection because HIV is less transmissible and because of the availability of HIV postexposure prophylaxis, which appears to be more effective than postexposure prophylaxis for HCV.
We propose a decision-making chart that consists of 3 different classes of procedures (categories I–III) and the 3 pathogens (and, for the hepatitis viruses, 2 groups of serostatus [A and B]), resulting in a matrix that entails a gradation of the separate risks, each accompanied by a specific practice recommendation. The recommendations are based on procedures being performed under standard antiseptic technique, with infected physicians wearing double gloves and observing universal precautions. Physicians whose serostatus is unknown should wear a single pair of gloves during invasive procedures unless they would usually wear double gloves as standard clinical practice.
Procedures, as illustrated by tables 1–3, are divided into 3 categories.
Category I: a list of categories of procedures with de minimis risk of viral transmission.
Category II: a list of categories of procedures for which viral transmission is theoretically possible but unlikely.
Category III: a list of categories of procedures for which there is definite risk of viral transmission or that are exposure-prone procedures.
This category entails all procedures that pose de minimis risk of viral transmission to patients. In the frank absence of documented cases of viral transmission, and considering the technical aspects of these procedures, they are considered to be safe until proven otherwise. The practical nature of these procedures is such that risk of blood-blood contact between physician and patient is minimal. The surgical field and the physicians' hands are well visualized at all times. The location of the procedures is either superficial, with minimal involvement of sharps, or there are no sharps involved at all. The likelihood of transmission ever occurring is thus deemed to be negligible.
This category lists procedures for which viral transmission from health care worker to patient is unlikely, although it is theoretically not impossible. Transmission has never been documented for these procedures. The presumed low likelihood of transmission was determined on the basis of extrapolated data and procedure characteristics. The practical nature of these procedures is such that risk of blood-blood contact between physician and patient is minimal, because the operative field and the physicians' hands are well visualized, and no deep spaces are reached other than with devices (e.g., scopes and catheters); therefore, the physician's fingers and sharps are never unseen in close proximity.
This category consists of those procedures that have proven to be exposure-prone—in other words, they are procedures during which a laboratory-confirmed clinical case of viral transmission from a health care worker to a patient has occurred. Documented risks are quantified, and the original literature is referred to in the list of such procedures. This category also includes those procedures that are highly likely to be exposure-prone—that is, viral transmission is theoretically possible but has not been documented to date. Transmission risks are extrapolated from available data on the basis of similarities in mechanical aspects of procedures for which viral transmission has occurred, or the practical nature of the procedures is such that needle-stick injuries are likely to occur. These procedures are not proven to be unsafe, but the risks are considered comparable to those that are. These procedures are considered dangerous when performed by infected personnel and are off-limits for infected physicians.
The committee acknowledges that there is a frank absence of data for categorizing most surgical and medical procedures. However, by consensus it decided to arrange procedures by these 3 categories, as presented in tables 1–3, on the basis of its interdisciplinary expertise. Items can move between categories in the face of new data.
The decision chart, as illustrated by table 4, combines the transmission risks per procedure as outlined in the categories of tables 1–3 and the viral characteristics for each pathogen, which groups infected physicians in group A or group B. For the hepatitis viruses, a specific boundary was set between low and increased infectivity. Infectivity of viruses is thought to be related to the volume injected and the viral titer of the donor [55].
Decision chart for safe-practice management for infected physicians who perform invasive procedures.
For HIV, no separate groups have been made. The committee recognizes that HIV loads are important when assessing infectivity. Higher viral loads have been proven to be more infectious than lower loads, both for maternal-fetal transmission and transmission between heterosexual partners. Yet determining where the cutoff lies in the health care setting is an impossible task. Although there admittedly are no definitive data on immune impairment and transmissibility, HIV load is a more important issue to consider than symptomatic AIDS, because viral loads during symptomatic (but treated) AIDS may be lower than viral loads during the episode of asymptomatic but untreated HIV infection. However, a physician with symptomatic AIDS may be too debilitated to practice and is susceptible to infection. Such physicians should be subjected to regular medical evaluations by their treating physician to assess their overall competence to practice. HIV-positive physicians who adhere to the guidelines can practice with certain restrictions, unless they have symptoms of AIDS, as defined by the CDC [56], and are too debilitated.
Group A includes HBV-positive physicians who are negative for HBeAg and viral DNA and whose serum samples yield <103 genome equivalents/mL [57]. Serum samples obtained from HCV-positive physicians may not yield >106 virions/mL for the physician to be included in group A. Admittedly, nondetectable virus does not equate with cure or noninfectivity, but transmission risks under such circumstances are exceedingly small.
Group B includes those physicians with the following viral characteristics: for HBV-infected persons, the group includes persons who are HBeAg or HBV DNA positive or whose serum samples yield ⩾103 genome equivalents/mL; and for HCV-infected persons, the group includes those with a high viral load (defined as >106 virions/mL). The rationale here is the fact that vertical transmission of HCV has only occurred when maternal serum titers were >1 × 106 viral equivalents/mL [58]. Admittedly, this boundary is arbitrary, but it is based on reasonable available data and expert consultation (by B.W., R.G.S., and R.L.N.).
The decision chart was created with 2 different risks weighed: the risk of blood-blood contact between physician and patient, and conversely, the risk of viral transmission per virus. If a physician has active disease, no procedure of any category should be performed, and full practice restrictions should take effect until recovery (or remission to a chronic, less infectious phase of disease). Active disease is defined by an AIDS-defining syndrome for HIV-positive physicians [56] and by active hepatitis with abnormal liver enzyme findings for HBV- and/or HCV-infected physicians.
We believe that these recommendations are reasonable because they are based on a compilation of historic and recent scientific data, as well as on the guiding principles and considerations considered by experts from a variety of disciplines who reviewed the issue from different vantage points. A limitation of our proposal is the fact that the categories of procedures are admittedly rudimentary and ideally require a far more detailed and extensive inventory. Also, because our committee did not have a dentistry representative during the development of this article, we did not feel qualified to address dentists and dental procedures in the decision-making process.
The guidelines are based on current knowledge, and as new scientific data on this topic emerge, they should be revised accordingly. Although the committee was charged with developing recommendations for infected physicians, it is clear that the guidelines presented are relevant to more than just infected physicians and should be applied to all infected health care workers, such as nurses or technicians who are similarly situated with regard to posing a risk of transmitting a bloodborne pathogen to a patient during an invasive procedure.
An all-important next step in the final resolution of this issue is obviously screening. Clearly, the decision chart presupposes and requires detailed and current knowledge about viral serum levels to enable sound practice advice to be given. There is no legal requirement for health care workers to undergo periodic testing for the 3 viruses except when they have received an accidental needle-stick. However, physicians at risk for acquiring and, thus, transmitting a bloodborne disease may have a moral obligation [59] to be tested regularly for HIV, HBV, and HCV.
We acknowledge a need for testing; otherwise, the recommendations we have made would be rendered impotent. However, this issue goes beyond the charge of this committee, its resources, and the scope of this article. Although recommendations for testing are paramount, we cannot provide guidance, because testing recommendations did not withstand the same rigor of discussion and review that the recommendations herein have. However, the defining criteria for disqualification of physicians inevitably lead in the direction of screening.
Wendy K. Mariner (Boston University School of Public Health, Boston, MA), Jeffrey Kahn (Center for Bioethics, University of Minnesota, Minneapolis), James F. Childress (Institute for Practical Ethics, University of Virginia [UVA], Charlottesville), David M. Kahler and Kornelis A. Poelstra (Department of Orthopaedic Surgery, UVA), Antal Solyom (psychiatrist and graduate student in bioethics, UVA), Farhat Moazam (graduate student in religious studies, UVA), Rafael Triana (Clinical Psychiatric Medicine, UVA), and Sue McCoy (retired general surgeon and graduate student in bioethics, UVA).
Financial support. Rockefeller Brothers Fund.
Potential conflicts of interest. All authors: no conflicts.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy